paint-brush
पीएचडी करने से पहले इसे देखें। मशीन लर्निंग में द्वारा@whatsai
718 रीडिंग
718 रीडिंग

पीएचडी करने से पहले इसे देखें। मशीन लर्निंग में

द्वारा Louis Bouchard
Louis Bouchard HackerNoon profile picture

Louis Bouchard

@whatsai

I explain Artificial Intelligence terms and news to non-experts.

2 मिनट read2023/04/23
Read on Terminal Reader
Read this story in a terminal
Print this story

बहुत लंबा; पढ़ने के लिए

पीएचडी करना। मशीन लर्निंग में क्षेत्र में प्रवेश करने का एक शानदार तरीका हो सकता है। अकादमिक बनाम उद्योग के पेशेवरों और विपक्षों पर विचार करना महत्वपूर्ण है। एक मजबूत व्यक्तिगत ब्रांड और ऑनलाइन प्रोफ़ाइल के प्रभाव को कम मत समझो। अपने काम को साझा करें, सहयोग करें और अपने लाभ के लिए सोशल मीडिया का उपयोग करें।
featured image - पीएचडी करने से पहले इसे देखें। मशीन लर्निंग में
Louis Bouchard HackerNoon profile picture
Louis Bouchard

Louis Bouchard

@whatsai

I explain Artificial Intelligence terms and news to non-experts.

0-item
1-item

STORY’S CREDIBILITY

Video

Video

The best videos on the Internet archived and shared on HackerNoon.

Podcast

Podcast

The best podcasts on the Internet archived and shared on HackerNoon.

यदि आप पीएचडी करने पर विचार कर रहे हैं। मशीन लर्निंग में, या केवल "उचित" पृष्ठभूमि के बिना क्षेत्र में उतरना", यहां कुछ महत्वपूर्ण अंतर्दृष्टि हैं जिन्हें आपको निर्णय लेने से पहले जानना चाहिए।


शिक्षा बनाम उद्योग:

एक पीएच.डी. मशीन लर्निंग में क्षेत्र में प्रवेश करने और विशिष्ट अनुसंधान लक्ष्यों को प्राप्त करने का एक शानदार तरीका हो सकता है।


हालांकि, अकादमिक बनाम उद्योग के पेशेवरों और विपक्षों पर विचार करना महत्वपूर्ण है। ओपन-सोर्स रिसर्च संगठनों में शामिल होने या साइड प्रोजेक्ट्स पर काम करने जैसे वैकल्पिक रास्ते वास्तव में कुछ व्यक्तियों के लिए स्नातक अध्ययन से ज्यादा उपयोगी साबित हो सकते हैं।


यहां ब्रायन के साक्षात्कार के कुछ बेहतरीन सुझावों की स्पष्ट सूची दी गई है:

  1. रेज़्यूमे को वैयक्तिकृत करना : जब मशीन लर्निंग में अपनी पहली नौकरी देने की बात आती है, तो आपके द्वारा किए गए मात्रात्मक प्रभावों को प्रदर्शित करना महत्वपूर्ण होता है। यह किसी सेवा की गति बढ़ा सकता है, किसी मॉडल के प्रदर्शन में सुधार कर सकता है, या किसी नियोक्ता के लिए राजस्व उत्पन्न कर सकता है। इन उपलब्धियों को उजागर करने के लिए अपने रिज्यूमे को वैयक्तिकृत करना महत्वपूर्ण है।


  2. ब्रांड और ऑनलाइन उपस्थिति की शक्ति जब भर्ती की बात आती है तो एक मजबूत व्यक्तिगत ब्रांड और ऑनलाइन प्रोफ़ाइल के प्रभाव को कम मत समझो। अपने काम को साझा करें, सहयोग करें और अपने लाभ के लिए सोशल मीडिया का उपयोग करें।


  3. कनेक्शन बनाएं : किसी भी क्षेत्र में नेटवर्किंग आवश्यक है, और मशीन लर्निंग कोई अपवाद नहीं है। पुराने स्कूल के सोशल नेटवर्किंग में गोता लगाएँ और उद्योग में संबंध बनाना शुरू करें। लिंक्डइन पर लोगों तक पहुंचें, नेटवर्किंग कार्यक्रमों में भाग लें, ऑनलाइन समुदायों में शामिल हों और प्रतिस्पर्धी बाजार में अलग दिखने के लिए अतिरिक्त प्रयास करें।


  4. एक कहानी साझा करें : एक बार जब आप साक्षात्कार में शामिल हो जाते हैं, तो अपनी कहानी साझा करना महत्वपूर्ण होता है। रिक्रूटर्स ने आपका रिज्यूमे पढ़ लिया है, इसलिए उस प्रोजेक्ट के बारे में बात करें जिस पर आपने काम किया और जिन चुनौतियों को आपने पार किया। जो कुछ भी आपको बनाता है, आप!


इन जानकारियों को ब्रायन बर्न्स , पीएच.डी. द्वारा साझा किया गया था। वाशिंगटन विश्वविद्यालय में उम्मीदवार और एआई पब ट्विटर अकाउंट के संस्थापक। एआई में आने के तरीके के बारे में और भी अधिक जानकारी जानने के लिए, एक ट्विटर पेज विकसित करें, एक पॉडकास्ट होस्ट करें, इक्का-दुक्का साक्षात्कार करें, और एक बेहतर रिज्यूमे बनाएं, नीचे दिए गए पूर्ण पॉडकास्ट में ट्यून करें!


पॉडकास्ट देखें


foreign

[00:00:00] : [00:00:03]

this is an interview with Brian Burns

[00:00:03] : [00:00:12]

PhD candidate at the University of

[00:00:12] : [00:00:14]

Washington and founder of the AI Pub

[00:00:14] : [00:00:16]

Twitter account now if over 60 000

[00:00:16] : [00:00:19]

followers Brian is trying to build a

[00:00:19] : [00:00:21]

talent referral business out of it Brian

[00:00:21] : [00:00:23]

also hosts a podcast where he explained

[00:00:23] : [00:00:25]

research papers with the authors in this

[00:00:25] : [00:00:28]

interview we go over all these different

[00:00:28] : [00:00:30]

facets the PHD I think a Twitter account

[00:00:30] : [00:00:34]

the podcast and more I hope you enjoy it

[00:00:34] : [00:00:37]

who are you and what's your background

[00:00:37] : [00:00:38]

sure

[00:00:38] : [00:00:40]

um my name is Brian I run this Twitter

[00:00:40] : [00:00:42]

account called aipob an Associated

[00:00:42] : [00:00:45]

podcast called Deep papers

[00:00:45] : [00:00:45]

um

[00:00:45] : [00:00:48]

what do I do

[00:00:48] : [00:00:49]

um with the with the Twitter account I

[00:00:49] : [00:00:52]

cover like high level

[00:00:52] : [00:00:55]

technical AI topics so it's like AI

[00:00:55] : [00:00:57]

research but not going like super deep

[00:00:57] : [00:00:59]

in the weeds but also for people who

[00:00:59] : [00:01:01]

have a technical background

[00:01:01] : [00:01:03]

um and then and then the podcast is like

[00:01:03] : [00:01:06]

interviewing uh AI scientists of like

[00:01:06] : [00:01:08]

recent like very interesting papers also

[00:01:08] : [00:01:11]

like on a technical but high level

[00:01:11] : [00:01:14]

um my background is a is uh before that

[00:01:14] : [00:01:16]

I'm I'm on leave from a machine learning

[00:01:16] : [00:01:18]

PhD within the stats department at uh at

[00:01:18] : [00:01:22]

UW I I did a year of a stats PhD uh

[00:01:22] : [00:01:23]

picked up a lot of like deep learning

[00:01:23] : [00:01:25]

stuff some computer vision stuff

[00:01:25] : [00:01:28]

um and started this Twitter account it

[00:01:28] : [00:01:29]

blew up and I'm kind of I'm kind of

[00:01:29] : [00:01:31]

running with it so that's uh that's my

[00:01:31] : [00:01:33]

background we'll definitely dive into

[00:01:33] : [00:01:35]

the podcast and the Twitter account

[00:01:35] : [00:01:38]

which is my main interest here but I

[00:01:38] : [00:01:40]

also want to take a small detour into

[00:01:40] : [00:01:42]

the PHD just because I'm myself doing

[00:01:42] : [00:01:44]

one and I know that a lot of people

[00:01:44] : [00:01:46]

listening are either doing one or

[00:01:46] : [00:01:49]

considering doing one so I'd love to

[00:01:49] : [00:01:52]

know why well first what was your your

[00:01:52] : [00:01:55]

reason in starting one

[00:01:55] : [00:01:57]

yeah I mean for me I feel like my

[00:01:57] : [00:01:59]

situation is kind of weird like for me I

[00:01:59] : [00:02:01]

wanted to break into machine learning uh

[00:02:01] : [00:02:03]

I didn't have I had a programming

[00:02:03] : [00:02:04]

background I've been programming since I

[00:02:04] : [00:02:07]

was a little kid or a teenager

[00:02:07] : [00:02:08]

um they had very very strong math

[00:02:08] : [00:02:10]

backgrounds so like growing up I wanted

[00:02:10] : [00:02:12]

to be a mathematician and like in

[00:02:12] : [00:02:14]

University it was kind of on on the

[00:02:14] : [00:02:16]

track to like do a math PhD

[00:02:16] : [00:02:17]

um I wanted to break in machine learning

[00:02:17] : [00:02:21]

like uh I thought that basically like

[00:02:21] : [00:02:24]

doing a statistics PhD like in a

[00:02:24] : [00:02:25]

department that had a really good

[00:02:25] : [00:02:27]

machine learning program would be a good

[00:02:27] : [00:02:28]

way to do that

[00:02:28] : [00:02:30]

um so that's like so I ended up applying

[00:02:30] : [00:02:32]

to a bunch of grad grad schools and I

[00:02:32] : [00:02:33]

got into UW

[00:02:33] : [00:02:36]

um so that's kind of like why I started

[00:02:36] : [00:02:37]

it

[00:02:37] : [00:02:39]

so to transition from mathematics into

[00:02:39] : [00:02:41]

something more applied

[00:02:41] : [00:02:43]

yeah I I just uh I I wanted to like

[00:02:43] : [00:02:45]

learn a bunch and like grad school is

[00:02:45] : [00:02:47]

like a really cool opportunity where you

[00:02:47] : [00:02:50]

know you get paid enough to pay rent and

[00:02:50] : [00:02:52]

buy food and you know you can use a lot

[00:02:52] : [00:02:53]

of your time not all of your time but a

[00:02:53] : [00:02:56]

lot of your time to to to learn um so

[00:02:56] : [00:02:58]

for me it was really useful

[00:02:58] : [00:03:00]

what would you say that it was maybe

[00:03:00] : [00:03:03]

more useful than directly going into

[00:03:03] : [00:03:05]

some industry roles like what's the

[00:03:05] : [00:03:07]

what's the difference for you or or why

[00:03:07] : [00:03:10]

did you choose the The Graduate Studies

[00:03:10] : [00:03:13]

path instead of going straight into the

[00:03:13] : [00:03:15]

industry yeah I mean honestly I think

[00:03:15] : [00:03:18]

that's a good question I I think in

[00:03:18] : [00:03:19]

hindsight

[00:03:19] : [00:03:21]

I think in hindsight I didn't understand

[00:03:21] : [00:03:23]

the breadth of opportunities that are

[00:03:23] : [00:03:24]

available

[00:03:24] : [00:03:27]

um I definitely think that having the

[00:03:27] : [00:03:29]

credential really helps even if even if

[00:03:29] : [00:03:31]

you do a fraction of a PhD like I have

[00:03:31] : [00:03:33]

I'm on leave I almost certainly am not

[00:03:33] : [00:03:34]

going to finish

[00:03:34] : [00:03:37]

um you know having that credential or

[00:03:37] : [00:03:38]

having you know master's degree or

[00:03:38] : [00:03:41]

something it does help but I do think

[00:03:41] : [00:03:42]

there are kind of a lot of ways to break

[00:03:42] : [00:03:44]

in and I

[00:03:44] : [00:03:46]

I definitely don't think that the PHD is

[00:03:46] : [00:03:49]

necessary at this point especially given

[00:03:49] : [00:03:51]

how quickly quickly things are moving

[00:03:51] : [00:03:52]

and how quickly kind of the dominant

[00:03:52] : [00:03:55]

paradigms of AI and machine learning are

[00:03:55] : [00:03:57]

changing um I I don't think it's I don't

[00:03:57] : [00:03:59]

think it's necessary so it definitely

[00:03:59] : [00:04:01]

like in hindsight I don't think it's

[00:04:01] : [00:04:04]

necessarily the optimal path and and I

[00:04:04] : [00:04:05]

think for for a lot of people it's

[00:04:05] : [00:04:07]

definitely not yeah and do you think

[00:04:07] : [00:04:10]

regarding the optimal path do you think

[00:04:10] : [00:04:12]

like of course the credential helps but

[00:04:12] : [00:04:14]

do you think that the four years put

[00:04:14] : [00:04:16]

into it well like minimum of four years

[00:04:16] : [00:04:19]

or something like this put into it is

[00:04:19] : [00:04:21]

worth the four years that you could have

[00:04:21] : [00:04:22]

put into

[00:04:22] : [00:04:25]

either working on your own thing or

[00:04:25] : [00:04:27]

working at a startup because let's say

[00:04:27] : [00:04:29]

we you won't be able to enter Google

[00:04:29] : [00:04:31]

right away if you don't have studies but

[00:04:31] : [00:04:34]

like do you think the time invested in

[00:04:34] : [00:04:36]

the PHD is worth the certification

[00:04:36] : [00:04:38]

I think it just really depends on what

[00:04:38] : [00:04:39]

your end goals are

[00:04:39] : [00:04:41]

um because I think I think there are a

[00:04:41] : [00:04:44]

lot of reasons people will pursue a PhD

[00:04:44] : [00:04:46]

um or kind of you won't want to do

[00:04:46] : [00:04:48]

something in machine learning I mean if

[00:04:48] : [00:04:50]

you're if your goal is just to like

[00:04:50] : [00:04:53]

learn and do novel research in machine

[00:04:53] : [00:04:55]

learning I think like if you can get

[00:04:55] : [00:04:58]

into a PhD program or somehow fund

[00:04:58] : [00:04:59]

yourself as an independent research

[00:04:59] : [00:05:01]

researcher uh or even like work with

[00:05:01] : [00:05:03]

these um

[00:05:03] : [00:05:05]

these open source Labs like like

[00:05:05] : [00:05:06]

stability I I think that's the best path

[00:05:06] : [00:05:08]

uh like if you just want to do research

[00:05:08] : [00:05:10]

like there's kind of not

[00:05:10] : [00:05:13]

there's not a huge societal slot for you

[00:05:13] : [00:05:16]

unless you either do a PhD or you you

[00:05:16] : [00:05:18]

join one of these open source research

[00:05:18] : [00:05:19]

Works

[00:05:19] : [00:05:22]

um I think for many of the other things

[00:05:22] : [00:05:24]

you want to do I don't think it's

[00:05:24] : [00:05:27]

optimal like I think I think if you want

[00:05:27] : [00:05:28]

to

[00:05:28] : [00:05:30]

just like get on the ground running and

[00:05:30] : [00:05:32]

learn how a lot of these modern machine

[00:05:32] : [00:05:34]

learning systems work and like you know

[00:05:34] : [00:05:36]

find good work in the area I think you

[00:05:36] : [00:05:37]

can you can learn a lot of this stuff

[00:05:37] : [00:05:39]

really quickly on your own um or even on

[00:05:39] : [00:05:40]

your job

[00:05:40] : [00:05:42]

um so I think for the vast majority of

[00:05:42] : [00:05:45]

situations I I wouldn't recommend it but

[00:05:45] : [00:05:47]

I think if you wanted if you if you're

[00:05:47] : [00:05:48]

really dead set on doing research and

[00:05:48] : [00:05:51]

you can get into a good PhD program

[00:05:51] : [00:05:52]

um I think it's probably the optimal

[00:05:52] : [00:05:53]

path

[00:05:53] : [00:05:54]

um

[00:05:54] : [00:05:55]

and then and then the other thing is

[00:05:55] : [00:05:57]

like again like the world is changing so

[00:05:57] : [00:05:58]

much like now there are these really

[00:05:58] : [00:06:00]

cool open source Labs like I have a lot

[00:06:00] : [00:06:01]

of friends who are like software

[00:06:01] : [00:06:03]

Engineers or machine learning Engineers

[00:06:03] : [00:06:05]

where they want to increase their

[00:06:05] : [00:06:07]

research shops so what they do is

[00:06:07] : [00:06:08]

instead of leaving for grad school they

[00:06:08] : [00:06:10]

just start working with like Carper AI

[00:06:10] : [00:06:11]

in their spare time

[00:06:11] : [00:06:13]

um so I think even if you're trying to

[00:06:13] : [00:06:14]

do research I think there are alternate

[00:06:14] : [00:06:16]

paths there

[00:06:16] : [00:06:19]

yeah definitely and I I don't know what

[00:06:19] : [00:06:21]

will happen with universities in general

[00:06:21] : [00:06:26]

but we will see and I I also think well

[00:06:26] : [00:06:28]

I assume that I assume it's obvious that

[00:06:28] : [00:06:30]

if you want to become a professor as

[00:06:30] : [00:06:32]

well it's pretty much necessary like if

[00:06:32] : [00:06:35]

if your goal is yet to teach it it

[00:06:35] : [00:06:37]

definitely helps and well it's not

[00:06:37] : [00:06:39]

necessary you can become like an online

[00:06:39] : [00:06:41]

Professor or something but

[00:06:41] : [00:06:43]

if you want to teach at at University

[00:06:43] : [00:06:48]

it's definitely required and to me that

[00:06:48] : [00:06:51]

well after speaking with many people I

[00:06:51] : [00:06:54]

think even doing research as you said

[00:06:54] : [00:06:58]

might be might not be the optimal path

[00:06:58] : [00:07:00]

the PHD might not be the table path you

[00:07:00] : [00:07:01]

may

[00:07:01] : [00:07:03]

be even better if you if you are able to

[00:07:03] : [00:07:05]

find some research position elsewhere

[00:07:05] : [00:07:09]

and the yeah the main reason to do a PhD

[00:07:09] : [00:07:12]

might be to become a professor and yep

[00:07:12] : [00:07:14]

the second one to to do research which

[00:07:14] : [00:07:17]

may not even be the the optimal path to

[00:07:17] : [00:07:18]

do that yeah I mean I think if you want

[00:07:18] : [00:07:20]

to do if you want to do research doing a

[00:07:20] : [00:07:22]

PhD is kind of the mainline way to do it

[00:07:22] : [00:07:24]

like I think there are alternate pads

[00:07:24] : [00:07:27]

but like whatever 80 90 percent of

[00:07:27] : [00:07:28]

people who are trying to pursue research

[00:07:28] : [00:07:31]

and who can get into some kind of PhD

[00:07:31] : [00:07:33]

program like I think that's probably the

[00:07:33] : [00:07:34]

safest bet and the most straightforward

[00:07:34] : [00:07:36]

way to do it that would be my guess

[00:07:36] : [00:07:38]

yeah it's also

[00:07:38] : [00:07:40]

but it's definitely worth doing if you

[00:07:40] : [00:07:43]

enjoy doing it like it's it's definitely

[00:07:43] : [00:07:46]

if you like to to control your own work

[00:07:46] : [00:07:49]

and pace and for example if you want to

[00:07:49] : [00:07:51]

take eight years to do the well I don't

[00:07:51] : [00:07:53]

know what's the maximum length but if

[00:07:53] : [00:07:55]

you want to do for example seven years

[00:07:55] : [00:07:58]

instead of four but work 20 hours per

[00:07:58] : [00:08:02]

week I most times you can well here in

[00:08:02] : [00:08:06]

Canada you can so like it's definitely a

[00:08:06] : [00:08:09]

good thing if you are more I don't know

[00:08:09] : [00:08:12]

that's a bit weird but I I can't see a

[00:08:12] : [00:08:13]

comparison between the PHD and

[00:08:13] : [00:08:15]

Entrepreneurship where you pretty much

[00:08:15] : [00:08:17]

decide on what you are doing and you

[00:08:17] : [00:08:19]

decide on your schedule and you work on

[00:08:19] : [00:08:21]

what you want to work so

[00:08:21] : [00:08:25]

that's that's also pretty cool and the

[00:08:25] : [00:08:27]

yeah the the flexibility that you have

[00:08:27] : [00:08:30]

is amazing to work on side projects just

[00:08:30] : [00:08:33]

like yours so I feel like it it may be a

[00:08:33] : [00:08:34]

good like

[00:08:34] : [00:08:38]

safe packet plan while trying other

[00:08:38] : [00:08:40]

stuff at the same time yeah I agree with

[00:08:40] : [00:08:43]

that I agree with that 100

[00:08:43] : [00:08:43]

and so you mentioned it but I assume you

[00:08:43] : [00:08:51]

most certainly will not go back to the

[00:08:51] : [00:08:54]

PHD and continue it

[00:08:54] : [00:08:56]

yeah I would say so

[00:08:56] : [00:08:57]

um I think

[00:08:57] : [00:08:59]

I think I've kind of explored this weird

[00:08:59] : [00:09:02]

alternate path uh of you know like

[00:09:02] : [00:09:03]

social media is just so cool like I just

[00:09:03] : [00:09:05]

have met all these different people in

[00:09:05] : [00:09:06]

social media over the last

[00:09:06] : [00:09:08]

six months or so seven months and

[00:09:08] : [00:09:11]

starting the account like I you know I

[00:09:11] : [00:09:12]

have this little recruiting business

[00:09:12] : [00:09:13]

that's going well

[00:09:13] : [00:09:15]

um I just there's so many kind of

[00:09:15] : [00:09:17]

opportunities that I feel like are here

[00:09:17] : [00:09:20]

um that I think at least at least for me

[00:09:20] : [00:09:23]

I think like I don't I don't necessarily

[00:09:23] : [00:09:25]

want to do another four years

[00:09:25] : [00:09:27]

yeah that makes sense yeah but it's very

[00:09:27] : [00:09:29]

very different because I I think for me

[00:09:29] : [00:09:30]

like I entered the PHD to kind of break

[00:09:30] : [00:09:33]

into machine learning and I think after

[00:09:33] : [00:09:34]

the first like six months to a year I

[00:09:34] : [00:09:36]

felt like I had learned enough to go out

[00:09:36] : [00:09:37]

on my own

[00:09:37] : [00:09:40]

um but but I think like I was not dead

[00:09:40] : [00:09:42]

set on like doing you know world-class

[00:09:42] : [00:09:44]

language modeling research or something

[00:09:44] : [00:09:45]

like that and I think I think for

[00:09:45] : [00:09:47]

someone else if if that's if that's your

[00:09:47] : [00:09:49]

aim then then staying is really the

[00:09:49] : [00:09:50]

right move

[00:09:50] : [00:09:53]

yeah and it's definitely something that

[00:09:53] : [00:09:56]

you are like in a PhD you are

[00:09:56] : [00:09:57]

specializing yourself in something

[00:09:57] : [00:10:01]

extremely specific and that's well that

[00:10:01] : [00:10:02]

that brings value but it may not be for

[00:10:02] : [00:10:05]

everyone yeah and it's also I mean this

[00:10:05] : [00:10:06]

is one of the things that like blew my

[00:10:06] : [00:10:09]

mind the most about uh machine learning

[00:10:09] : [00:10:11]

kind of the history machine learning is

[00:10:11] : [00:10:12]

um is almost the trend of

[00:10:12] : [00:10:15]

anti-specialization with time like most

[00:10:15] : [00:10:18]

Fields as they go on you know they get

[00:10:18] : [00:10:20]

more and more Niche it's like if you do

[00:10:20] : [00:10:22]

like a physics PhD or something you know

[00:10:22] : [00:10:25]

you specialize in like

[00:10:25] : [00:10:28]

theoretical Quark String Theory type B

[00:10:28] : [00:10:30]

or something and you do like seven years

[00:10:30] : [00:10:31]

on that whereas the really interesting

[00:10:31] : [00:10:33]

thing with with ML over the last like 10

[00:10:33] : [00:10:35]

years you know post post deep learning

[00:10:35] : [00:10:37]

and especially with the kind of like uh

[00:10:37] : [00:10:39]

you know Transformer sequence modeling

[00:10:39] : [00:10:41]

Paradigm is like this unification of

[00:10:41] : [00:10:43]

architectures right where it's like yeah

[00:10:43] : [00:10:45]

you know I I'm not an expert on this

[00:10:45] : [00:10:46]

stuff but it's like 10 years ago like

[00:10:46] : [00:10:48]

people were doing computer vision people

[00:10:48] : [00:10:49]

who were doing like NLP were doing like

[00:10:49] : [00:10:51]

very very different things

[00:10:51] : [00:10:53]

um and now like a lot of the methods are

[00:10:53] : [00:10:56]

actually becoming more and more similar

[00:10:56] : [00:10:57]

um so it's it's also kind of an

[00:10:57] : [00:10:58]

interesting time to like think about

[00:10:58] : [00:11:00]

specialization or like really deep

[00:11:00] : [00:11:01]

specialization

[00:11:01] : [00:11:03]

you know

[00:11:03] : [00:11:06]

yeah and I also I've read a great book

[00:11:06] : [00:11:09]

well listen to a great audiobook about

[00:11:09] : [00:11:11]

uh I don't remember the name exactly but

[00:11:11] : [00:11:13]

it's like why generalist Triumph or

[00:11:13] : [00:11:15]

something like this which was really

[00:11:15] : [00:11:19]

interesting and about how

[00:11:19] : [00:11:22]

generators are often the ones that that

[00:11:22] : [00:11:24]

succeeds even in in some very specific

[00:11:24] : [00:11:27]

fields for example I remember one of the

[00:11:27] : [00:11:30]

examples he gave was Roger Federer the

[00:11:30] : [00:11:33]

tennis player he they in in the book he

[00:11:33] : [00:11:36]

told he he wrote about his background

[00:11:36] : [00:11:39]

and said that like

[00:11:39] : [00:11:42]

um compared to other tennis professional

[00:11:42] : [00:11:46]

players Roger Federer where was playing

[00:11:46] : [00:11:47]

like

[00:11:47] : [00:11:50]

every other sports and some tennis until

[00:11:50] : [00:11:53]

he was yeah I would say random numbers

[00:11:53] : [00:11:56]

but until his things or something

[00:11:56] : [00:11:59]

compared to others that like at six they

[00:11:59] : [00:12:01]

already started only playing tennis and

[00:12:01] : [00:12:04]

nothing else whereas he was playing a

[00:12:04] : [00:12:06]

lot of different sports and developing

[00:12:06] : [00:12:08]

different skills and so I I don't know

[00:12:08] : [00:12:10]

why I'm biased towards the

[00:12:10] : [00:12:13]

generalization because I just like

[00:12:13] : [00:12:15]

touching a lot of stuff but I feel like

[00:12:15] : [00:12:17]

there's there's definitely some value to

[00:12:17] : [00:12:20]

to staying a bit more general or at

[00:12:20] : [00:12:21]

least

[00:12:21] : [00:12:23]

even if you are focused on one area

[00:12:23] : [00:12:26]

still getting out of your way to

[00:12:26] : [00:12:28]

understand the different things that are

[00:12:28] : [00:12:30]

happening or developing in other

[00:12:30] : [00:12:31]

Industries

[00:12:31] : [00:12:34]

oh yeah I mean yeah I kind of I see this

[00:12:34] : [00:12:36]

a little bit in recruiting as well like

[00:12:36] : [00:12:38]

I do Talent referrals to different AI

[00:12:38] : [00:12:41]

startups and I think like the candidates

[00:12:41] : [00:12:43]

that I find like get the most job offers

[00:12:43] : [00:12:45]

or like end up like progressing in their

[00:12:45] : [00:12:47]

careers the fastest they're definitely

[00:12:47] : [00:12:48]

not like total generalists like they

[00:12:48] : [00:12:50]

tend to be really good at one thing

[00:12:50] : [00:12:51]

right yeah

[00:12:51] : [00:12:52]

um you know they tend to be like really

[00:12:52] : [00:12:54]

really good at NLP and like have a long

[00:12:54] : [00:12:56]

history of projects and like you know

[00:12:56] : [00:12:58]

successful work history doing that kind

[00:12:58] : [00:12:59]

of thing but I think it's kind of like

[00:12:59] : [00:13:01]

you're they're not entirely t-shaped

[00:13:01] : [00:13:02]

it's like they're really good at one

[00:13:02] : [00:13:05]

thing but then for basically everything

[00:13:05] : [00:13:07]

else they also are like good enough or

[00:13:07] : [00:13:09]

or actually like even a little bit

[00:13:09] : [00:13:11]

beyond that so it's like they're really

[00:13:11] : [00:13:13]

good at NLP but then like they're also

[00:13:13] : [00:13:15]

like really fun to work with and you

[00:13:15] : [00:13:17]

know they have at least some social

[00:13:17] : [00:13:19]

media profile showing off their work and

[00:13:19] : [00:13:20]

like they Network every once in a while

[00:13:20] : [00:13:23]

you know so that I I I've noticed like

[00:13:23] : [00:13:26]

that that additional kind of marginal

[00:13:26] : [00:13:28]

stuff kind of across the skill

[00:13:28] : [00:13:30]

distribution like really helps people in

[00:13:30] : [00:13:33]

their careers yeah or even for example

[00:13:33] : [00:13:36]

just just for myself in my internships

[00:13:36] : [00:13:39]

and things like that if you are good at

[00:13:39] : [00:13:41]

something else it can also help just I I

[00:13:41] : [00:13:45]

remember that I I

[00:13:45] : [00:13:48]

we I wrote in my resume that I was doing

[00:13:48] : [00:13:51]

some climbing and like one of the

[00:13:51] : [00:13:53]

interviewers were was a big climber

[00:13:53] : [00:13:55]

climbing fan and we just talked about

[00:13:55] : [00:13:57]

this during the whole interview and I

[00:13:57] : [00:13:58]

ended up getting the job and we just

[00:13:58] : [00:14:00]

went climbing and it was

[00:14:00] : [00:14:02]

perfect like I I feel like a lot of

[00:14:02] : [00:14:05]

recruiting is based on

[00:14:05] : [00:14:09]

and not preferences but like just fine

[00:14:09] : [00:14:10]

it's just like finding a friend

[00:14:10] : [00:14:12]

basically you want to match with the

[00:14:12] : [00:14:15]

person you like the person can learn and

[00:14:15] : [00:14:17]

as long as he's he or she is motivate

[00:14:17] : [00:14:21]

motivated it's it should be fine so I so

[00:14:21] : [00:14:22]

you definitely try to

[00:14:22] : [00:14:25]

to find a good fit but you are the

[00:14:25] : [00:14:29]

recruiter expert so what what would you

[00:14:29] : [00:14:33]

say that it's most important when when

[00:14:33] : [00:14:36]

matching like it I don't know matching a

[00:14:36] : [00:14:39]

company with with an individual like or

[00:14:39] : [00:14:42]

what what would you say that companies

[00:14:42] : [00:14:44]

are looking for mostly is it strictly

[00:14:44] : [00:14:46]

the skills or the personality is even

[00:14:46] : [00:14:48]

more important what's the most important

[00:14:48] : [00:14:50]

thing oh man that's that's so hard

[00:14:50] : [00:14:52]

because I don't think there's any one

[00:14:52] : [00:14:53]

specific thing and I think there's also

[00:14:53] : [00:14:55]

a selection for a lot of different roles

[00:14:55] : [00:14:57]

right like I think what kids you had as

[00:14:57] : [00:14:58]

a research engineer is different from

[00:14:58] : [00:15:00]

what gets you out as a software engineer

[00:15:00] : [00:15:03]

research scientist so I do think it's I

[00:15:03] : [00:15:05]

think it varies a lot across roles

[00:15:05] : [00:15:07]

um but let me let me answer directly

[00:15:07] : [00:15:09]

um

[00:15:09] : [00:15:09]

so I think so there's there's kind of

[00:15:09] : [00:15:14]

like two levels of this like I think uh

[00:15:14] : [00:15:16]

you know first and foremost I think just

[00:15:16] : [00:15:19]

like having impressive technical

[00:15:19] : [00:15:21]

accomplishments on your resume is really

[00:15:21] : [00:15:22]

useful like this is just very standard

[00:15:22] : [00:15:24]

recruiting advice but it's just like

[00:15:24] : [00:15:26]

you've shipped a bunch of projects with

[00:15:26] : [00:15:29]

like very quantifiable impact yeah you

[00:15:29] : [00:15:30]

know n dollars a year for your business

[00:15:30] : [00:15:33]

for your employer or you know you sped

[00:15:33] : [00:15:37]

up sped up X service by X percent or you

[00:15:37] : [00:15:39]

know you improved X models performance

[00:15:39] : [00:15:42]

from you know A to B

[00:15:42] : [00:15:43]

um so I think that stuff matters first

[00:15:43] : [00:15:46]

and foremost I hate to say it but I

[00:15:46] : [00:15:48]

think second is like brand I think

[00:15:48] : [00:15:51]

recruiter like people who are hiring

[00:15:51] : [00:15:53]

CEOs or startups you know or careers

[00:15:53] : [00:15:55]

like the the name on the resume matters

[00:15:55] : [00:15:58]

a lot you know like if you're working at

[00:15:58] : [00:16:00]

whatever Google Facebook or like you

[00:16:00] : [00:16:03]

know a big startup uh stripe Airbnb or

[00:16:03] : [00:16:05]

something like um that it really greases

[00:16:05] : [00:16:07]

the wheels to to get you in the door

[00:16:07] : [00:16:08]

um you know whether whether or not

[00:16:08] : [00:16:10]

people like that the other thing to say

[00:16:10] : [00:16:12]

about the brand stuff is I think um I

[00:16:12] : [00:16:14]

think it's more in your control than you

[00:16:14] : [00:16:15]

think

[00:16:15] : [00:16:17]

um so it's like it's harder to you know

[00:16:17] : [00:16:18]

change whatever company like maybe you

[00:16:18] : [00:16:20]

can't work for a really prestigious

[00:16:20] : [00:16:21]

company but

[00:16:21] : [00:16:23]

um I think if you like if you end up

[00:16:23] : [00:16:25]

going on social media and you make a

[00:16:25] : [00:16:27]

profile for yourself and you like share

[00:16:27] : [00:16:28]

really cool work of yours you have like

[00:16:28] : [00:16:30]

a really slick website

[00:16:30] : [00:16:31]

um there's a way that that actually can

[00:16:31] : [00:16:33]

really like alter people's perceptions

[00:16:33] : [00:16:35]

of you and like show off stuff that

[00:16:35] : [00:16:36]

doesn't show in the resume so I think I

[00:16:36] : [00:16:39]

think second is brand like below the

[00:16:39] : [00:16:41]

technical stuff

[00:16:41] : [00:16:43]

um and then I think third I think third

[00:16:43] : [00:16:44]

the thing that really helps with

[00:16:44] : [00:16:46]

recruiting is kind of exactly what you

[00:16:46] : [00:16:47]

said it's actually a lot of this like

[00:16:47] : [00:16:52]

really like non-scalable social stuff

[00:16:52] : [00:16:54]

um it's really funny I think and again

[00:16:54] : [00:16:55]

I'm like not an expert with recruiting

[00:16:55] : [00:16:57]

I've been doing this for maybe six

[00:16:57] : [00:16:59]

months or something five months

[00:16:59] : [00:17:01]

um you know there's all this stuff like

[00:17:01] : [00:17:03]

there's job boards you know there's the

[00:17:03] : [00:17:06]

job application page and

[00:17:06] : [00:17:08]

one of the things I've learned is that

[00:17:08] : [00:17:09]

just like a lot of this stuff just gets

[00:17:09] : [00:17:11]

totally ignored

[00:17:11] : [00:17:13]

um there's just so much volume on the

[00:17:13] : [00:17:15]

internet that like these job boards or

[00:17:15] : [00:17:16]

these these job applications which get

[00:17:16] : [00:17:18]

totally filled up and often the the CEOs

[00:17:18] : [00:17:20]

of the hiring managers are so

[00:17:20] : [00:17:21]

overwhelmed that they just don't even

[00:17:21] : [00:17:24]

look at the uh at at the applications at

[00:17:24] : [00:17:26]

all like it's it's really surprising so

[00:17:26] : [00:17:28]

like the funny thing is kind of I think

[00:17:28] : [00:17:30]

the thing I've learned about recruiting

[00:17:30] : [00:17:31]

and like job stuff is like if you want

[00:17:31] : [00:17:33]

to get good at it you have to get good

[00:17:33] : [00:17:35]

at like all the old school stuff so just

[00:17:35] : [00:17:37]

like networking or like finding out who

[00:17:37] : [00:17:38]

the hiring manager is on LinkedIn and

[00:17:38] : [00:17:40]

like sending them a really nice cold

[00:17:40] : [00:17:43]

email about why you think you're great

[00:17:43] : [00:17:45]

for this position so that kind of thing

[00:17:45] : [00:17:47]

um so to recap One technical stuff

[00:17:47] : [00:17:51]

matters the most to Brand three like

[00:17:51] : [00:17:54]

non-scalable old school social stuff I

[00:17:54] : [00:17:56]

think referrals to like like having a

[00:17:56] : [00:17:57]

friend who's part of the company or

[00:17:57] : [00:17:58]

something like that

[00:17:58] : [00:18:01]

and I'd say like fourth that is outside

[00:18:01] : [00:18:03]

of your CV would be your online presence

[00:18:03] : [00:18:05]

as you mentioned just it's really

[00:18:05] : [00:18:08]

important to to be to be visible

[00:18:08] : [00:18:11]

basically it's it's not it's I mean I'm

[00:18:11] : [00:18:12]

sure you've experienced this like just

[00:18:12] : [00:18:14]

the kind of opportunities it opens up or

[00:18:14] : [00:18:16]

the people you mean it's totally totally

[00:18:16] : [00:18:19]

Bonkers totally bonkers

[00:18:19] : [00:18:22]

yeah it's really like if you guys are

[00:18:22] : [00:18:25]

not on either Twitter or LinkedIn just

[00:18:25] : [00:18:27]

go create an account right now and at

[00:18:27] : [00:18:29]

least share what you've done in the past

[00:18:29] : [00:18:32]

year or something just just be online a

[00:18:32] : [00:18:34]

little bit and if you are coding on

[00:18:34] : [00:18:36]

GitHub you can also put it to just

[00:18:36] : [00:18:39]

public and if you can put your code to

[00:18:39] : [00:18:42]

public and clean it a bit before but

[00:18:42] : [00:18:44]

yeah it's

[00:18:44] : [00:18:46]

really a game changer to to have an

[00:18:46] : [00:18:48]

online presence especially in this new

[00:18:48] : [00:18:51]

field that where we see for example I've

[00:18:51] : [00:18:52]

interviewed someone

[00:18:52] : [00:18:55]

um the VP of of research and development

[00:18:55] : [00:18:59]

at a startup called did and they he said

[00:18:59] : [00:19:01]

that for a research scientist role they

[00:19:01] : [00:19:05]

they received in the thousands of like

[00:19:05] : [00:19:07]

um potential candidates that were

[00:19:07] : [00:19:10]

qualified that's just crazy like they

[00:19:10] : [00:19:12]

cannot go through all of them so

[00:19:12] : [00:19:15]

definitely you need to stand out one way

[00:19:15] : [00:19:17]

or the other and as you said either by

[00:19:17] : [00:19:20]

reaching out and sending a personalized

[00:19:20] : [00:19:22]

letter or at least use chatgpt or

[00:19:22] : [00:19:24]

something to write it like more personal

[00:19:24] : [00:19:27]

but like not not spam them but at least

[00:19:27] : [00:19:30]

send something that seems personal is is

[00:19:30] : [00:19:34]

already a good step to to to get out of

[00:19:34] : [00:19:37]

the out of the lot out of the the bunch

[00:19:37] : [00:19:40]

so yep

[00:19:40] : [00:19:43]

so you mentioned that yeah we talked

[00:19:43] : [00:19:45]

about recruiting and talent referral

[00:19:45] : [00:19:47]

your what is I believe your current

[00:19:47] : [00:19:50]

business so or like main business model

[00:19:50] : [00:19:53]

maybe but what does this mean exactly

[00:19:53] : [00:19:55]

what what are you doing and how do you

[00:19:55] : [00:19:57]

do that how how can you

[00:19:57] : [00:20:00]

how are you matching people and finding

[00:20:00] : [00:20:02]

candidates and everything around that

[00:20:02] : [00:20:04]

yeah it's um it's it's actually quite

[00:20:04] : [00:20:08]

simple and I uh a lot of it I'm still

[00:20:08] : [00:20:09]

experimenting with it and I think there

[00:20:09] : [00:20:10]

are a lot of different things that can

[00:20:10] : [00:20:12]

be done but

[00:20:12] : [00:20:15]

um at least for now how it works is I

[00:20:15] : [00:20:17]

just have a link on my Twitter page and

[00:20:17] : [00:20:19]

people can click on the link and they'll

[00:20:19] : [00:20:20]

you know submit some information about

[00:20:20] : [00:20:22]

themselves their name you know what what

[00:20:22] : [00:20:24]

company they're working at like what

[00:20:24] : [00:20:26]

they're looking for in their next role

[00:20:26] : [00:20:29]

um and then you know typically like if

[00:20:29] : [00:20:33]

if they uh if they uh meet certain

[00:20:33] : [00:20:34]

parameters and I'm explicit about this

[00:20:34] : [00:20:36]

stuff it's like you know fortunately or

[00:20:36] : [00:20:39]

unfortunately uh almost all the startups

[00:20:39] : [00:20:41]

are referred to are in the US so it's

[00:20:41] : [00:20:43]

like if they're in the US and they have

[00:20:43] : [00:20:44]

like two years of experience

[00:20:44] : [00:20:46]

um and they submit that stuff I'll reach

[00:20:46] : [00:20:47]

out to them for for like a phone or a

[00:20:47] : [00:20:48]

zoom call just to get a better sense of

[00:20:48] : [00:20:50]

what they're looking for

[00:20:50] : [00:20:51]

um and then they're just kind of like a

[00:20:51] : [00:20:54]

passive Matchmaker so like uh at that

[00:20:54] : [00:20:55]

point maybe you'll really understand

[00:20:55] : [00:20:56]

what they're looking for and already

[00:20:56] : [00:20:58]

know a company that's that's a good

[00:20:58] : [00:20:59]

match

[00:20:59] : [00:21:00]

um or it's happened where it's like

[00:21:00] : [00:21:02]

WeChat I don't actually know anyone

[00:21:02] : [00:21:04]

who's a great fit for them but then like

[00:21:04] : [00:21:05]

three months later I meet startup

[00:21:05] : [00:21:07]

founder who's like looking for someone

[00:21:07] : [00:21:09]

who you know fit the profile

[00:21:09] : [00:21:11]

um so then you just put people in touch

[00:21:11] : [00:21:12]

via email

[00:21:12] : [00:21:14]

um and that's um that's how it works I

[00:21:14] : [00:21:16]

I've started to there's one company I've

[00:21:16] : [00:21:18]

been recruiting for that won't say much

[00:21:18] : [00:21:19]

about them

[00:21:19] : [00:21:21]

um but they're like really cool I'm like

[00:21:21] : [00:21:23]

a huge fan of them uh you probably

[00:21:23] : [00:21:24]

guessed who they are if you like follow

[00:21:24] : [00:21:25]

me on social media

[00:21:25] : [00:21:28]

um I've started to work with them a

[00:21:28] : [00:21:30]

little bit more closely So like um

[00:21:30] : [00:21:32]

started to like do more Outreach like

[00:21:32] : [00:21:34]

like actually doing outbound and you

[00:21:34] : [00:21:35]

know either reaching out to a couple

[00:21:35] : [00:21:37]

people who I follow on Twitter saying

[00:21:37] : [00:21:38]

like hey I think you'd be a really good

[00:21:38] : [00:21:39]

fit here

[00:21:39] : [00:21:41]

um or even like on LinkedIn

[00:21:41] : [00:21:42]

um

[00:21:42] : [00:21:43]

but that's actually this is the first

[00:21:43] : [00:21:44]

time I've ever done like any kind of

[00:21:44] : [00:21:46]

outbound Outreach

[00:21:46] : [00:21:49]

um so that's that's currently how how my

[00:21:49] : [00:21:50]

thing works

[00:21:50] : [00:21:54]

and you mentioned that this was mainly

[00:21:54] : [00:21:56]

thanks to your Twitter account so I

[00:21:56] : [00:21:59]

think we we should definitely dive into

[00:21:59] : [00:22:01]

the the Twitter which is called AI Pub

[00:22:01] : [00:22:04]

and for the people that are not familiar

[00:22:04] : [00:22:07]

with with it if you're in the field I I

[00:22:07] : [00:22:09]

think we can assume that you are

[00:22:09] : [00:22:11]

familiar with it just because of all the

[00:22:11] : [00:22:13]

reads that you've got in the recent

[00:22:13] : [00:22:17]

uh I don't know six seven month or

[00:22:17] : [00:22:19]

is it that much around there I think it

[00:22:19] : [00:22:23]

started in like September again

[00:22:23] : [00:22:25]

I think that you you've got something

[00:22:25] : [00:22:30]

like 8 million reads or even more now oh

[00:22:30] : [00:22:31]

I I actually haven't checked the numbers

[00:22:31] : [00:22:33]

in a long time I know I have 60 000

[00:22:33] : [00:22:36]

followers uh yeah something like that

[00:22:36] : [00:22:38]

yeah definitely I I would assume the

[00:22:38] : [00:22:39]

fastest growing Twitter account

[00:22:39] : [00:22:42]

especially in in artificial intelligence

[00:22:42] : [00:22:45]

and you already explained a bit what we

[00:22:45] : [00:22:48]

you were doing but could you go a bit

[00:22:48] : [00:22:50]

more in detail of what is your goal with

[00:22:50] : [00:22:52]

the satorical as well as

[00:22:52] : [00:22:54]

who is it for who should follow you and

[00:22:54] : [00:22:58]

and go check it out yeah I think um

[00:22:58] : [00:22:59]

again like all this is kind of an

[00:22:59] : [00:23:00]

experiment like here it's kind of

[00:23:00] : [00:23:02]

there's nothing really set in stone

[00:23:02] : [00:23:04]

um so I'm kind of playing with a lot of

[00:23:04] : [00:23:05]

different things

[00:23:05] : [00:23:07]

um I'll start by saying like who is my

[00:23:07] : [00:23:09]

target audience I think target audience

[00:23:09] : [00:23:12]

is like technical people who are in the

[00:23:12] : [00:23:14]

industry or like very adjacent to the

[00:23:14] : [00:23:15]

industry

[00:23:15] : [00:23:17]

um so like people are professional

[00:23:17] : [00:23:18]

software Engineers people with

[00:23:18] : [00:23:21]

professional machine learning Engineers

[00:23:21] : [00:23:22]

um who either like read machine learning

[00:23:22] : [00:23:23]

papers themselves or like kind of want

[00:23:23] : [00:23:25]

to be kept abreast of it all

[00:23:25] : [00:23:26]

um

[00:23:26] : [00:23:29]

you know uh so that my target audience

[00:23:29] : [00:23:31]

is like generally technical people

[00:23:31] : [00:23:33]

um and then what I'm kind of trying to

[00:23:33] : [00:23:36]

do is cover

[00:23:36] : [00:23:38]

technical AI topics at like a high level

[00:23:38] : [00:23:42]

so like research summaries but like not

[00:23:42] : [00:23:45]

going super deep in the weeds so that

[00:23:45] : [00:23:46]

people can understand a lot of the high

[00:23:46] : [00:23:47]

level Concepts

[00:23:47] : [00:23:49]

um without spending hours and hours

[00:23:49] : [00:23:52]

going into the nitty-gritty uh details

[00:23:52] : [00:23:53]

so there's just a lot of stuff that

[00:23:53] : [00:23:54]

interests me

[00:23:54] : [00:23:56]

um I think I think over the last couple

[00:23:56] : [00:23:59]

months it's mostly been

[00:23:59] : [00:23:59]

um

[00:23:59] : [00:24:04]

just like summaries of research papers

[00:24:04] : [00:24:05]

um for a while I did this thing that I

[00:24:05] : [00:24:07]

feel really bad about I've just been so

[00:24:07] : [00:24:08]

slammed with the recruiting I haven't

[00:24:08] : [00:24:09]

kept up with it but like for a while I

[00:24:09] : [00:24:12]

did I did like a weekly Twitter thread

[00:24:12] : [00:24:14]

that like summarized all all the things

[00:24:14] : [00:24:16]

that happened this week on AI Twitter

[00:24:16] : [00:24:18]

that must be crazy right now if you were

[00:24:18] : [00:24:20]

to do it no yeah I actually again I feel

[00:24:20] : [00:24:21]

terrible about this because I haven't

[00:24:21] : [00:24:22]

done this in a couple weeks because the

[00:24:22] : [00:24:25]

recruiting stuff um but yeah like I

[00:24:25] : [00:24:26]

don't even know how I would do that now

[00:24:26] : [00:24:27]

because it's like the Twitter thread

[00:24:27] : [00:24:29]

like you know you should probably do

[00:24:29] : [00:24:30]

that Twitter thread every two days or

[00:24:30] : [00:24:33]

something like that at this point are

[00:24:33] : [00:24:35]

you familiar with Alpha signal the yeah

[00:24:35] : [00:24:39]

so like he he posted I think it was

[00:24:39] : [00:24:41]

Tuesday like people said that summary of

[00:24:41] : [00:24:44]

what happened Tuesday and it was like

[00:24:44] : [00:24:46]

more more new things that what you were

[00:24:46] : [00:24:49]

sharing in a weekly thread yeah no it's

[00:24:49] : [00:24:52]

it's it is it's it's it's it's kind of I

[00:24:52] : [00:24:54]

I my emotions are rare mix it's like

[00:24:54] : [00:24:56]

very exciting very overwhelming a little

[00:24:56] : [00:24:57]

bit terrifying depending on how you

[00:24:57] : [00:24:58]

think about the future

[00:24:58] : [00:25:00]

um so I think I think all that stuff is

[00:25:00] : [00:25:01]

interesting I think one thing like I

[00:25:01] : [00:25:02]

kind of want to explore in the future

[00:25:02] : [00:25:04]

with the account is not just like like

[00:25:04] : [00:25:07]

research progress but also talking about

[00:25:07] : [00:25:11]

kind of like how AI or is affecting

[00:25:11] : [00:25:13]

Society or is affecting the economy or

[00:25:13] : [00:25:15]

is affecting the startup scene um

[00:25:15] : [00:25:17]

there's there's a podcast that's run by

[00:25:17] : [00:25:19]

a friend that I really really like that

[00:25:19] : [00:25:20]

kind of covers all this stuff super well

[00:25:20] : [00:25:23]

it's called the cognitive Revolution

[00:25:23] : [00:25:25]

um and I think like

[00:25:25] : [00:25:27]

my plate is kind of full right now but I

[00:25:27] : [00:25:28]

think if I get more bandwidth I think I

[00:25:28] : [00:25:30]

want to do more stuff like that because

[00:25:30] : [00:25:31]

it's just a personal interest of mine

[00:25:31] : [00:25:33]

like aside from just keeping up with the

[00:25:33] : [00:25:34]

research papers like I think this stuff

[00:25:34] : [00:25:36]

is just going to be so transformative

[00:25:36] : [00:25:38]

for society and like even just like

[00:25:38] : [00:25:39]

walking around or driving or doing

[00:25:39] : [00:25:41]

errands like I'm just thinking about

[00:25:41] : [00:25:43]

like what's the economy going to be look

[00:25:43] : [00:25:45]

like in three years or like how you know

[00:25:45] : [00:25:46]

how are gonna people going to have to

[00:25:46] : [00:25:48]

adapt you know what what should be

[00:25:48] : [00:25:52]

people be doing to like ride that wave

[00:25:52] : [00:25:53]

um so I think like that's also a

[00:25:53] : [00:25:55]

direction I kind of want to take the

[00:25:55] : [00:25:57]

account of you once I get more bandwidth

[00:25:57] : [00:25:59]

and I think my Powers would be really

[00:25:59] : [00:26:00]

interesting because I think I think

[00:26:00] : [00:26:01]

there's a thing that's on like

[00:26:01] : [00:26:03]

everyone's mind right now anyone who's

[00:26:03] : [00:26:05]

following gpt4 or any of this stuff like

[00:26:05] : [00:26:06]

they're thinking about this kind of

[00:26:06] : [00:26:07]

thing

[00:26:07] : [00:26:10]

yeah exactly and as you just said it but

[00:26:10] : [00:26:12]

I think that would be extremely relevant

[00:26:12] : [00:26:13]

to

[00:26:13] : [00:26:14]

technical people that usually don't

[00:26:14] : [00:26:17]

think about this stuff but they are

[00:26:17] : [00:26:18]

developing those models and working with

[00:26:18] : [00:26:20]

them and Etc but they

[00:26:20] : [00:26:22]

they are

[00:26:22] : [00:26:24]

for sure interested in what will happen

[00:26:24] : [00:26:26]

and how it affects Society but they

[00:26:26] : [00:26:27]

don't have really the time to

[00:26:27] : [00:26:30]

investigate or think about it so it will

[00:26:30] : [00:26:33]

definitely be a great transition for you

[00:26:33] : [00:26:37]

I think well just just because you are

[00:26:37] : [00:26:40]

you already are building the to my mind

[00:26:40] : [00:26:44]

the perfect audience that should be

[00:26:44] : [00:26:47]

um sensibilized with that so that that

[00:26:47] : [00:26:49]

would be definitely great to see I'd

[00:26:49] : [00:26:52]

love to see that yeah if I I'll just say

[00:26:52] : [00:26:54]

one thing I I think they're like all

[00:26:54] : [00:26:56]

these different like media niches even

[00:26:56] : [00:26:57]

like within machine learning like

[00:26:57] : [00:26:59]

there's kind of like machine learning

[00:26:59] : [00:27:01]

for like people who don't have a

[00:27:01] : [00:27:03]

technical background that just want to

[00:27:03] : [00:27:04]

be crap you know kept abreast of the

[00:27:04] : [00:27:07]

news you know I'm trying to do like you

[00:27:07] : [00:27:08]

know a more technical audience but even

[00:27:08] : [00:27:10]

within a more technical audience there's

[00:27:10] : [00:27:12]

so many different like niches like I

[00:27:12] : [00:27:13]

think one thing that would be really

[00:27:13] : [00:27:16]

cool that I haven't seen too much of

[00:27:16] : [00:27:19]

um maybe to some degree is like a media

[00:27:19] : [00:27:22]

Outlet like specifically for AI tools

[00:27:22] : [00:27:25]

just like for developers like like just

[00:27:25] : [00:27:27]

like and you keep it really simple but

[00:27:27] : [00:27:29]

just keeping keeping software developers

[00:27:29] : [00:27:30]

and machine learning Engineers like

[00:27:30] : [00:27:33]

abreast with just like you know the best

[00:27:33] : [00:27:34]

tools that they could be using in their

[00:27:34] : [00:27:36]

in their work like even something like

[00:27:36] : [00:27:37]

that would just be so extraordinarily

[00:27:37] : [00:27:39]

valuable

[00:27:39] : [00:27:40]

um so I'm not doing that but I think it

[00:27:40] : [00:27:43]

would be really cool if someone did yeah

[00:27:43] : [00:27:44]

it's crazy how

[00:27:44] : [00:27:47]

I don't know if it like it wasn't like

[00:27:47] : [00:27:49]

this I feel like in with the internet in

[00:27:49] : [00:27:53]

the beginning but it's crazy now how any

[00:27:53] : [00:27:57]

very Niche topic can blow up and work

[00:27:57] : [00:27:59]

really well sure no I think we're both

[00:27:59] : [00:28:01]

on Twitter like one of my favorite

[00:28:01] : [00:28:02]

accounts is this guy called strip mall

[00:28:02] : [00:28:04]

guy I think he's like kind of famous on

[00:28:04] : [00:28:06]

Twitter and he's just like he's this guy

[00:28:06] : [00:28:09]

and he's he only tweets about like

[00:28:09] : [00:28:11]

buying and flipping strip malls and like

[00:28:11] : [00:28:13]

renovating strip balls and he has like a

[00:28:13] : [00:28:14]

hundred thousand followers or something

[00:28:14] : [00:28:16]

like that

[00:28:16] : [00:28:17]

um so there's there's people like this

[00:28:17] : [00:28:18]

all over

[00:28:18] : [00:28:19]

um

[00:28:19] : [00:28:22]

that's so cool you can like any passion

[00:28:22] : [00:28:24]

that you have or just something that you

[00:28:24] : [00:28:27]

do you can share about it and some other

[00:28:27] : [00:28:31]

person will certainly be interested do

[00:28:31] : [00:28:32]

you uh do you have like a favorite like

[00:28:32] : [00:28:34]

weirder Niche Twitter account I don't

[00:28:34] : [00:28:36]

mean it if you know I don't mean to put

[00:28:36] : [00:28:38]

you on the spot but I'm curious uh you

[00:28:38] : [00:28:40]

you kind of are because I'm I'm a really

[00:28:40] : [00:28:43]

bad twitter user so I I actually wanted

[00:28:43] : [00:28:45]

to like this was part of my questions

[00:28:45] : [00:28:49]

that I actually skipped but I I will ask

[00:28:49] : [00:28:52]

them so since whatever

[00:28:52] : [00:28:53]

um

[00:28:53] : [00:28:55]

for myself personally but also anyone

[00:28:55] : [00:29:00]

else that is that is listening because I

[00:29:00] : [00:29:01]

just give a small background to my

[00:29:01] : [00:29:05]

question it's basically all my friends

[00:29:05] : [00:29:08]

and the people I know like because I I

[00:29:08] : [00:29:10]

assume it may be because we are from a

[00:29:10] : [00:29:12]

small City and here it's French not in

[00:29:12] : [00:29:14]

English or something but nobody is using

[00:29:14] : [00:29:17]

Twitter like back in back in the days it

[00:29:17] : [00:29:19]

was Facebook and now it's messenger

[00:29:19] : [00:29:21]

nobody is using Twitter or even have an

[00:29:21] : [00:29:24]

account how would you convince them that

[00:29:24] : [00:29:26]

Twitter is useful or like why is Twitter

[00:29:26] : [00:29:28]

useful and why should should they be on

[00:29:28] : [00:29:30]

there

[00:29:30] : [00:29:30]

yeah I would say I would say two things

[00:29:30] : [00:29:34]

on a high level and it just depends so

[00:29:34] : [00:29:35]

some people don't have these interests

[00:29:35] : [00:29:38]

but I think the two things that are most

[00:29:38] : [00:29:40]

useful that are kind of separate Are One

[00:29:40] : [00:29:43]

Professional and then uh two are just

[00:29:43] : [00:29:45]

keeping abreast of news

[00:29:45] : [00:29:46]

um

[00:29:46] : [00:29:48]

you know professional like just the

[00:29:48] : [00:29:51]

networking that you can do not like a

[00:29:51] : [00:29:53]

weird sleazy way but just like meeting

[00:29:53] : [00:29:55]

people who are in your Niche

[00:29:55] : [00:29:57]

um is is just wild and it's just so much

[00:29:57] : [00:29:59]

better than LinkedIn like LinkedIn is

[00:29:59] : [00:30:00]

the site where you're supposed to be

[00:30:00] : [00:30:02]

doing this but like I don't know like

[00:30:02] : [00:30:03]

people don't really like hang out and

[00:30:03] : [00:30:05]

have genuine conversations on LinkedIn

[00:30:05] : [00:30:06]

or like people like I don't know it's

[00:30:06] : [00:30:09]

it's not it's much more it's it's much

[00:30:09] : [00:30:12]

more like genuine on Twitter

[00:30:12] : [00:30:14]

um so I think if you're interested in

[00:30:14] : [00:30:17]

doing that like uh you know you should

[00:30:17] : [00:30:19]

try it out uh and they're and they're

[00:30:19] : [00:30:21]

just crazy stories of like people people

[00:30:21] : [00:30:22]

who have made crazy professional

[00:30:22] : [00:30:25]

progress by just going on Twitter

[00:30:25] : [00:30:26]

um you know so I think there's

[00:30:26] : [00:30:28]

professional and then the other is just

[00:30:28] : [00:30:29]

like I think it's the best place to get

[00:30:29] : [00:30:32]

news uh across a variety of niches you

[00:30:32] : [00:30:33]

know

[00:30:33] : [00:30:35]

um you can just find all of these random

[00:30:35] : [00:30:37]

people who are like domain experts again

[00:30:37] : [00:30:39]

it's like you know I follow like Jan Le

[00:30:39] : [00:30:42]

and like Sam Altman but like I also

[00:30:42] : [00:30:44]

follow like strip mall guy or like yeah

[00:30:44] : [00:30:46]

olaji and it's like this really cool

[00:30:46] : [00:30:48]

like it's kind of like you craft your

[00:30:48] : [00:30:50]

own newspaper that you get every morning

[00:30:50] : [00:30:52]

from like all of these different like

[00:30:52] : [00:30:54]

Niche experts on exactly what you want

[00:30:54] : [00:30:56]

to learn about

[00:30:56] : [00:30:58]

um so I and and also just like

[00:30:58] : [00:30:59]

empirically like the the quality of

[00:30:59] : [00:31:01]

information you get

[00:31:01] : [00:31:04]

um is frankly higher and faster than

[00:31:04] : [00:31:07]

what you'd get out of traditional media

[00:31:07] : [00:31:08]

um you know I think with like a lot of

[00:31:08] : [00:31:11]

political issues uh things like kovid

[00:31:11] : [00:31:14]

um you know thing things like inflation

[00:31:14] : [00:31:16]

at least in the US like Twitter at least

[00:31:16] : [00:31:19]

I found was way ahead of the curve

[00:31:19] : [00:31:21]

um on like finding up-to-date info about

[00:31:21] : [00:31:23]

this stuff

[00:31:23] : [00:31:23]

um

[00:31:23] : [00:31:25]

so I think just even keeping abreast of

[00:31:25] : [00:31:27]

what's going on like Twitter's great

[00:31:27] : [00:31:29]

now once you you spend the time to craft

[00:31:29] : [00:31:32]

your your list or like the people that

[00:31:32] : [00:31:34]

you are following this seems like a

[00:31:34] : [00:31:36]

really good tool to to stay in the farm

[00:31:36] : [00:31:39]

in as you said I'll Niche I'll

[00:31:39] : [00:31:41]

most important

[00:31:41] : [00:31:45]

subject for you so definitely and coming

[00:31:45] : [00:31:48]

back to just before we I forget

[00:31:48] : [00:31:52]

coming back to your AI Pub the the

[00:31:52] : [00:31:54]

Twitter account that's this specific

[00:31:54] : [00:31:59]

tutorial how how would you say that

[00:31:59] : [00:32:02]

like higher how are you balancing

[00:32:02] : [00:32:05]

the need to share super Cutting Edge

[00:32:05] : [00:32:08]

research like in your weekly reviews

[00:32:08] : [00:32:10]

well also need to communicate the need

[00:32:10] : [00:32:14]

to communicate them simply and make it

[00:32:14] : [00:32:18]

accessible like how is it for you is it

[00:32:18] : [00:32:21]

difficult to share the most advanced

[00:32:21] : [00:32:24]

algorithm while making it accessible and

[00:32:24] : [00:32:25]

all this and

[00:32:25] : [00:32:27]

Twitter it's like basically really short

[00:32:27] : [00:32:30]

threads so it what are the challenges

[00:32:30] : [00:32:32]

and how are you doing that yeah it's

[00:32:32] : [00:32:33]

definitely hard I mean the biggest

[00:32:33] : [00:32:35]

challenge is just volume

[00:32:35] : [00:32:37]

um and I don't just don't even pretend

[00:32:37] : [00:32:40]

to be a source of like all of The

[00:32:40] : [00:32:42]

Cutting Edge stuff it's more I feel like

[00:32:42] : [00:32:44]

I'll pick out a gem every once in a

[00:32:44] : [00:32:45]

while and maybe it helps people

[00:32:45] : [00:32:47]

understand it but it's like there's just

[00:32:47] : [00:32:50]

even if I was doing my job like 10x

[00:32:50] : [00:32:52]

better like I just being able to cover

[00:32:52] : [00:32:54]

just all of the things that are coming

[00:32:54] : [00:32:56]

out is just like really really hard

[00:32:56] : [00:32:58]

um so for me like the biggest challenge

[00:32:58] : [00:32:59]

is just volume

[00:32:59] : [00:33:01]

um and also Pace like just you know a

[00:33:01] : [00:33:03]

paper will come out you know

[00:33:03] : [00:33:05]

you know Day end like if you can

[00:33:05] : [00:33:07]

figure out what's going on in it and

[00:33:07] : [00:33:09]

summarize it on Day N plus one like

[00:33:09] : [00:33:10]

that's that's great but sometimes it's

[00:33:10] : [00:33:12]

just hard to do

[00:33:12] : [00:33:13]

um

[00:33:13] : [00:33:17]

I actually like in some ways I don't I

[00:33:17] : [00:33:18]

don't actually find there's a ton of

[00:33:18] : [00:33:21]

tension between

[00:33:21] : [00:33:22]

um

[00:33:22] : [00:33:24]

you know there's this question like kind

[00:33:24] : [00:33:26]

of explaining the research versus you

[00:33:26] : [00:33:28]

know crafting a Twitter thread like like

[00:33:28] : [00:33:30]

in some ways like you're kind of

[00:33:30] : [00:33:31]

compressing information and how much

[00:33:31] : [00:33:33]

information is lost when you do that

[00:33:33] : [00:33:36]

yeah um for me I don't I don't find

[00:33:36] : [00:33:37]

there's like a huge trade-off there

[00:33:37] : [00:33:40]

because at least what I'm expressly

[00:33:40] : [00:33:43]

trying to do is is kind of trying to do

[00:33:43] : [00:33:45]

some of that lossy compression uh you

[00:33:45] : [00:33:47]

know so for me like what I'm trying to

[00:33:47] : [00:33:49]

do is trying to kind of distill as much

[00:33:49] : [00:33:50]

as I can

[00:33:50] : [00:33:52]

in the paper into something that could

[00:33:52] : [00:33:53]

be a Twitter thread or could be

[00:33:53] : [00:33:55]

explained yeah you know a page or two of

[00:33:55] : [00:33:58]

writing um I think for other people who

[00:33:58] : [00:34:00]

are going much deeper into the technical

[00:34:00] : [00:34:03]

weeds or like are you know trying to

[00:34:03] : [00:34:04]

explain things for people who have much

[00:34:04] : [00:34:05]

more of a background I think there's

[00:34:05] : [00:34:07]

more of a trade-off but uh for me for me

[00:34:07] : [00:34:09]

not as much and that's kind of why I

[00:34:09] : [00:34:10]

like Twitter like I like that there's

[00:34:10] : [00:34:12]

at least to some degree in the UI like

[00:34:12] : [00:34:14]

the 240 character limit because it

[00:34:14] : [00:34:17]

really actually imposes a constraint of

[00:34:17] : [00:34:20]

being of being succinct and pithy yeah

[00:34:20] : [00:34:21]

and like just really summarizing like

[00:34:21] : [00:34:23]

the key ideas

[00:34:23] : [00:34:23]

um

[00:34:23] : [00:34:25]

so I don't actually find a huge tension

[00:34:25] : [00:34:27]

there personally

[00:34:27] : [00:34:28]

it's really funny actually because I

[00:34:28] : [00:34:31]

feel like oh sorry I I feel like I feel

[00:34:31] : [00:34:32]

like for people who don't use Twitter

[00:34:32] : [00:34:34]

they're historically has been this

[00:34:34] : [00:34:35]

reputation of Twitter as like this kind

[00:34:35] : [00:34:37]

of like

[00:34:37] : [00:34:41]

I feel like low IQ yeah like platform

[00:34:41] : [00:34:43]

it's like oh yeah it's just 140

[00:34:43] : [00:34:44]

characters like what kind of

[00:34:44] : [00:34:46]

intellectual content are you gonna get

[00:34:46] : [00:34:48]

there and it's like surprise actually no

[00:34:48] : [00:34:49]

it's like one of the best places to

[00:34:49] : [00:34:50]

learn

[00:34:50] : [00:34:52]

um so so it seems like there's a trade

[00:34:52] : [00:34:53]

off there but but at least I haven't

[00:34:53] : [00:34:55]

found it personally

[00:34:55] : [00:34:56]

um

[00:34:56] : [00:34:59]

yeah indeed and making stuff simple is

[00:34:59] : [00:35:02]

also a super useful skill for just

[00:35:02] : [00:35:04]

Professionals in general so would you

[00:35:04] : [00:35:06]

have any tips based on how you

[00:35:06] : [00:35:08]

personally do that to for anyone to

[00:35:08] : [00:35:11]

improve on explaining their research or

[00:35:11] : [00:35:14]

their own work more simply

[00:35:14] : [00:35:16]

well that's hard

[00:35:16] : [00:35:18]

um well the first thing that's kind of

[00:35:18] : [00:35:20]

obvious advice is get on Twitter

[00:35:20] : [00:35:21]

um it's really funny it's like a lot of

[00:35:21] : [00:35:24]

the research that ends up getting the

[00:35:24] : [00:35:26]

most attention or like most follow-up

[00:35:26] : [00:35:28]

work is just the research that gets

[00:35:28] : [00:35:29]

shared on Twitter because everyone's

[00:35:29] : [00:35:31]

using Twitter as their source of info so

[00:35:31] : [00:35:32]

I really the first and most obvious

[00:35:32] : [00:35:34]

source of advice is like just go on

[00:35:34] : [00:35:36]

Twitter and like every time you publish

[00:35:36] : [00:35:38]

research like dedicate a good deal of

[00:35:38] : [00:35:39]

time to making like a really good

[00:35:39] : [00:35:42]

Twitter thread with you know photos and

[00:35:42] : [00:35:44]

media and stuff uh that explains what

[00:35:44] : [00:35:46]

you did at a high level

[00:35:46] : [00:35:47]

um

[00:35:47] : [00:35:51]

was in terms of summarizing

[00:35:51] : [00:35:51]

I don't I don't feel like I have like

[00:35:51] : [00:35:54]

super interesting thoughts there it's

[00:35:54] : [00:35:56]

like you can write short sentences kind

[00:35:56] : [00:35:58]

of space it out like you know

[00:35:58] : [00:36:00]

yeah well I think those are already

[00:36:00] : [00:36:03]

great insights just basically I think

[00:36:03] : [00:36:04]

it's it's interesting to force yourself

[00:36:04] : [00:36:07]

to cut some words like make it more

[00:36:07] : [00:36:10]

concise by intention this will just

[00:36:10] : [00:36:13]

allow you to practice your brain to to

[00:36:13] : [00:36:16]

even verbalize things more simply in

[00:36:16] : [00:36:19]

their General Life just because you you

[00:36:19] : [00:36:21]

are doing that you are forcing yourself

[00:36:21] : [00:36:23]

one more thought here which is

[00:36:23] : [00:36:24]

um

[00:36:24] : [00:36:26]

trying to condense something to like

[00:36:26] : [00:36:28]

three major ideas or like trying to

[00:36:28] : [00:36:31]

understand why something is obvious it's

[00:36:31] : [00:36:32]

I when I was an undergrad I was like

[00:36:32] : [00:36:34]

really interested in doing math and I

[00:36:34] : [00:36:36]

was I was mentored by this like really

[00:36:36] : [00:36:38]

awesome mathematician

[00:36:38] : [00:36:40]

um and like one of the pieces of advice

[00:36:40] : [00:36:42]

he had like even for like extremely

[00:36:42] : [00:36:44]

technical work these are like math

[00:36:44] : [00:36:46]

papers which are like way harder to read

[00:36:46] : [00:36:48]

than than machine learning papers this

[00:36:48] : [00:36:50]

is totally esoteric like very awfully

[00:36:50] : [00:36:52]

put together often

[00:36:52] : [00:36:54]

um his advice was often like there are

[00:36:54] : [00:36:55]

three major ideas and like your goal

[00:36:55] : [00:36:57]

reading this like really hard esoteric

[00:36:57] : [00:37:00]

thing is to just figure out what those

[00:37:00] : [00:37:03]

like three major ideas are so like kind

[00:37:03] : [00:37:04]

of deliberately approaching your

[00:37:04] : [00:37:07]

explanations or your

[00:37:07] : [00:37:10]

uh reading from the perspective of like

[00:37:10] : [00:37:11]

what are what are the three bullet

[00:37:11] : [00:37:13]

points that explain this whole thing or

[00:37:13] : [00:37:15]

the other the other kind of like useful

[00:37:15] : [00:37:17]

intuition is like why is this obvious

[00:37:17] : [00:37:20]

like because the funny thing is like all

[00:37:20] : [00:37:21]

this intellectual work like some really

[00:37:21] : [00:37:23]

interesting paper comes out sometimes it

[00:37:23] : [00:37:25]

seems like this incredible technical

[00:37:25] : [00:37:27]

feat or like this leap of Genius of the

[00:37:27] : [00:37:28]

mind

[00:37:28] : [00:37:31]

um but but the person who discovered it

[00:37:31] : [00:37:33]

it was obvious right like to the person

[00:37:33] : [00:37:35]

who discovered it was actually very

[00:37:35] : [00:37:37]

intuitive so then sometimes I feel like

[00:37:37] : [00:37:39]

the goal of either coming to understand

[00:37:39] : [00:37:42]

the stuff well or explain it well is

[00:37:42] : [00:37:43]

trying to like figure out like trying to

[00:37:43] : [00:37:46]

reverse engineer the headspace to the

[00:37:46] : [00:37:47]

point where the work becomes obvious so

[00:37:47] : [00:37:49]

you're kind of asking yourself the

[00:37:49] : [00:37:50]

question like why is this really

[00:37:50] : [00:37:52]

intuitive even if it seems hard like

[00:37:52] : [00:37:55]

what's the intuition behind it um

[00:37:55] : [00:37:57]

these are two ideas I learned from math

[00:37:57] : [00:37:59]

that yeah they are good tips and also

[00:37:59] : [00:38:01]

what you mentioned with using external

[00:38:01] : [00:38:05]

media like images or videos or even just

[00:38:05] : [00:38:07]

real world example just like something

[00:38:07] : [00:38:08]

that anyone can understand but you can

[00:38:08] : [00:38:11]

compare that's extremely good to just

[00:38:11] : [00:38:14]

make things sound simpler and more

[00:38:14] : [00:38:16]

understandable I I think well most

[00:38:16] : [00:38:18]

professors do that they start to most

[00:38:18] : [00:38:20]

good professors do that they start with

[00:38:20] : [00:38:22]

a real world example or real life like

[00:38:22] : [00:38:24]

they talk about tennis or whatever and

[00:38:24] : [00:38:26]

then they talk about math

[00:38:26] : [00:38:28]

whether you like it or not a lot of this

[00:38:28] : [00:38:31]

stuff is marketing you know like you

[00:38:31] : [00:38:33]

know you you do the research that's

[00:38:33] : [00:38:34]

making the product but then in order to

[00:38:34] : [00:38:36]

get people to pay attention to it it's

[00:38:36] : [00:38:38]

marketing so you know learning the

[00:38:38] : [00:38:40]

basics of copywriting and all these

[00:38:40] : [00:38:41]

network marketing things that a lot of

[00:38:41] : [00:38:42]

people are reverse to it's useful

[00:38:42] : [00:38:45]

exactly and just just like selling

[00:38:45] : [00:38:47]

yourself in an interview or anything

[00:38:47] : [00:38:50]

it's it's all about marketing yep so you

[00:38:50] : [00:38:52]

you mentioned that you were

[00:38:52] : [00:38:54]

mainly staying up to date with Twitter

[00:38:54] : [00:38:56]

and

[00:38:56] : [00:38:57]

yeah Twitter

[00:38:57] : [00:39:00]

and so is there since you are yourself a

[00:39:00] : [00:39:02]

Twitter account trying to share news in

[00:39:02] : [00:39:05]

Ai and what my knee and AI

[00:39:05] : [00:39:08]

are there any other tools that you are

[00:39:08] : [00:39:10]

using to help stay up to date or

[00:39:10] : [00:39:12]

newsletters or I don't know YouTube

[00:39:12] : [00:39:15]

accounts or anything else than Twitter

[00:39:15] : [00:39:18]

accounts or if if there are other useful

[00:39:18] : [00:39:21]

Twitter accounts that you can share yes

[00:39:21] : [00:39:22]

um okay so you don't want to like play

[00:39:22] : [00:39:25]

favorites here but I'll and I I'm I'm a

[00:39:25] : [00:39:27]

little bit sleep deprived today so I'm

[00:39:27] : [00:39:28]

sure like some great accounts are gonna

[00:39:28] : [00:39:32]

skip my memory um nonetheless uh a

[00:39:32] : [00:39:34]

couple that come to mind

[00:39:34] : [00:39:36]

um Dar AI or I don't know how you

[00:39:36] : [00:39:37]

pronounce it but their AI is like

[00:39:37] : [00:39:40]

they're really really good

[00:39:40] : [00:39:42]

um they I feel like there are just

[00:39:42] : [00:39:43]

actually a lot of Twitter accounts that

[00:39:43] : [00:39:45]

are now just doing a better job at what

[00:39:45] : [00:39:46]

I'm trying to do than I am

[00:39:46] : [00:39:48]

um so I think I think like Dairy eye is

[00:39:48] : [00:39:50]

really really good uh they're doing

[00:39:50] : [00:39:52]

weekly paper summaries which is like

[00:39:52] : [00:39:53]

super cool so if you're trying to keep

[00:39:53] : [00:39:55]

up to date with research like I think

[00:39:55] : [00:39:57]

it's every Sunday they publish a thread

[00:39:57] : [00:39:59]

where it's like hey here like the top 10

[00:39:59] : [00:40:01]

like papers that came out this week and

[00:40:01] : [00:40:03]

here's like one or two sentences about

[00:40:03] : [00:40:05]

like what they contain

[00:40:05] : [00:40:07]

um so dare AI is really good

[00:40:07] : [00:40:09]

um Alpha signal you know uh Weekly

[00:40:09] : [00:40:11]

Newsletter with kind of like some of the

[00:40:11] : [00:40:14]

top AI papers that's really cool

[00:40:14] : [00:40:15]

um

[00:40:15] : [00:40:18]

uh I really like uh Dr Jim fan

[00:40:18] : [00:40:20]

um he's he's on Twitter he like he makes

[00:40:20] : [00:40:22]

like really cool explainer threads of

[00:40:22] : [00:40:24]

various AI research stuff

[00:40:24] : [00:40:27]

um and then and then another uh kind of

[00:40:27] : [00:40:30]

Follow that that I really like that's a

[00:40:30] : [00:40:31]

little bit different from the research

[00:40:31] : [00:40:32]

stuff

[00:40:32] : [00:40:34]

um is is this podcast called the

[00:40:34] : [00:40:36]

cognitive Revolution that I mentioned

[00:40:36] : [00:40:39]

um is run by uh this guy Nathan lebens

[00:40:39] : [00:40:42]

uh who's just a great guy and uh it it

[00:40:42] : [00:40:44]

covers some AI research stuff but also

[00:40:44] : [00:40:46]

just like how AI is affecting society

[00:40:46] : [00:40:49]

and they like interview some of the top

[00:40:49] : [00:40:51]

like startup Founders or like VCS in the

[00:40:51] : [00:40:53]

space and also researchers so it's like

[00:40:53] : [00:40:56]

more of a kind of like broad uh like

[00:40:56] : [00:40:58]

podcast uh that that I think is good and

[00:40:58] : [00:41:00]

they also they also have like a sub

[00:41:00] : [00:41:02]

stack or a newsletter not exactly sure

[00:41:02] : [00:41:04]

on that or they they do have one I I

[00:41:04] : [00:41:06]

mostly just listen to the podcast

[00:41:06] : [00:41:09]

I I will also add just it's basically

[00:41:09] : [00:41:12]

this the only with yours and I follow

[00:41:12] : [00:41:15]

Alpha signal AI Pub and one other

[00:41:15] : [00:41:19]

Twitter page account

[00:41:19] : [00:41:22]

um for this is mainly interesting for

[00:41:22] : [00:41:24]

people enter in into research AI

[00:41:24] : [00:41:26]

research and

[00:41:26] : [00:41:29]

even more towards

[00:41:29] : [00:41:31]

like researchers applying those research

[00:41:31] : [00:41:35]

but it's obviously AK oh yeah yes you

[00:41:35] : [00:41:37]

already know but this this was a game

[00:41:37] : [00:41:39]

changer to me like a few I think years

[00:41:39] : [00:41:43]

ago now well and it's definitely cool he

[00:41:43] : [00:41:46]

posts he basically shares

[00:41:46] : [00:41:49]

every new relevant papers and even

[00:41:49] : [00:41:51]

non-relevant but like any new papers

[00:41:51] : [00:41:53]

with a short one-liner that is usually

[00:41:53] : [00:41:56]

taken from the abstract but still a

[00:41:56] : [00:41:58]

short useful one-liner to know what's

[00:41:58] : [00:42:00]

the paper about with a picture and it's

[00:42:00] : [00:42:04]

it's just really cool to to be easily to

[00:42:04] : [00:42:06]

have easy updates on on the new papers

[00:42:06] : [00:42:08]

the new like thousands of daily papers

[00:42:08] : [00:42:10]

so that's that's a really good one I

[00:42:10] : [00:42:12]

feel like yep I totally agree totally

[00:42:12] : [00:42:15]

forgot uh another person who's in a

[00:42:15] : [00:42:18]

similar vein is Iran Iran kamat suzaki

[00:42:18] : [00:42:20]

um he's he's at he's doing a machine

[00:42:20] : [00:42:21]

learning PhD at Georgia Tech I think

[00:42:21] : [00:42:24]

he's finishing up uh and he he does

[00:42:24] : [00:42:25]

similar stuff like he posts a lot of

[00:42:25] : [00:42:27]

papers and kind of summaries and papers

[00:42:27] : [00:42:29]

uh awesome

[00:42:29] : [00:42:32]

oh I'm I'm sure you do uh he's he's very

[00:42:32] : [00:42:35]

active on on AI yeah I must have seen it

[00:42:35] : [00:42:37]

but I my mirror is extremely bad anyway

[00:42:37] : [00:42:40]

so I probably know know him but yeah to

[00:42:40] : [00:42:43]

continue on on Twitter since uh this is

[00:42:43] : [00:42:45]

the pretty much the main topic of the

[00:42:45] : [00:42:48]

past 30 minutes or so I I have still

[00:42:48] : [00:42:50]

follow-up questions on this and it's

[00:42:50] : [00:42:54]

also for people like me that maybe do

[00:42:54] : [00:42:56]

not understand Twitter enough

[00:42:56] : [00:43:00]

and it's more about growth so

[00:43:00] : [00:43:03]

a very short question that may be

[00:43:03] : [00:43:06]

extremely hard to answer is how to do a

[00:43:06] : [00:43:07]

viral Twitter post

[00:43:07] : [00:43:10]

oh that's um that's what I'll put I I

[00:43:10] : [00:43:12]

don't mean to toot my own horn too much

[00:43:12] : [00:43:14]

uh but I do actually think I've Gotten

[00:43:14] : [00:43:15]

Good at this

[00:43:15] : [00:43:18]

um especially yeah yeah now I've been

[00:43:18] : [00:43:20]

actually kind of like slammed with some

[00:43:20] : [00:43:21]

of this recruiting stuff and other

[00:43:21] : [00:43:22]

things I'm doing so I actually haven't

[00:43:22] : [00:43:23]

been focused on growing my Twitter

[00:43:23] : [00:43:25]

account but yeah like first first couple

[00:43:25] : [00:43:27]

months it was like very deliberate very

[00:43:27] : [00:43:28]

focused on it and I think it did a very

[00:43:28] : [00:43:30]

good job

[00:43:30] : [00:43:30]

um

[00:43:30] : [00:43:33]

yeah I think so there are a lot of

[00:43:33] : [00:43:34]

thoughts

[00:43:34] : [00:43:36]

um how do you write yeah so there I mean

[00:43:36] : [00:43:39]

and maybe this is too vague or too broad

[00:43:39] : [00:43:40]

but I think you gotta get like three

[00:43:40] : [00:43:42]

things right I think one getting the

[00:43:42] : [00:43:44]

audience right like kind of it's kind of

[00:43:44] : [00:43:46]

like finding product Market fit it's

[00:43:46] : [00:43:49]

like finding like info audience fit like

[00:43:49] : [00:43:51]

like like being really deliberate like

[00:43:51] : [00:43:53]

think like a think like a business like

[00:43:53] : [00:43:56]

ask who are your customers and ask like

[00:43:56] : [00:43:58]

what do they actually want like what

[00:43:58] : [00:43:59]

would help them like what do they

[00:43:59] : [00:44:01]

actually care about rather than like

[00:44:01] : [00:44:04]

what's exactly interesting to you is

[00:44:04] : [00:44:05]

because often like those very subtle

[00:44:05] : [00:44:07]

differences in that

[00:44:07] : [00:44:11]

um so like kind of like finding like a

[00:44:11] : [00:44:13]

niche audience or once you have an

[00:44:13] : [00:44:14]

audience like figuring out exactly who

[00:44:14] : [00:44:16]

that is and like really being deliberate

[00:44:16] : [00:44:18]

like what did what would actually be

[00:44:18] : [00:44:19]

very useful to them

[00:44:19] : [00:44:23]

um you know and um you know thinking

[00:44:23] : [00:44:24]

about that

[00:44:24] : [00:44:26]

um you know

[00:44:26] : [00:44:28]

um I think so so that the second is is

[00:44:28] : [00:44:30]

content like finding content that's

[00:44:30] : [00:44:32]

interesting and relevant

[00:44:32] : [00:44:34]

um I think one thing that I definitely

[00:44:34] : [00:44:38]

found is posting posting like long form

[00:44:38] : [00:44:40]

content

[00:44:40] : [00:44:42]

um that's like really thought out and

[00:44:42] : [00:44:45]

valuable like occasionally does much

[00:44:45] : [00:44:47]

better than

[00:44:47] : [00:44:50]

um really frequently posting content

[00:44:50] : [00:44:52]

that's not necessarily like stuff that

[00:44:52] : [00:44:55]

you have added or like reposting stuff

[00:44:55] : [00:44:56]

for just like sharing a random things I

[00:44:56] : [00:44:58]

think like there's a huge return like

[00:44:58] : [00:45:00]

one of the reasons I actually don't post

[00:45:00] : [00:45:01]

too much on Twitter is because like

[00:45:01] : [00:45:03]

often it takes me like several hours to

[00:45:03] : [00:45:05]

write a single thread because it's like

[00:45:05] : [00:45:06]

I'll read the paper like come up with

[00:45:06] : [00:45:08]

some of the major ideas and all that

[00:45:08] : [00:45:10]

like it takes a long time

[00:45:10] : [00:45:12]

um so I think like you know figuring out

[00:45:12] : [00:45:14]

what your audience wants then like doing

[00:45:14] : [00:45:16]

really like long-form valuable stuff in

[00:45:16] : [00:45:18]

that domain uh that's like detailed and

[00:45:18] : [00:45:19]

helpful

[00:45:19] : [00:45:21]

um and then the third thing you kind of

[00:45:21] : [00:45:23]

have to get all these right too uh

[00:45:23] : [00:45:24]

because I feel like if you get one of

[00:45:24] : [00:45:26]

them wrong like you you it really you

[00:45:26] : [00:45:28]

don't grow as much or you don't get get

[00:45:28] : [00:45:30]

as much um

[00:45:30] : [00:45:31]

views

[00:45:31] : [00:45:33]

um the last is just like kind of

[00:45:33] : [00:45:35]

straightforward like marketing

[00:45:35] : [00:45:37]

copywriting stuff

[00:45:37] : [00:45:39]

um so it's like you know just the way

[00:45:39] : [00:45:41]

you structure your sentences or like the

[00:45:41] : [00:45:44]

way you structure your tweets like each

[00:45:44] : [00:45:46]

you know have each tweet like not have a

[00:45:46] : [00:45:48]

ton of text in it like have each tweet

[00:45:48] : [00:45:51]

ideally just contain one idea

[00:45:51] : [00:45:52]

um you know structure your sentences so

[00:45:52] : [00:45:54]

that they're like short easy to follow

[00:45:54] : [00:45:56]

structure your sentences so that they're

[00:45:56] : [00:45:57]

kind of interesting so it's like you

[00:45:57] : [00:46:00]

have one idea and then the next idea is

[00:46:00] : [00:46:04]

like a plot twist or like you know uh

[00:46:04] : [00:46:05]

you just you don't want it to be too

[00:46:05] : [00:46:08]

monotonic right

[00:46:08] : [00:46:10]

um you know and then also including

[00:46:10] : [00:46:12]

media like including like images for

[00:46:12] : [00:46:15]

each each uh each tweet like these these

[00:46:15] : [00:46:16]

kind of subtle things actually end up

[00:46:16] : [00:46:18]

mattering a lot so I think

[00:46:18] : [00:46:20]

maybe that's not super explanatory but

[00:46:20] : [00:46:22]

that's kind of my like high level view

[00:46:22] : [00:46:24]

it's like you get infinite you know

[00:46:24] : [00:46:26]

audience info fit think about what your

[00:46:26] : [00:46:28]

audience actually wants like think about

[00:46:28] : [00:46:30]

them like a customer how can you serve

[00:46:30] : [00:46:32]

them uh you know long form detailed

[00:46:32] : [00:46:34]

stuff and then like doing the

[00:46:34] : [00:46:36]

copywriting and the and the images and

[00:46:36] : [00:46:37]

stuff

[00:46:37] : [00:46:39]

and how much time did it take when you

[00:46:39] : [00:46:41]

were trying to really grow on Twitter

[00:46:41] : [00:46:44]

like per day or per week

[00:46:44] : [00:46:46]

approximately it really depends like I

[00:46:46] : [00:46:47]

wouldn't say actually a ton of time

[00:46:47] : [00:46:49]

honestly

[00:46:49] : [00:46:51]

um I think like I think when I was

[00:46:51] : [00:46:53]

starting maybe like 10 hours a week or

[00:46:53] : [00:46:55]

something like that

[00:46:55] : [00:46:57]

um because I was doing well actually

[00:46:57] : [00:46:58]

maybe more than that because I think

[00:46:58] : [00:47:00]

like when I was starting out even when I

[00:47:00] : [00:47:01]

was starting out and I was growing

[00:47:01] : [00:47:03]

really fast like I wasn't actually

[00:47:03] : [00:47:04]

posting that much stuff I was posting

[00:47:04] : [00:47:06]

like twice a week basically like I would

[00:47:06] : [00:47:07]

post

[00:47:07] : [00:47:10]

roughly every week like one long form

[00:47:10] : [00:47:11]

explainer thread on like some cool paper

[00:47:11] : [00:47:14]

or some cool machine learning topic

[00:47:14] : [00:47:16]

um and then I would do like a weekly

[00:47:16] : [00:47:20]

like uh summary of the stuff that

[00:47:20] : [00:47:22]

happened in AI or AI research uh and

[00:47:22] : [00:47:24]

each of those okay maybe it's actually a

[00:47:24] : [00:47:27]

little bit over 10 like the the summary

[00:47:27] : [00:47:30]

thing I think even to put the thread

[00:47:30] : [00:47:32]

together would take like a couple hours

[00:47:32] : [00:47:34]

uh but then throughout the course of the

[00:47:34] : [00:47:36]

week I would just save all of the links

[00:47:36] : [00:47:38]

that I found on Twitter to like filter

[00:47:38] : [00:47:39]

through that so maybe it's like 15 hours

[00:47:39] : [00:47:41]

something like that and then and then

[00:47:41] : [00:47:43]

you know uh writing a thread on a paper

[00:47:43] : [00:47:44]

like

[00:47:44] : [00:47:46]

maybe it's like seven hours or something

[00:47:46] : [00:47:49]

like that like you find the paper you

[00:47:49] : [00:47:51]

print it yeah I'm boring I you actually

[00:47:51] : [00:47:53]

read it on paper and Mark it up

[00:47:53] : [00:47:55]

um you know that can take a couple hours

[00:47:55] : [00:47:58]

to really understand and then um you

[00:47:58] : [00:47:59]

know writing the thread and doing all

[00:47:59] : [00:48:01]

the like the copywriting like I think a

[00:48:01] : [00:48:02]

lot of people just don't like doing this

[00:48:02] : [00:48:03]

because it's actually kind of boring

[00:48:03] : [00:48:05]

like just getting getting the images

[00:48:05] : [00:48:06]

right and getting the sentences right

[00:48:06] : [00:48:08]

like that actually takes a good deal of

[00:48:08] : [00:48:10]

time yeah I think for me it was like 10

[00:48:10] : [00:48:12]

to 15 hours a week something like that

[00:48:12] : [00:48:16]

and right now how long does it take you

[00:48:16] : [00:48:16]

I think I'm actually spending maybe a

[00:48:16] : [00:48:19]

similar amount of time because I'm also

[00:48:19] : [00:48:21]

doing a podcast now

[00:48:21] : [00:48:22]

um

[00:48:22] : [00:48:24]

maybe actually maybe a little bit less

[00:48:24] : [00:48:25]

than ten something like 10 I would say

[00:48:25] : [00:48:27]

something like that

[00:48:27] : [00:48:30]

yeah that is uh it's not really that

[00:48:30] : [00:48:33]

many hours to get that much views and

[00:48:33] : [00:48:36]

and just so many opportunities oh yeah

[00:48:36] : [00:48:38]

yeah less than we think total life hack

[00:48:38] : [00:48:40]

like like even you know someone who's

[00:48:40] : [00:48:42]

just you have a full-time job you know

[00:48:42] : [00:48:45]

you do a couple hours a week like and it

[00:48:45] : [00:48:47]

just it totally changed your life

[00:48:47] : [00:48:50]

indeed and you mentioned cool research

[00:48:50] : [00:48:53]

papers but how do you decide if they are

[00:48:53] : [00:48:55]

cool or worth sharing you you said that

[00:48:55] : [00:48:58]

you have to see as if you you were

[00:48:58] : [00:49:01]

basically sharing with customers but

[00:49:01] : [00:49:03]

it's still

[00:49:03] : [00:49:05]

you in the end you still have to decide

[00:49:05] : [00:49:08]

if you are explaining a paper or not so

[00:49:08] : [00:49:11]

how how do you take these decisions yeah

[00:49:11] : [00:49:13]

I think in terms of finding the papers

[00:49:13] : [00:49:14]

and maybe you're just a little bit more

[00:49:14] : [00:49:16]

selfish I think I'm actually thinking

[00:49:16] : [00:49:18]

less from the audience perspective and I

[00:49:18] : [00:49:19]

but it's it's more like I'm just using

[00:49:19] : [00:49:21]

my own taste or my own judgment it's

[00:49:21] : [00:49:23]

kind of like with even with math like

[00:49:23] : [00:49:24]

people just have different tastes you

[00:49:24] : [00:49:26]

know some people like analysis other

[00:49:26] : [00:49:27]

people like number Theory or something

[00:49:27] : [00:49:28]

you know

[00:49:28] : [00:49:30]

um and with machine learning papers like

[00:49:30] : [00:49:32]

it's just kind of like what I what I

[00:49:32] : [00:49:33]

find most interesting and typically what

[00:49:33] : [00:49:36]

I find most interesting is like

[00:49:36] : [00:49:38]

I would say less like really cool demos

[00:49:38] : [00:49:41]

or something or like a really impressive

[00:49:41] : [00:49:43]

state of the art result or something but

[00:49:43] : [00:49:46]

I think what I find very interesting is

[00:49:46] : [00:49:49]

like papers that open up new paradigms

[00:49:49] : [00:49:52]

or papers that

[00:49:52] : [00:49:53]

um kind of have something to say about

[00:49:53] : [00:49:56]

the long-term direction of uh AI

[00:49:56] : [00:49:59]

research I also like to some degree I'm

[00:49:59] : [00:50:02]

in kind of an AGI believer like maybe

[00:50:02] : [00:50:04]

even short timelines type person so like

[00:50:04] : [00:50:06]

I also find papers that kind of like

[00:50:06] : [00:50:09]

play into that thing interesting you

[00:50:09] : [00:50:11]

know so it's like

[00:50:11] : [00:50:13]

uh some of the papers that I've covered

[00:50:13] : [00:50:15]

recently on the podcast that that I

[00:50:15] : [00:50:17]

found really interesting or like tool

[00:50:17] : [00:50:19]

former so it's like okay that's like

[00:50:19] : [00:50:21]

just directly Society societally

[00:50:21] : [00:50:24]

relevant um or like you know recently

[00:50:24] : [00:50:27]

did one on um Hungry Hungry Hippos which

[00:50:27] : [00:50:28]

is like a new language modeling

[00:50:28] : [00:50:30]

architecture that like has much better

[00:50:30] : [00:50:33]

scaling with um context uh the context

[00:50:33] : [00:50:35]

length so like stuff like that like

[00:50:35] : [00:50:36]

where it's like oh if you get that right

[00:50:36] : [00:50:37]

then the language models are gonna get

[00:50:37] : [00:50:38]

way better

[00:50:38] : [00:50:42]

um so it's mostly it's mostly my taste I

[00:50:42] : [00:50:43]

think I'm mostly interested in language

[00:50:43] : [00:50:44]

models

[00:50:44] : [00:50:47]

um or like or or like Foundation models

[00:50:47] : [00:50:49]

that incorporate language so I think

[00:50:49] : [00:50:50]

there's also a lot more bias towards

[00:50:50] : [00:50:52]

that than like vision

[00:50:52] : [00:50:55]

yeah you you definitely have a bias Just

[00:50:55] : [00:50:57]

Like Me on YouTube I I share what I find

[00:50:57] : [00:50:59]

interesting and what I like it's and but

[00:50:59] : [00:51:02]

I think that's a good thing it's also it

[00:51:02] : [00:51:04]

also gives personality to your Twitter

[00:51:04] : [00:51:06]

account and if people enjoy what you

[00:51:06] : [00:51:09]

seem to be sharing well they won't be

[00:51:09] : [00:51:11]

disappointed since you will keep sharing

[00:51:11] : [00:51:13]

similar stuff

[00:51:13] : [00:51:15]

yeah yeah there's always I think this is

[00:51:15] : [00:51:16]

kind of kind of interesting Trend in

[00:51:16] : [00:51:18]

like media there's always this like

[00:51:18] : [00:51:20]

interesting tension that I've seen

[00:51:20] : [00:51:21]

between

[00:51:21] : [00:51:22]

um

[00:51:22] : [00:51:25]

kind of doing like writing for yourself

[00:51:25] : [00:51:27]

or producing media for yourself that's

[00:51:27] : [00:51:30]

like unique to your character and

[00:51:30] : [00:51:33]

like sort of serving your audience as if

[00:51:33] : [00:51:36]

you were a business or a servant or you

[00:51:36] : [00:51:38]

know kind of like a more of a giving

[00:51:38] : [00:51:40]

mentality

[00:51:40] : [00:51:41]

um because you kind of have to do both

[00:51:41] : [00:51:43]

like if you're just so obsessed with

[00:51:43] : [00:51:44]

like what you think is cool like

[00:51:44] : [00:51:46]

actually often people don't care like

[00:51:46] : [00:51:47]

it's just not interesting to other

[00:51:47] : [00:51:49]

people but then on the other side like

[00:51:49] : [00:51:51]

there's definitely a phenomenon among

[00:51:51] : [00:51:53]

like influencers of like audience

[00:51:53] : [00:51:56]

capture where it's like they start doing

[00:51:56] : [00:51:57]

something that's really interesting and

[00:51:57] : [00:51:59]

unique but then they find that Their

[00:51:59] : [00:52:01]

audience actually wants more clickbait

[00:52:01] : [00:52:03]

and then like over time they end up just

[00:52:03] : [00:52:06]

becoming like a clickbait Channel or

[00:52:06] : [00:52:08]

they end up just being completely

[00:52:08] : [00:52:09]

transformed and captured by Their

[00:52:09] : [00:52:11]

audience so it's

[00:52:11] : [00:52:13]

it's definitely something that you have

[00:52:13] : [00:52:15]

to like think about yeah and speaking of

[00:52:15] : [00:52:18]

audience how how are you engaging with

[00:52:18] : [00:52:20]

your Twitter audience or just building

[00:52:20] : [00:52:23]

some kind of real engaged Community

[00:52:23] : [00:52:27]

rather than just being a anonymized

[00:52:27] : [00:52:29]

Twitter account that doesn't talk to

[00:52:29] : [00:52:32]

them do you do anything more specific

[00:52:32] : [00:52:34]

than sharing your your weekly posts or

[00:52:34] : [00:52:37]

all your posts well I'm doing much a

[00:52:37] : [00:52:39]

much worse job than you are because you

[00:52:39] : [00:52:42]

actually run a community I I'm I I feel

[00:52:42] : [00:52:43]

like I could actually improve a lot in

[00:52:43] : [00:52:45]

this respect I think insofar as I'm

[00:52:45] : [00:52:47]

doing anything community

[00:52:47] : [00:52:49]

oriented I think it's actually much less

[00:52:49] : [00:52:50]

like scalable

[00:52:50] : [00:52:52]

um or it's very like I think I'm

[00:52:52] : [00:52:53]

actually doing more like in-person stuff

[00:52:53] : [00:52:56]

so like yeah occasionally all host

[00:52:56] : [00:52:58]

events you know I feel bad because it's

[00:52:58] : [00:52:59]

like they're NSF like if you're not NSF

[00:52:59] : [00:53:01]

but like you know a whole host events

[00:53:01] : [00:53:03]

are all sponsor events that are like

[00:53:03] : [00:53:05]

open to anyone who wants to show up

[00:53:05] : [00:53:07]

um and then some of the some of the

[00:53:07] : [00:53:10]

recruiting stuff uh is also like useful

[00:53:10] : [00:53:11]

so like a lot of the stuff that I've

[00:53:11] : [00:53:12]

been doing community wise is like in

[00:53:12] : [00:53:14]

person where it's like there are just

[00:53:14] : [00:53:16]

people who I've been chatting with in

[00:53:16] : [00:53:17]

the DMS for like a couple months and

[00:53:17] : [00:53:19]

they're like looking for a job in this

[00:53:19] : [00:53:21]

area and like I'll help them out

[00:53:21] : [00:53:22]

um or just like connecting people but

[00:53:22] : [00:53:24]

it's very it's very like non-scalable

[00:53:24] : [00:53:26]

like relative to I don't know

[00:53:26] : [00:53:28]

do you have any plan in in creating

[00:53:28] : [00:53:30]

something that is more oriented towards

[00:53:30] : [00:53:33]

some kind of community and where people

[00:53:33] : [00:53:34]

can

[00:53:34] : [00:53:36]

help each other or you can help them

[00:53:36] : [00:53:38]

more easily than in person because of

[00:53:38] : [00:53:40]

course in person is like the best thing

[00:53:40] : [00:53:42]

but it as you said it's definitely not

[00:53:42] : [00:53:44]

scalable

[00:53:44] : [00:53:46]

to be honest not really like right right

[00:53:46] : [00:53:49]

now uh I feel like I'm still like

[00:53:49] : [00:53:50]

figuring out all this recruiting stuff

[00:53:50] : [00:53:52]

like a little bit slammed uh just with

[00:53:52] : [00:53:54]

all the things that are going on so I

[00:53:54] : [00:53:55]

don't

[00:53:55] : [00:53:57]

I think one thing that I've tried to do

[00:53:57] : [00:53:59]

with the account is like if I'm doing

[00:53:59] : [00:54:00]

anything I want to do like really good

[00:54:00] : [00:54:02]

jobs so I I don't feel like I do a very

[00:54:02] : [00:54:03]

good job with the community thing right

[00:54:03] : [00:54:04]

now

[00:54:04] : [00:54:07]

um so there's nothing on my docket

[00:54:07] : [00:54:10]

yeah that makes sense you you are just

[00:54:10] : [00:54:12]

prioritizing and focusing on other stuff

[00:54:12] : [00:54:14]

that is more important to you and that's

[00:54:14] : [00:54:17]

that's just perfectly fine and well

[00:54:17] : [00:54:18]

speaking of managing the the Twitter

[00:54:18] : [00:54:21]

account is there any challenges or

[00:54:21] : [00:54:25]

difficult things that we people that are

[00:54:25] : [00:54:26]

not really good with Twitter or are not

[00:54:26] : [00:54:28]

managing such a popular Twitter account

[00:54:28] : [00:54:32]

uh yeah sorry I will just before it are

[00:54:32] : [00:54:33]

there any

[00:54:33] : [00:54:35]

challenges that we do not expect to see

[00:54:35] : [00:54:40]

from managing oh

[00:54:40] : [00:54:40]

something that maybe you didn't expect

[00:54:40] : [00:54:45]

at the beginning yourself

[00:54:45] : [00:54:47]

that is that takes you more time than

[00:54:47] : [00:54:49]

expected or yeah I mean it I think well

[00:54:49] : [00:54:51]

I would say two things

[00:54:51] : [00:54:53]

um one is it depends on what kind of

[00:54:53] : [00:54:54]

account you're running like if you're

[00:54:54] : [00:54:56]

running a newsletter for example I know

[00:54:56] : [00:54:57]

people are running like daily AI

[00:54:57] : [00:55:00]

newsletters you just I used to you know

[00:55:00] : [00:55:02]

yeah you have to keep up you have to

[00:55:02] : [00:55:04]

every day put that out like there are

[00:55:04] : [00:55:06]

people who are depending on you yeah I

[00:55:06] : [00:55:07]

feel fortunate enough that like kind of

[00:55:07] : [00:55:09]

what I'm doing can be at my own pace so

[00:55:09] : [00:55:11]

it's like if I'm really slammed like

[00:55:11] : [00:55:14]

you know I'll just you know I will post

[00:55:14] : [00:55:15]

something about a research paper that

[00:55:15] : [00:55:18]

week yeah um so I I don't have as much

[00:55:18] : [00:55:21]

of that pressure but I think

[00:55:21] : [00:55:22]

um you know other people would

[00:55:22] : [00:55:24]

experience that

[00:55:24] : [00:55:25]

um

[00:55:25] : [00:55:28]

I I think the the main thing for me

[00:55:28] : [00:55:30]

that's also true of like some of the

[00:55:30] : [00:55:32]

recruiting stuff is just like inbound

[00:55:32] : [00:55:34]

and just like DMs I feel so bad I'm like

[00:55:34] : [00:55:37]

I and it's it's also a skill so it's

[00:55:37] : [00:55:39]

like it becomes hard but then you get

[00:55:39] : [00:55:40]

better at it

[00:55:40] : [00:55:42]

um I still like not good at it but it's

[00:55:42] : [00:55:44]

just like you get a lot of like inbound

[00:55:44] : [00:55:45]

and a lot of people are like oh can you

[00:55:45] : [00:55:47]

share this and I'm kind of like yeah I

[00:55:47] : [00:55:49]

don't really know like just lots of

[00:55:49] : [00:55:50]

people who want to share things to you

[00:55:50] : [00:55:52]

or like do you want to sponsor this

[00:55:52] : [00:55:54]

event or like with recruiting too the

[00:55:54] : [00:55:56]

funny thing is like recruiting is like

[00:55:56] : [00:55:59]

it's a pretty it's pretty unsexy

[00:55:59] : [00:56:01]

business but there aren't actually like

[00:56:01] : [00:56:02]

that many people it's so funny it's like

[00:56:02] : [00:56:04]

there's so many AI startups so many AI

[00:56:04] : [00:56:07]

VCS really not that many ad recruiters

[00:56:07] : [00:56:09]

like I kind of don't know of any

[00:56:09] : [00:56:11]

um so then if you start to get a

[00:56:11] : [00:56:13]

reputation as like the guy who does AI

[00:56:13] : [00:56:14]

recruiting like a lot of people will

[00:56:14] : [00:56:15]

just come to you and they're like can

[00:56:15] : [00:56:17]

you send us candidates or you know uh

[00:56:17] : [00:56:20]

so just kind of keeping up with that uh

[00:56:20] : [00:56:22]

I'm trying I'm trying to do a better job

[00:56:22] : [00:56:25]

that's really cool so you are a Pioneer

[00:56:25] : [00:56:28]

in this in this subfield a little bit

[00:56:28] : [00:56:30]

yeah it's just some degree

[00:56:30] : [00:56:32]

and that's also the main way you manage

[00:56:32] : [00:56:34]

to monetize your Twitter account right

[00:56:34] : [00:56:37]

were there any other ways that you you

[00:56:37] : [00:56:40]

succeeded in in making money through

[00:56:40] : [00:56:43]

that like that's such a big amount of

[00:56:43] : [00:56:45]

views that that you get

[00:56:45] : [00:56:47]

um that's the only way I've done it so

[00:56:47] : [00:56:49]

far I for a while I started I started a

[00:56:49] : [00:56:51]

job board I had like a job board yeah

[00:56:51] : [00:56:53]

what I found is that I don't know like

[00:56:53] : [00:56:54]

because again I kind of was talking

[00:56:54] : [00:56:56]

about like how recruiting is pretty

[00:56:56] : [00:56:58]

non-scalable and like I would recruit I

[00:56:58] : [00:57:01]

will recruit for startup CEOs and

[00:57:01] : [00:57:02]

they'll tell me that they just don't

[00:57:02] : [00:57:04]

even look at their job submissions

[00:57:04] : [00:57:06]

anymore so then I was kind of like man

[00:57:06] : [00:57:07]

like even if you could make money doing

[00:57:07] : [00:57:10]

this like are you really helping

[00:57:10] : [00:57:12]

um you know so I had this job board for

[00:57:12] : [00:57:14]

a little bit but I kind of just wasn't

[00:57:14] : [00:57:16]

clear if I was actually getting anyone a

[00:57:16] : [00:57:17]

job from it

[00:57:17] : [00:57:19]

um so that was the main way of kind of

[00:57:19] : [00:57:21]

approach monetization

[00:57:21] : [00:57:23]

um I don't know if it's the best way

[00:57:23] : [00:57:26]

um recruiting especially in AI is is

[00:57:26] : [00:57:29]

very high ticket you know if you charge

[00:57:29] : [00:57:31]

a 20 placement fee

[00:57:31] : [00:57:34]

um and someone gets paid a salary of a

[00:57:34] : [00:57:35]

hundred and fifty thousand dollars is

[00:57:35] : [00:57:38]

thirty thousand dollars uh so it's it's

[00:57:38] : [00:57:41]

the the ticket size is very high

[00:57:41] : [00:57:42]

um the the one thing is is very

[00:57:42] : [00:57:45]

non-scalable so it's like I just spend a

[00:57:45] : [00:57:46]

lot of time like sending people emails

[00:57:46] : [00:57:49]

or like pinging Founders or like hopping

[00:57:49] : [00:57:50]

on calls with people whereas I think

[00:57:50] : [00:57:52]

like a lot of the ways that people tend

[00:57:52] : [00:57:53]

to monetize media

[00:57:53] : [00:57:58]

uh via sponsorships or like ads or

[00:57:58] : [00:57:59]

things like that are like much more

[00:57:59] : [00:58:01]

scalable right it's like if you have a

[00:58:01] : [00:58:02]

newsletter or someone just like clicks

[00:58:02] : [00:58:04]

something to post on your job board or

[00:58:04] : [00:58:06]

like post it yeah you know something

[00:58:06] : [00:58:08]

it's just it's it doesn't take an

[00:58:08] : [00:58:09]

additional time

[00:58:09] : [00:58:11]

um on your end I think if your audience

[00:58:11] : [00:58:13]

is interested or like I have thought

[00:58:13] : [00:58:14]

about this stuff like I do think about

[00:58:14] : [00:58:16]

this like a business and do I do

[00:58:16] : [00:58:18]

actually think about making money

[00:58:18] : [00:58:20]

um one thing that I think you just

[00:58:20] : [00:58:22]

make tons of money from that I haven't

[00:58:22] : [00:58:23]

experimented from but I've seen other

[00:58:23] : [00:58:26]

people do is like events um there's just

[00:58:26] : [00:58:29]

so much demand for events in the space

[00:58:29] : [00:58:33]

like hackathons or things like that uh

[00:58:33] : [00:58:35]

real world events right because I've

[00:58:35] : [00:58:38]

I've seen like a drastic drop in online

[00:58:38] : [00:58:40]

events attendance just well just for

[00:58:40] : [00:58:43]

myself I'm not really that interested in

[00:58:43] : [00:58:45]

online events anymore I think we we've

[00:58:45] : [00:58:47]

just all did a

[00:58:47] : [00:58:50]

uh I forgot the word but we we had way

[00:58:50] : [00:58:53]

too much online events during covid and

[00:58:53] : [00:58:57]

everything so yeah I I think a lot of

[00:58:57] : [00:59:00]

people are super into real world events

[00:59:00] : [00:59:02]

now even even more than before Colvin

[00:59:02] : [00:59:06]

maybe and I I assume that's what you

[00:59:06] : [00:59:08]

meant by events and hackathon like

[00:59:08] : [00:59:11]

mainly in person yeah I would I would

[00:59:11] : [00:59:12]

say in person and like I'm not an expert

[00:59:12] : [00:59:14]

here and like I'm so busy that I'm like

[00:59:14] : [00:59:16]

not not doing this personally but it's

[00:59:16] : [00:59:18]

like the economics are really

[00:59:18] : [00:59:19]

interesting you know because it's like

[00:59:19] : [00:59:21]

you you can like even a hackathon you

[00:59:21] : [00:59:23]

don't charge anyone any thing to show up

[00:59:23] : [00:59:25]

like often you can get the event space

[00:59:25] : [00:59:28]

for free from a sponsor and then you

[00:59:28] : [00:59:29]

have like 50 people who will do 50

[00:59:29] : [00:59:31]

different companies that sponsor the

[00:59:31] : [00:59:33]

event for like two thousand dollars or

[00:59:33] : [00:59:34]

something it's like surprise you've made

[00:59:34] : [00:59:36]

a hundred thousand dollars in a day you

[00:59:36] : [00:59:37]

know

[00:59:37] : [00:59:39]

um then I so I've just seen this like

[00:59:39] : [00:59:41]

events space and I've started to talk to

[00:59:41] : [00:59:43]

some people who like do this

[00:59:43] : [00:59:44]

professionally and it's like really

[00:59:44] : [00:59:48]

interesting yeah also I went to I don't

[00:59:48] : [00:59:50]

know if you know about ai4

[00:59:50] : [00:59:53]

oh yeah yeah yeah yeah yeah it's an

[00:59:53] : [00:59:54]

in-person event with lots of

[00:59:54] : [00:59:57]

professionals and these events the

[00:59:57] : [00:59:58]

tickets are like extremely expensive

[00:59:58] : [01:00:00]

it's for companies to basically learn

[01:00:00] : [01:00:03]

about other companies and connect and

[01:00:03] : [01:00:07]

like find Opportunities and it's I'm

[01:00:07] : [01:00:10]

pretty sure it is it it makes a lot of

[01:00:10] : [01:00:13]

money I I don't know personally but it

[01:00:13] : [01:00:17]

seems it seems quite quite beneficial

[01:00:17] : [01:00:19]

yeah

[01:00:19] : [01:00:21]

and okay so my last questions about

[01:00:21] : [01:00:24]

Twitter then I will just conclude with

[01:00:24] : [01:00:25]

some other questions about your other

[01:00:25] : [01:00:28]

projects and and you but

[01:00:28] : [01:00:33]

um so Twitter we all know about like

[01:00:33] : [01:00:35]

Elon Musk and Trump and everything going

[01:00:35] : [01:00:39]

on with Twitter so yeah do you have to

[01:00:39] : [01:00:41]

handle some controversies or like

[01:00:41] : [01:00:45]

polarizing topics or and if you do how

[01:00:45] : [01:00:47]

how do you handle those or can you just

[01:00:47] : [01:00:51]

completely like not respond and not not

[01:00:51] : [01:00:53]

do anything about that

[01:00:53] : [01:00:55]

I think at least for me the the niche

[01:00:55] : [01:00:57]

that I'm in I don't have to engage with

[01:00:57] : [01:00:59]

any of that stuff which I'm I feel very

[01:00:59] : [01:01:00]

fortunate about

[01:01:00] : [01:01:02]

um and I actually try to avoid it

[01:01:02] : [01:01:03]

because it's like I you know I think a

[01:01:03] : [01:01:05]

lot of people are following me for like

[01:01:05] : [01:01:07]

AI research so it's like

[01:01:07] : [01:01:09]

if I started to post about political

[01:01:09] : [01:01:11]

stuff or start liking political things

[01:01:11] : [01:01:13]

like I just think people wouldn't like

[01:01:13] : [01:01:15]

it so I'm I'm very fortunate I think

[01:01:15] : [01:01:16]

like people are following me they want

[01:01:16] : [01:01:19]

AI research they don't you know they

[01:01:19] : [01:01:20]

follow other people for you know Trump

[01:01:20] : [01:01:23]

or Elon

[01:01:23] : [01:01:23]

awesome and that definitely makes sense

[01:01:23] : [01:01:29]

as well I I don't have any issues either

[01:01:29] : [01:01:32]

but I'm also much smarter on Twitter I'm

[01:01:32] : [01:01:34]

just basically sharing my videos so it's

[01:01:34] : [01:01:37]

not really a good way of using Twitter

[01:01:37] : [01:01:39]

but anyways I completely understand that

[01:01:39] : [01:01:42]

and my

[01:01:42] : [01:01:44]

I wanted to talk a bit about your

[01:01:44] : [01:01:46]

podcast but you already mentioned what

[01:01:46] : [01:01:49]

you are doing on there and I I have I've

[01:01:49] : [01:01:51]

also listened to the recent one with uh

[01:01:51] : [01:01:54]

the two the two people I don't remember

[01:01:54] : [01:01:56]

the name but the two people from openai

[01:01:56] : [01:01:58]

which was really cool yeah yeah

[01:01:58] : [01:02:02]

and uh yeah I was just okay I wanted to

[01:02:02] : [01:02:03]

talk just a little bit about the podcast

[01:02:03] : [01:02:06]

because I'm now hosting one as well and

[01:02:06] : [01:02:09]

just maybe that that could be

[01:02:09] : [01:02:12]

interesting to some people because

[01:02:12] : [01:02:15]

um we know that podcasts are becoming

[01:02:15] : [01:02:18]

incredibly popular like a lot of even

[01:02:18] : [01:02:20]

myself everyone is listening to podcasts

[01:02:20] : [01:02:22]

I assume you are too well you mentioned

[01:02:22] : [01:02:24]

that you were listening to One so

[01:02:24] : [01:02:28]

we all are listening to to podcasts and

[01:02:28] : [01:02:31]

so some of our audience may want to even

[01:02:31] : [01:02:33]

create one themselves if they are

[01:02:33] : [01:02:35]

interested into that that's also a good

[01:02:35] : [01:02:37]

way to to be online to have an online

[01:02:37] : [01:02:38]

presence

[01:02:38] : [01:02:40]

and so

[01:02:40] : [01:02:42]

what's to you the most challenging thing

[01:02:42] : [01:02:46]

about hosting a podcast

[01:02:46] : [01:02:46]

oh I don't have a good answer here

[01:02:46] : [01:02:49]

because I feel like

[01:02:49] : [01:02:51]

um I have a very fortunate setup so the

[01:02:51] : [01:02:54]

the podcast was actually proposed to me

[01:02:54] : [01:02:56]

by uh the CEO of this model modern

[01:02:56] : [01:02:59]

monitoring company a rising a guy so

[01:02:59] : [01:03:01]

this is this guy he's he's crazy he runs

[01:03:01] : [01:03:04]

this huge AI company huge AI startup

[01:03:04] : [01:03:05]

he's you know working around the clock

[01:03:05] : [01:03:07]

he somehow he also his name is Jason

[01:03:07] : [01:03:11]

lopateki he also manages to find time to

[01:03:11] : [01:03:14]

like just read AI papers like like I

[01:03:14] : [01:03:15]

don't I don't know how he does it but he

[01:03:15] : [01:03:17]

reads AI research like on a day-to-day

[01:03:17] : [01:03:19]

basis so he he actually reached out to

[01:03:19] : [01:03:20]

me and he's like hey let's do this

[01:03:20] : [01:03:21]

podcast

[01:03:21] : [01:03:23]

um so for me it's actually been very

[01:03:23] : [01:03:24]

easy

[01:03:24] : [01:03:26]

um because he has a marketing team where

[01:03:26] : [01:03:29]

basically we do the recording but then

[01:03:29] : [01:03:30]

um you know his marketing team is

[01:03:30] : [01:03:32]

absolutely excellent and they'll handle

[01:03:32] : [01:03:34]

something like the editing and things

[01:03:34] : [01:03:36]

like that I think for me the hard the

[01:03:36] : [01:03:38]

hard part or the the typically the hard

[01:03:38] : [01:03:40]

part with doing podcast and media is is

[01:03:40] : [01:03:42]

just a lot of the editing stuff and the

[01:03:42] : [01:03:43]

the marketing things like that

[01:03:43] : [01:03:45]

fortunately I do not have to handle that

[01:03:45] : [01:03:47]

um foreign

[01:03:47] : [01:03:49]

on my end like and the funny thing is

[01:03:49] : [01:03:51]

even the the format of the podcast is

[01:03:51] : [01:03:52]

not that hard because it's I don't even

[01:03:52] : [01:03:55]

have to talk too much because I'm just

[01:03:55] : [01:03:57]

I'm interviewing you know AI researchers

[01:03:57] : [01:03:59]

so just ask them about the paper so the

[01:03:59] : [01:04:01]

really the only hard part for me is just

[01:04:01] : [01:04:03]

is is reading is actually reading the

[01:04:03] : [01:04:05]

papers uh but that's that's pretty fun

[01:04:05] : [01:04:08]

for me uh so so typically I'll read you

[01:04:08] : [01:04:10]

know read the paper in advance and come

[01:04:10] : [01:04:11]

up with you know some high-level

[01:04:11] : [01:04:12]

questions

[01:04:12] : [01:04:13]

um I've actually I was like really

[01:04:13] : [01:04:15]

hesitant to take on the podcast because

[01:04:15] : [01:04:17]

I I didn't I wanted to do a really good

[01:04:17] : [01:04:19]

job with it and I thought it would be

[01:04:19] : [01:04:21]

really hard but it hasn't been super

[01:04:21] : [01:04:24]

hard I think I think I think I want to

[01:04:24] : [01:04:26]

increase the volume again you know we're

[01:04:26] : [01:04:29]

doing something like once every three

[01:04:29] : [01:04:30]

weeks or something I think a lot of

[01:04:30] : [01:04:31]

really good podcasts are doing like once

[01:04:31] : [01:04:33]

a week so I think if we could up the

[01:04:33] : [01:04:34]

volume that would be cool

[01:04:34] : [01:04:36]

yeah that's what we call I'd love that

[01:04:36] : [01:04:39]

as well myself and

[01:04:39] : [01:04:41]

what you mentioned that you you like

[01:04:41] : [01:04:45]

doing that and there are no like

[01:04:45] : [01:04:48]

there aren't big challenges for you just

[01:04:48] : [01:04:49]

reading the papers that takes time

[01:04:49] : [01:04:52]

obviously and and coming coming up with

[01:04:52] : [01:04:55]

the questions which I for a tip for

[01:04:55] : [01:04:57]

anyone that hosts an interview I think

[01:04:57] : [01:04:59]

that like of course you need to draft

[01:04:59] : [01:05:02]

some questions and think about the the

[01:05:02] : [01:05:06]

person and and do some research but also

[01:05:06] : [01:05:08]

I've done that in the past it doesn't it

[01:05:08] : [01:05:10]

is the vast major minority of my

[01:05:10] : [01:05:12]

questions but I have also asked

[01:05:12] : [01:05:15]

some AI is like charge GPT to generate

[01:05:15] : [01:05:17]

more questions given the background of

[01:05:17] : [01:05:20]

someone and some question examples it's

[01:05:20] : [01:05:22]

a really good idea yeah it's for for

[01:05:22] : [01:05:26]

real it works pretty good like it's it's

[01:05:26] : [01:05:29]

quite repetitive but if you ask if you

[01:05:29] : [01:05:30]

ask it you will definitely go it will

[01:05:30] : [01:05:32]

definitely come up with at least one

[01:05:32] : [01:05:34]

interesting question that you haven't

[01:05:34] : [01:05:38]

thought of so definitely was gpt4 in the

[01:05:38] : [01:05:40]

long context length like yeah I see I

[01:05:40] : [01:05:42]

feel terrible I actually I I haven't

[01:05:42] : [01:05:44]

used these tools like

[01:05:44] : [01:05:46]

a ton in my day-to-day work but it's

[01:05:46] : [01:05:48]

just the value-add is so obvious like

[01:05:48] : [01:05:49]

yeah especially with a long context

[01:05:49] : [01:05:52]

length of gpt4 like you just copy paste

[01:05:52] : [01:05:54]

like a whole paper in it or even just

[01:05:54] : [01:05:56]

take a screenshot of the paper and just

[01:05:56] : [01:05:58]

say like hey what are some cool high

[01:05:58] : [01:05:59]

level questions that you have about this

[01:05:59] : [01:06:02]

yeah yeah I just started using it well I

[01:06:02] : [01:06:05]

I'm using it more and more now it's it's

[01:06:05] : [01:06:07]

funny because it's my PhD Professor that

[01:06:07] : [01:06:11]

is pushing his students well us to to

[01:06:11] : [01:06:15]

use GPT more so it's it's really it's

[01:06:15] : [01:06:17]

crazy I I wouldn't think that a

[01:06:17] : [01:06:19]

professor would actually recommend using

[01:06:19] : [01:06:22]

an AI to to write and code but he's

[01:06:22] : [01:06:24]

doing that and it's it's extremely

[01:06:24] : [01:06:26]

powerful of course you need to double

[01:06:26] : [01:06:28]

check and like be sure that it's fine

[01:06:28] : [01:06:30]

but it's really helpful especially for

[01:06:30] : [01:06:33]

me that and other people that are not

[01:06:33] : [01:06:36]

native English speaker it's so good to

[01:06:36] : [01:06:38]

just like for example I write an article

[01:06:38] : [01:06:40]

and just and then I just copy paste it

[01:06:40] : [01:06:42]

and ask to keep the information the same

[01:06:42] : [01:06:45]

but rephrase it and it transforms it

[01:06:45] : [01:06:47]

into like a very professional article

[01:06:47] : [01:06:51]

written by like a perfect thing with a

[01:06:51] : [01:06:52]

perfect English this perfect grammar

[01:06:52] : [01:06:54]

everything so it's that's pretty cool

[01:06:54] : [01:06:56]

this is like truly truly magical

[01:06:56] : [01:07:00]

technology that's that's so wild yeah

[01:07:00] : [01:07:01]

and

[01:07:01] : [01:07:03]

yeah what what do you enjoy most about

[01:07:03] : [01:07:05]

hosting a podcast

[01:07:05] : [01:07:05]

I think it's either I just like talking

[01:07:05] : [01:07:09]

to people I think I think it's really

[01:07:09] : [01:07:10]

cool to like talk to the people who are

[01:07:10] : [01:07:11]

actually figuring this stuff out or

[01:07:11] : [01:07:13]

actually building these systems

[01:07:13] : [01:07:16]

um you know I just I feel very fortunate

[01:07:16] : [01:07:18]

that you know I can kind of uh at least

[01:07:18] : [01:07:21]

make a living uh

[01:07:21] : [01:07:24]

understanding and following a lot of

[01:07:24] : [01:07:25]

high-level trends that really matter to

[01:07:25] : [01:07:27]

me so just like just feel very fortunate

[01:07:27] : [01:07:28]

to like

[01:07:28] : [01:07:31]

be able to at least spend a good chunk

[01:07:31] : [01:07:32]

of my time like reading papers and

[01:07:32] : [01:07:33]

thinking about it

[01:07:33] : [01:07:34]

um

[01:07:34] : [01:07:36]

yeah I think I think the coolest thing

[01:07:36] : [01:07:37]

is just being able to follow these

[01:07:37] : [01:07:39]

Trends um and also like talk talk to the

[01:07:39] : [01:07:40]

people because they have like really

[01:07:40] : [01:07:42]

really insightful things to say yeah and

[01:07:42] : [01:07:43]

they're often really good communicators

[01:07:43] : [01:07:45]

I I feel like especially like the last

[01:07:45] : [01:07:47]

episode

[01:07:47] : [01:07:49]

um the last episode on tool former just

[01:07:49] : [01:07:51]

put it out like this week like

[01:07:51] : [01:07:53]

especially Teemo like Teemo just like

[01:07:53] : [01:07:55]

gave a really really good explanation

[01:07:55] : [01:07:56]

like he was just really good and kind of

[01:07:56] : [01:07:58]

lively Communicator so I think this is

[01:07:58] : [01:08:01]

also just fun are you planning on

[01:08:01] : [01:08:03]

interviewing other kind of background

[01:08:03] : [01:08:05]

and researchers with their research

[01:08:05] : [01:08:06]

papers

[01:08:06] : [01:08:08]

yes I don't actually know who I'm gonna

[01:08:08] : [01:08:11]

do next I don't know who's next I I want

[01:08:11] : [01:08:13]

to maybe do something that's not just

[01:08:13] : [01:08:15]

language models because I I I'm I'm

[01:08:15] : [01:08:16]

really into this stuff but I feel like

[01:08:16] : [01:08:19]

maybe I cover it too much

[01:08:19] : [01:08:20]

um

[01:08:20] : [01:08:22]

I it's like obviously the the next

[01:08:22] : [01:08:24]

obvious candidate would be gpt4 but if

[01:08:24] : [01:08:25]

you read the paper they don't really

[01:08:25] : [01:08:27]

want to talk about what they did yeah

[01:08:27] : [01:08:30]

yeah so I don't I don't know if they're

[01:08:30] : [01:08:32]

the right people to to interview

[01:08:32] : [01:08:34]

but all right I also want to say I don't

[01:08:34] : [01:08:36]

mean that I I don't uh I don't want I

[01:08:36] : [01:08:38]

don't mean to say that to like knock

[01:08:38] : [01:08:40]

open Ai No no of course of course it's

[01:08:40] : [01:08:42]

incredible but no it's it's really

[01:08:42] : [01:08:44]

incredible what they did and yeah

[01:08:44] : [01:08:48]

obviously and I just I mainly wanted to

[01:08:48] : [01:08:51]

ask if you were planning to to change a

[01:08:51] : [01:08:53]

bit your interview is like not only

[01:08:53] : [01:08:55]

cover research papers but maybe maybe

[01:08:55] : [01:08:58]

get people from startups or

[01:08:58] : [01:09:00]

anything different in the AI world or

[01:09:00] : [01:09:03]

will this just stick you will stick with

[01:09:03] : [01:09:04]

research papers for this particular but

[01:09:04] : [01:09:05]

yes

[01:09:05] : [01:09:07]

for this particular podcast probably

[01:09:07] : [01:09:08]

just stick with research papers again

[01:09:08] : [01:09:10]

there's just this whole thing where I I

[01:09:10] : [01:09:12]

think the kind of the AI media space is

[01:09:12] : [01:09:14]

pretty underserved

[01:09:14] : [01:09:16]

um there's just there's so much that I

[01:09:16] : [01:09:17]

would like to do that like I don't

[01:09:17] : [01:09:19]

exactly have bandwidth to do like I

[01:09:19] : [01:09:20]

totally I think there's so much more to

[01:09:20] : [01:09:22]

do in the space I don't mean to like

[01:09:22] : [01:09:24]

show this one podcast too much but I

[01:09:24] : [01:09:25]

just do really like them I do think the

[01:09:25] : [01:09:27]

cognitive Revolution is doing a really

[01:09:27] : [01:09:28]

good job there

[01:09:28] : [01:09:29]

um but yeah I like I think there would

[01:09:29] : [01:09:31]

be a ton of cool stuff just to like

[01:09:31] : [01:09:33]

interview people about like machine

[01:09:33] : [01:09:35]

learning careers like just like

[01:09:35] : [01:09:38]

interview people who's like uh machine

[01:09:38] : [01:09:40]

learning Tech lead or like a machine

[01:09:40] : [01:09:41]

learning manager and talk about like how

[01:09:41] : [01:09:43]

did they get there like what's what

[01:09:43] : [01:09:45]

actually matters like in terms of

[01:09:45] : [01:09:46]

learning on the job or getting a good

[01:09:46] : [01:09:48]

job

[01:09:48] : [01:09:50]

um there's there's there's one guy who's

[01:09:50] : [01:09:53]

on uh machine learning Twitter I I hope

[01:09:53] : [01:09:55]

I get his name right or like I I'm not

[01:09:55] : [01:09:57]

sure if I exactly remember his name but

[01:09:57] : [01:09:59]

I think his name is Alexa gordick uh

[01:09:59] : [01:10:00]

he's he's a deep mind I think he

[01:10:00] : [01:10:02]

recently is taking a break from deepmind

[01:10:02] : [01:10:04]

but like his story is really cool where

[01:10:04] : [01:10:07]

he kind of took a an alternate path to

[01:10:07] : [01:10:09]

get into being a research engineer at D

[01:10:09] : [01:10:10]

mind and like doing an interview with

[01:10:10] : [01:10:11]

him would be like super cool just like

[01:10:11] : [01:10:13]

how did he do that how do you pull that

[01:10:13] : [01:10:16]

off and they did yeah

[01:10:16] : [01:10:19]

you're I I will actually have it have

[01:10:19] : [01:10:21]

him on the on the podcast and he he's

[01:10:21] : [01:10:24]

taking a break in Japan I believe for a

[01:10:24] : [01:10:26]

few months and then in I believe in may

[01:10:26] : [01:10:28]

we will have a podcast together so oh

[01:10:28] : [01:10:30]

cool yeah he has a great background

[01:10:30] : [01:10:33]

um okay speaking of uh GPT generated

[01:10:33] : [01:10:36]

questions I think this is the question

[01:10:36] : [01:10:38]

that GPT gave me and I I really liked it

[01:10:38] : [01:10:41]

I had to rephrase it but it it had a

[01:10:41] : [01:10:44]

really good ID and so I will read it

[01:10:44] : [01:10:47]

just to keep it keep the credit to GPT

[01:10:47] : [01:10:49]

um since you are really active on there

[01:10:49] : [01:10:52]

how has social media impacted the way AI

[01:10:52] : [01:10:54]

research is communicated and shared with

[01:10:54] : [01:10:57]

the broader public and I can like maybe

[01:10:57] : [01:10:58]

spend

[01:10:58] : [01:10:59]

um

[01:10:59] : [01:11:01]

specialized to Twitter like how how

[01:11:01] : [01:11:04]

would you see Twitter how are you seeing

[01:11:04] : [01:11:06]

Twitter

[01:11:06] : [01:11:08]

um

[01:11:08] : [01:11:10]

modifying the way resources shared

[01:11:10] : [01:11:14]

compared to the Past

[01:11:14] : [01:11:16]

yeah I mean I think

[01:11:16] : [01:11:18]

that's an interesting question I think

[01:11:18] : [01:11:20]

you can comment Less on how the research

[01:11:20] : [01:11:22]

gets shared to the public because I just

[01:11:22] : [01:11:24]

kind of don't know

[01:11:24] : [01:11:25]

um

[01:11:25] : [01:11:26]

I don't know if I have anything to say

[01:11:26] : [01:11:28]

that's not kind of obvious I mean one

[01:11:28] : [01:11:30]

one thing that is just totally true is

[01:11:30] : [01:11:33]

there's kind of this formal reality of

[01:11:33] : [01:11:36]

how AI research gets disseminated and

[01:11:36] : [01:11:38]

then there's the actual reality like

[01:11:38] : [01:11:40]

there's there's all these conferences

[01:11:40] : [01:11:42]

you know

[01:11:42] : [01:11:43]

um all these journals and stuff but then

[01:11:43] : [01:11:46]

you talk to any of these researchers and

[01:11:46] : [01:11:47]

they're kind of like none of this stuff

[01:11:47] : [01:11:49]

matters anymore it's like it does matter

[01:11:49] : [01:11:51]

especially if you're trying to further

[01:11:51] : [01:11:54]

your academic career or you put you know

[01:11:54] : [01:11:59]

or iclr on your resume or something but

[01:11:59] : [01:12:00]

um insofar as like what are people

[01:12:00] : [01:12:02]

actually using to keep up to date like

[01:12:02] : [01:12:05]

the top researchers in the space it's

[01:12:05] : [01:12:06]

just Twitter

[01:12:06] : [01:12:07]

um it's really like it's the place that

[01:12:07] : [01:12:10]

people are using to uh

[01:12:10] : [01:12:12]

to learn about what's going on

[01:12:12] : [01:12:13]

um

[01:12:13] : [01:12:14]

so I feel like that's kind of my only

[01:12:14] : [01:12:16]

observation is like it's actually even

[01:12:16] : [01:12:18]

even if it's not formally the mechanism

[01:12:18] : [01:12:20]

that people are using to aggregate a lot

[01:12:20] : [01:12:22]

of the most interesting work it's

[01:12:22] : [01:12:24]

I think factually what people use

[01:12:24] : [01:12:27]

um and then as a result like I do think

[01:12:27] : [01:12:29]

you just get

[01:12:29] : [01:12:32]

you can totally change the impact of

[01:12:32] : [01:12:34]

your work by just focusing on marketing

[01:12:34] : [01:12:36]

on on Twitter and there's there's

[01:12:36] : [01:12:37]

there's kind of a hard effort there's a

[01:12:37] : [01:12:40]

weird ethical line to

[01:12:40] : [01:12:43]

tread here where it's like maybe you

[01:12:43] : [01:12:44]

will increase the breadth of your kind

[01:12:44] : [01:12:47]

of the the extent of the replication of

[01:12:47] : [01:12:49]

your research by you know making lots of

[01:12:49] : [01:12:51]

like click-baity threads about your

[01:12:51] : [01:12:53]

research or like making lots of threads

[01:12:53] : [01:12:56]

that kind of overstate the the promises

[01:12:56] : [01:12:57]

the results of your research I think

[01:12:57] : [01:12:59]

that's true

[01:12:59] : [01:13:00]

um but nonetheless I think it's like a

[01:13:00] : [01:13:02]

lot of the research that really gets um

[01:13:02] : [01:13:04]

gets followed up on is is the stuff that

[01:13:04] : [01:13:06]

does what does does well on Twitter so

[01:13:06] : [01:13:08]

if you're an academic I I think it

[01:13:08] : [01:13:09]

actually is quite imperative if you want

[01:13:09] : [01:13:11]

to you know yeah if you actually think

[01:13:11] : [01:13:13]

your work is worth following up on to to

[01:13:13] : [01:13:15]

like you know make Twitter threads about

[01:13:15] : [01:13:17]

it I think

[01:13:17] : [01:13:19]

yeah there I I don't have the reference

[01:13:19] : [01:13:22]

but there has been a study as well where

[01:13:22] : [01:13:25]

like they shared the researchers shared

[01:13:25] : [01:13:28]

on Twitter the half of the papers and

[01:13:28] : [01:13:30]

didn't share the other half and it

[01:13:30] : [01:13:32]

clearly had an impact the Twitter paper

[01:13:32] : [01:13:34]

clearly had an impact in the reads

[01:13:34] : [01:13:36]

obviously but it even had an impact in

[01:13:36] : [01:13:39]

the number of citations the paper had so

[01:13:39] : [01:13:39]

that's

[01:13:39] : [01:13:42]

I don't know if it's a good thing but

[01:13:42] : [01:13:45]

it's definitely worthwhile to share it

[01:13:45] : [01:13:48]

on Twitter if you have published a paper

[01:13:48] : [01:13:51]

but yeah speaking of a good thing or not

[01:13:51] : [01:13:52]

would you say that Twitter has a

[01:13:52] : [01:13:56]

positive impact on the research and

[01:13:56] : [01:13:59]

their research industry like the

[01:13:59] : [01:14:01]

that's really hard to say I don't think

[01:14:01] : [01:14:02]

I'm an expert like I did a year of a

[01:14:02] : [01:14:04]

machine learning PhD I definitely can

[01:14:04] : [01:14:06]

read and understand these things at a

[01:14:06] : [01:14:08]

high level but like I'm not you know I'm

[01:14:08] : [01:14:11]

not you know a 10-year Professor I'm not

[01:14:11] : [01:14:13]

no someone who's done really world-class

[01:14:13] : [01:14:16]

research at deepmind or something

[01:14:16] : [01:14:18]

um so I don't actually feel like I'm

[01:14:18] : [01:14:19]

qualified to comment

[01:14:19] : [01:14:21]

um I think the one like meta comment I

[01:14:21] : [01:14:24]

would have is generally in life I don't

[01:14:24] : [01:14:26]

think of things as being like single

[01:14:26] : [01:14:27]

valued but I think of them as being kind

[01:14:27] : [01:14:29]

of like multi-dimensional or vector

[01:14:29] : [01:14:33]

valued so insofar as I wouldn't know if

[01:14:33] : [01:14:34]

it's like good or bad but I think on

[01:14:34] : [01:14:36]

various Dimensions it would be good or

[01:14:36] : [01:14:39]

bad so it's like one dimension would be

[01:14:39] : [01:14:41]

like yeah probably the more clickbaity

[01:14:41] : [01:14:43]

stuff like gets shared more and people

[01:14:43] : [01:14:45]

believe in it like maybe that's bad

[01:14:45] : [01:14:49]

um but then also like um it's very uh

[01:14:49] : [01:14:52]

it's in many ways very democratized like

[01:14:52] : [01:14:54]

like truly if you're like on the other

[01:14:54] : [01:14:55]

side of globe and you do cool research

[01:14:55] : [01:14:57]

and like you write a cool Twitter Thread

[01:14:57] : [01:14:59]

about it like people will see it um and

[01:14:59] : [01:15:00]

like that's that's really cool so like

[01:15:00] : [01:15:02]

that would be one area where I think

[01:15:02] : [01:15:04]

it's really good

[01:15:04] : [01:15:06]

it definitely lowers the barrier to

[01:15:06] : [01:15:09]

entry like a lot of these kind of old

[01:15:09] : [01:15:11]

boys clubs uh you know

[01:15:11] : [01:15:13]

uh don't matter as much anymore of

[01:15:13] : [01:15:15]

course sir for sure

[01:15:15] : [01:15:18]

and so my my last question

[01:15:18] : [01:15:21]

is you you already talked a lot about

[01:15:21] : [01:15:23]

this so I assume we already know the

[01:15:23] : [01:15:27]

answer but still I I want to

[01:15:27] : [01:15:29]

to ask it it's

[01:15:29] : [01:15:32]

basically well it's it's twofold very

[01:15:32] : [01:15:34]

simple it's just what is your next

[01:15:34] : [01:15:36]

project or like current Obsession that

[01:15:36] : [01:15:40]

you are working on and also what would

[01:15:40] : [01:15:42]

you like the people to know or follow

[01:15:42] : [01:15:45]

from you

[01:15:45] : [01:15:45]

um next project what I'm working on I I

[01:15:45] : [01:15:50]

feel kind of weird it's not super

[01:15:50] : [01:15:52]

related to Twitter but it's kind of an

[01:15:52] : [01:15:54]

outgrowth of the recruiting stuff is um

[01:15:54] : [01:15:56]

there's this one company I'm really

[01:15:56] : [01:15:57]

excited about they're doing really well

[01:15:57] : [01:15:59]

they're trying to hire like top one

[01:15:59] : [01:16:01]

percent Talent um trying to be careful

[01:16:01] : [01:16:03]

about what I say about them so I I won't

[01:16:03] : [01:16:05]

mention who they are um but

[01:16:05] : [01:16:08]

you know if things work out I think uh

[01:16:08] : [01:16:10]

I'm gonna not like work for them as an

[01:16:10] : [01:16:12]

employee but like uh partner more

[01:16:12] : [01:16:14]

closely and just like try to fill a ton

[01:16:14] : [01:16:16]

of their roles um so that for me is like

[01:16:16] : [01:16:18]

a kind of interesting experience where

[01:16:18] : [01:16:19]

it's uh it's more high pressure and it's

[01:16:19] : [01:16:21]

just definitely like

[01:16:21] : [01:16:22]

it's different from what I've been doing

[01:16:22] : [01:16:23]

in the past because in the past I've

[01:16:23] : [01:16:25]

just been this like passive Matchmaker

[01:16:25] : [01:16:26]

like people show up and I'll be like oh

[01:16:26] : [01:16:28]

I think you'd enjoy this job whereas now

[01:16:28] : [01:16:30]

it's like okay here's the org chart like

[01:16:30] : [01:16:32]

here's all the roles we want to fill in

[01:16:32] : [01:16:32]

the next like

[01:16:32] : [01:16:34]

you know four months or something like

[01:16:34] : [01:16:37]

let's do it you know um so I think I

[01:16:37] : [01:16:39]

think that's one of the biggest projects

[01:16:39] : [01:16:40]

that I'm approaching right now that I

[01:16:40] : [01:16:42]

think will be really fun and and really

[01:16:42] : [01:16:44]

hard and really new to me

[01:16:44] : [01:16:45]

um

[01:16:45] : [01:16:48]

and what what to follow me on Twitter or

[01:16:48] : [01:16:50]

uh or what it is yeah would you like to

[01:16:50] : [01:16:53]

mainly share your Twitter account or

[01:16:53] : [01:16:57]

maybe well this is undisclosed so you

[01:16:57] : [01:16:58]

cannot really talk about it that much

[01:16:58] : [01:17:01]

but that much but like I assume AI Pub

[01:17:01] : [01:17:04]

is the best place best place to to learn

[01:17:04] : [01:17:06]

more about you and just follow your work

[01:17:06] : [01:17:08]

but is there would you like to share

[01:17:08] : [01:17:12]

your recruiting platform or oh sure just

[01:17:12] : [01:17:15]

oh yeah sure yeah I mean you it's all

[01:17:15] : [01:17:16]

kind of via Twitter like I don't have a

[01:17:16] : [01:17:17]

newsletter I don't have a website or

[01:17:17] : [01:17:19]

anything so if you just go to my Twitter

[01:17:19] : [01:17:21]

page you'll find everything uh you know

[01:17:21] : [01:17:22]

that is it's all very simple like you

[01:17:22] : [01:17:24]

just click on the link and it'll take

[01:17:24] : [01:17:27]

you to the little form just to submit uh

[01:17:27] : [01:17:29]

you know your preferences for for talent

[01:17:29] : [01:17:30]

referrals

[01:17:30] : [01:17:32]

um I have a personal Twitter account but

[01:17:32] : [01:17:33]

I don't really actually put much effort

[01:17:33] : [01:17:35]

into it like I'm actually really

[01:17:35] : [01:17:37]

deliberate like on my personal account I

[01:17:37] : [01:17:38]

just feel like I can post whatever and I

[01:17:38] : [01:17:40]

don't have a high bar and a lot of it's

[01:17:40] : [01:17:42]

just like totally random personal stuff

[01:17:42] : [01:17:44]

it's like experimenting with sleep or

[01:17:44] : [01:17:47]

Fitness or things like that or oh I

[01:17:47] : [01:17:48]

don't you know I don't do this as much

[01:17:48] : [01:17:50]

anymore but I'll like post about books

[01:17:50] : [01:17:52]

I'm reading uh I was reading some like

[01:17:52] : [01:17:53]

interesting math textbooks in the summer

[01:17:53] : [01:17:55]

so it's like posting about that I'm like

[01:17:55] : [01:17:56]

reading an interesting book on

[01:17:56] : [01:17:59]

negotiation right now so is it the the

[01:17:59] : [01:18:03]

one from the CIA agent or no I will

[01:18:03] : [01:18:05]

maybe I maybe this is the same guy it's

[01:18:05] : [01:18:06]

it's actually a really good book it's

[01:18:06] : [01:18:08]

called Uh the the title sounds very

[01:18:08] : [01:18:11]

Machiavellian the title is um bargaining

[01:18:11] : [01:18:13]

for Advantage and then oh I think yeah

[01:18:13] : [01:18:16]

yeah this the subtitle is um negotiation

[01:18:16] : [01:18:18]

strategies for reasonable people the

[01:18:18] : [01:18:20]

book is much closer to this subtitle

[01:18:20] : [01:18:22]

where it's actually it's very like civil

[01:18:22] : [01:18:25]

yeah about being a good person and

[01:18:25] : [01:18:27]

maintaining relationships and like that

[01:18:27] : [01:18:29]

but it's really good I I'm enjoying it

[01:18:29] : [01:18:31]

okay then I I

[01:18:31] : [01:18:34]

forgot the name obviously but there's

[01:18:34] : [01:18:37]

one called negotiate something and it's

[01:18:37] : [01:18:41]

by an xcia agent and it's like I think

[01:18:41] : [01:18:43]

it's the best book on negotiation and

[01:18:43] : [01:18:45]

it's just applied to the real world

[01:18:45] : [01:18:47]

getting a job your your personal life

[01:18:47] : [01:18:49]

anything it's

[01:18:49] : [01:18:51]

incredibly good it's a really it's a

[01:18:51] : [01:18:54]

really good book so it's I'm just

[01:18:54] : [01:18:56]

opening a tab I'm just curious this is

[01:18:56] : [01:18:58]

yeah

[01:18:58] : [01:18:58]

yeah negotiate and CIA agent you you

[01:18:58] : [01:19:03]

should you should find it cool

[01:19:03] : [01:19:07]

I think just pulled it up never split

[01:19:07] : [01:19:08]

the difference yeah exactly that's

[01:19:08] : [01:19:11]

that's the one yeah a really good one I

[01:19:11] : [01:19:14]

certainly recommend but awesome so I I

[01:19:14] : [01:19:16]

might follow your your private account

[01:19:16] : [01:19:19]

just because I'm really into better

[01:19:19] : [01:19:21]

sleeping habits and and Sports in

[01:19:21] : [01:19:23]

general and just eating and everything

[01:19:23] : [01:19:26]

like for well

[01:19:26] : [01:19:28]

uberman and everything related to that

[01:19:28] : [01:19:32]

I'm I'm super into into this and I'm

[01:19:32] : [01:19:34]

also reading lots of book and you seem

[01:19:34] : [01:19:34]

to be

[01:19:34] : [01:19:38]

reading similar format as as I enjoy so

[01:19:38] : [01:19:40]

well anyways I'll definitely check that

[01:19:40] : [01:19:44]

out as well and I I really invite anyone

[01:19:44] : [01:19:48]

interested in AI a little bit I assume

[01:19:48] : [01:19:50]

you are if you are still listening to

[01:19:50] : [01:19:52]

follow AI Pub on Twitter it's a really

[01:19:52] : [01:19:55]

amazing account and I personally love it

[01:19:55] : [01:19:58]

and thank you very much Brian for your

[01:19:58] : [01:20:00]

time it was amazing to learn from you

[01:20:00] : [01:20:03]

and chat with you and yeah a huge thanks

[01:20:03] : [01:20:05]

for for taking the time to to be there

[01:20:05] : [01:20:07]

no thank you this is like super fun

[01:20:07] : [01:20:11]

really really enjoyed this so

[01:20:11] : [01:20:11]

[Music]

[01:20:11] : [01:20:28]

thank you

[01:20:28] : [01:20:31]

[Music]

[01:20:31] : [01:20:31]


( Spotify या Apple Podcasts पर भी उपलब्ध)।

L O A D I N G
. . . comments & more!

About Author

Louis Bouchard HackerNoon profile picture
Louis Bouchard@whatsai
I explain Artificial Intelligence terms and news to non-experts.

लेबल

इस लेख में चित्रित किया गया था...

Read on Terminal Reader
Read this story in a terminal
 Terminal
Read this story w/o Javascript
Read this story w/o Javascript
 Lite