American International Group, Inc. is an international insurance organization serving commercial, institutional and individual customers. AIG provides property-casualty insurance, life insurance, and retirement services.

Aig Claim Services, Inc was founded in 1977. The company's line of business includes providing insurance agent and broker services for a range of insurance types.

The buyer had previously sold off its Atlanta portfolio but has gained a new foothold in the market with the Midtown acquisition.

AIG Sells Midtown Apartments For $61M

The following video showcases a snippet of the Apache SeaTunnel developer's daily workflow, aided by Co-Pilot. If you haven't yet embraced tools like Co-Pilot, ChatGPT, or private large-scale models for development assistance, you might find yourself quickly outpaced by the industry over the next five years. Those skilled in leveraging AI-powered coding can accelerate their development speed tenfold compared to those without this skill. This isn't an exaggeration – after reading this, you'll likely have a whole new perspective on how AIGC enhances development efficiency. https://www.youtube.com/watch?v=-pBmwLZZtIw&embedable=true Disrupting Traditional Junior Programmer Training with Large Models: Empowering Skills and Experience In the past, junior programmers would be given tasks by mentors, taught basic approaches, and then guided and corrected as they wrote code. They learned from experience and guidance, gradually gaining proficiency. However, the arrival of large models has completely transformed this process. Large models possess extensive knowledge and even some rudimentary reasoning abilities. They've undergone countless practices, learning various codes and business definitions within a company. The scenarios they've encountered far exceed those faced by mentors in their time. They provide potential answers based on developers' requirements and goals. It's akin to having an all-knowing "mentor" available at your beck and call, capable of instantly generating code for your reference and learning. After learning and adjustment, a junior programmer can submit code that surpasses their individual skill level, ready for peer review. So, why wouldn't we utilize large models to boost our development efficiency? How to Utilize Large Models for Assisted Programming? Common tools currently include ChatGPT, Co-Pilot, and private large-scale models. Different methods are applicable depending on the programming scenario: It's evident that humans are best suited for challenging and innovative architecture or new business scenario code. For reused or similar algorithms, using Co-Pilot can rapidly enhance efficiency. ChatGPT can be employed for generating code in similar scenarios with minor modifications. Private large models are suitable for scenarios where data and code security are crucial, and where the model's familiarity with your business knowledge matters – requiring Fine-Tuning. One might argue that not every company can afford private large models. However, this is a misconception. You don't necessarily need to train an entirely new private large model. For most companies, optimizing (Fine-Tuning) an existing open-source large model to understand your business context is sufficient. This process requires only 1-2 graphics cards (e.g., 3090/4090) and a few hours for configuration.(If interested, you can refer to the following article: In this article, it discusses using Apache DolphinScheduler and pre-configured templates to easily train a large model by drag-and-drop.) . Train Your Own Private ChatGPT with the Cost of a Starbucks Coffee Using private large models can directly facilitate the following: Easy access to desired software features. Efficient navigation of complex user manuals and rules. Assisted programming, Txt2SQL, enhancing data programmer efficiency. Leveraging private large models for programming assistance is closer than you think. If you're still skeptical about the era of AIGC-driven automation in programming, consider the example below on how an open-source project utilizes AIGC to enhance development efficiency. Example of Large Model-Driven Automated Programming: Apache SeaTunnel The vision of Apache SeaTunnel is to " " This ambitious goal aims to connect all data sources on the market, including databases, SaaS, middleware, and BinLogs, achieving optimal synchronization. This is an impossible feat for any single company, and even humans struggle to accomplish it amidst thousands of SaaS software options and ever-changing interfaces. So, how did the core team behind Apache SeaTunnel design such software in the AI era? The overall structure is depicted in the following diagram: Connect Everything, Sync as Fast as Lightning. Unlike Flink or Spark, it focuses on efficient memory, CPU, and bandwidth utilization, and ensures data consistency – aspects that lack existing benchmarks. Consequently, most of the code isn't borrowed but developed directly by core engineers, with continuous input and refinements from experts worldwide to keep pace with cutting-edge technology. Firstly, the core computational engine is purpose-built for synchronization. While these still rely primarily on human expertise, they can leverage existing codes and cloud-based practices, with Co-Pilot as a supportive tool. Large models assist by complementing conventional algorithms and reusing code snippets, as demonstrated in the introductory video. Secondly, the core database connectors, such as the Iceberg Connector, are complex implementations that prioritize both accuracy and high data transmission efficiency. In the face of a vast sea of SaaS interfaces – over 5000 in the MarTech sector alone – manual integration is unfeasible. The SeaTunnel core team devised a strategy: abstracting over a dozen interfaces that humans had previously coded, collaborating with ChatGPT to turn them into two interfaces capable of producing elegant code. ChatGPT can comprehend SaaS interface documentation and directly generate relevant code snippets. Within SeaTunnel, this feature is known as "AI Compatibility," facilitating collaboration between AI and humans. It's a kind of "reconciliation" between programmers and AI, ensuring that each contributes where the other excels. However, as an evolving open-source project, it still has room for improvement. Given the open nature of the code, more enthusiasts are likely to fine-tune and automate this feature. For instance, I've heard of a developer planning to create a GPT Coder that monitors SaaS issues on GitHub, uses ChatGPT to generate code, and submits pull requests automatically – taking human-bot collaboration to the extreme. This feature was released in Apache SeaTunnel 2.3.1. Challenges of Large Model-Driven Automated Programming While ChatGPT, Co-Pilot, and similar tools enhance programming, they're not invincible. There are several challenges to consider when generating code with large models: Accuracy concerns. Inability to conduct code reviews. Limited capability for automated testing. No accountability. – even ChatGPT4 may produce code with 90% accuracy. Hence, it's crucial to simplify the process of code generation, as generating incorrect code is a potential pitfall. After rapid code generation by large models, human code review may struggle to keep up, as machines cannot verify if the final code correctly implements the business logic. Attempting code review using large models reveals a plethora of seemingly important but ultimately trivial improvements, obscuring the assessment of logic correctness. Large models can still make mistakes, and this will likely remain the norm for some time Moreover, automated testing and test case generation are still challenging aspects for large models. While TestPilot is actively discussed in academic circles (see Cornell's articles on the topic), its engineering-level application remains distant. Lastly, there's the issue of responsibility – a philosophical rather than technical challenge. Even with a 90% accuracy rate for ChatGPT, let alone a hypothetical 99.9999%, would you trust it to automatically calculate the entire company's payroll and initiate direct salary transfers with banks? If a problem arises, who's accountable? Not all business challenges can be solved with technology, and the same applies to large models. Future Outlook We're currently in the early stages of large model-driven automated programming. Many developers are still experimenting with Co-Pilot and ChatGPT, and most programmers haven't yet utilized private large models to boost coding efficiency based on their company's context. : However, over the next 3-5 years, automated assistance in programming will undoubtedly become a staple tool for our generation of developers Foundational large models will narrow the gap with ChatGPT, improving usability. Open-source large models will enhance accuracy and performance, resulting in improved ROI for more companies that adopt private large models. The threshold for large model-driven automation will decrease, with the emergence of more tools for democratized large model training – beyond . Apache DolphinScheduler Awareness of large model-driven automated programming will increase among technical managers, leading to adaptations in technical management processes. Amid the current economic cycle, leveraging large models to enhance efficiency becomes imperative. Therefore, over the next few years, if your development process remains limited to CRUD operations and you can't harness large models to amplify your expertise and business comprehension by 10 or 100 times, you might not need to wait until you're 35 to be replaced by programmers adept at large model programming. Once they've scaled up by a factor of 10, you might find yourself among the 9 who are left behind. While this might sound a bit alarmist, the trend toward large model-driven automation in programming is an unstoppable force. I'm just a novice explorer in the realm of large model-driven automated programming, and I believe many technical managers and architects will join the wave of large model-driven automation, continuously iterating and optimizing the relationship between humans and large models within the development field. Ultimately, this synergy will make programmers, AI, and technical development processes more effective in serving businesses.

Programmers Not Proficient in AIGC Programming May Face Obsolescence in the Next 5 Years

today announced their plans to expand into Virtual Reality (VR) and Game Content (AIGC). Building upon their GameFi expertise, BinaryX has reportedly entered into agreements with their VR and AIGC partners, AiGC Labs, to develop a groundbreaking new GameFi genre of AI-based VR games. BinaryX Artificial Intelligence First AI-Powered Virtual Reality Game on the Metaverse According to the team, BinaryX is preparing for the launch of the first-ever VR game which heavily features both VR and AIGC in collaboration with AiGC Labs. The team has not released the full details of the game yet, but has hinted that the game will be set in a character-rich universe, inspired by heroes and characters from East Asian folklore. This comes just after Apple’s official reveal of their VR headset, Apple Vision Pro, during the Worldwide Developers Conference event on June 5. Using ChatGPT and GameGPT for Game Design The team revealed that they have been experimenting with ChatGPT and GameGPT to generate engaging gameplay, and characters and build compelling narratives for their games. The AI models helped the team to explore and refine ideas for world-building as well, including ideas for environmental layout, challenges, and level design. Besides the AI-powered VR game in collaboration with AiGC Labs, BinaryX has also been working on a separate AI-powered game that will be released sometime this year. "We are incredibly excited to unveil our latest venture into VR and AIGC technologies," said Rudy, Head of Growth at BinaryX. "This is a significant milestone for our companies as we continue to push the boundaries of innovation in the Web3 gaming industry. Together with our partners, our team has poured talent and dedication into creating a game that will redefine the gaming experience and captivate players like never before. We’re excited to reveal the game to our players soon." The game is currently still under development and more details about the official release will be shared on their social media. This story was distributed as a release by Chainwire under HackerNoon’s Brand As An Author Program. Learn more about the program here: https://business.hackernoon.com/brand-as-author About BinaryX BinaryX is a GameFi and IGO platform, and the force behind popular blockchain games such as CyberChess, CyberDragon, and CyberLand. These games run on the BNB Chain. Aside from developing games, BinaryX has now evolved into a full-fledged GameFi platform by offering services, giving game developers the opportunity to launch their GameFi projects on their platform, and for users to get early access to innovative new games. Initial Game Offering (IGO) As one of the top 10 projects on the BNB Chain, BinaryX has a vast community of more than 100k coin holders and 30K monthly active wallets. It is also one of the top few metaverse projects by trading volume on the BNB chain, with more than 200 million in market cap. BinaryX also has a token, $BNX, that has consistently demonstrated strong performance despite the bear market. For more details and information about BinaryX, please visit | | | | Official website Medium Discord Telegram Twitter Check out BinaryX's games on their and on website YouTube Contact: marketing@binaryx.pro

BinaryX and AiGC Labs Set to Launch AI-Powered VR Game on Metaverse

BinaryX y AiGC Labs se preparan para lanzar un juego de realidad virtual impulsado por IA en Metaverse

BinaryX et AiGC Labs s'apprêtent à lancer un jeu VR alimenté par l'IA sur Metaverse

बाइनरीएक्स और एआईजीसी लैब्स मेटावर्स पर एआई-पावर्ड वीआर गेम लॉन्च करने के लिए तैयार हैं

BinaryX と AiGC Labs がメタバース上で AI 搭載 VR ゲームを開始予定

BinaryX e AiGC Labs preparam-se para lançar um jogo de realidade virtual baseado em IA no Metaverse

Phòng thí nghiệm BinaryX và AiGC được thiết lập để khởi chạy trò chơi VR do AI cung cấp trên Metaverse

BinaryX 和 AiGC Labs 将在元界推出 AI 驱动的 VR 游戏

bench parallel is an ideal promise library that I created to improve inefficiencies that I noticed when using other promise libraries like . On top of being an impressive benchmark exercise, it is a production-ready library that implements the Promise A+ standard. [Aigle](https://github.com/suguru03/aigle) Bluebird Benchmark indicates that is faster than . This was achieved by avoiding unnecessary memory allocation and function calls. For instance, is able to optimize it’s dependencies while Bluebird’s performance is affected if it’s dependencies use different versions of it. For more information on how is faster than , I would encourage you to read on it. The above benchmark exercises Aigle Bluebird Aigle Aigle Bluebird my article Functions makes your code less verbose and contains many functions similar to and . As a result, you would be able to code faster and have a much easier time maintaining code. Aigle [Lodash](https://github.com/lodash/lodash) [Async](https://github.com/caolan/async) Aigle.map vs Bluebird.map Compared to , is 1.5x faster and also supports an object as a first argument unlike . Bluebird.map Aigle.map Bluebird.map const object = {a: 1,b: 2,c: 3};Blubird.map(Object.keys(object), key => {const value = object[key];...});Aigle.map(object, (value, key) => { ... }); If you want to use concurrency, has which is similar to . Aigle Aigle.mapLimit Async.mapLimit const array = [1, 2, 3, 4, 5];Bluebird.map(array, value => {console.log(value); // 1, 2, 5, 4, 3 the order is not ensuredreturn Bluebird.delay(10);}, { concurrency: 2 }); Aigle.mapLimit(array, 2, value => {console.log(value); // 1, 2, 3, 4, 5 the order is ensuredreturn Aigle.delay(10);}); // you could use default limit, the limit is 8Aigle.mapLimit(array, value => {console.log(value); // 1, 2, 3, 4, 5return Aigle.delay(10);}); In addition, supports shorthand notation similar to for cleaner code. Aigle Lodash const array = [{ name: 'barney' }, { name: 'fred' }];Aigle.resolve(array).map('name') // the function is called asynchronously.then(names => console.log(names)); // ['barney', 'fred'] Aigle.mapValues The function is similar to and . It is also similar to , but it returns an object instead. Below is a code comparision between and . Async.mapValues Lodash.mapValues Aigle.map Aigle Bluebird const object = { a: 1, b: 2, c: 3 };const iterator = value => Promise.resolve(value * 2); // Bluebirdconst result = {};Bluebird.each(Object.keys(object), key => {const value = object[key];return iterator(value).then(val => {result[key] = val;});}).then(() => console.log(result)); // { a: 2, b: 4, c: 6 } // AigleAigle.mapValues(object, iterator).then(result => console.log(result)); // { a: 2, b: 4, c: 6 } Aigle.times The function is similar to and . It calls an iterator specified times. Async.times Lodash.times const n = 10;const iterator = value => Promise.resolve(value * 2); // Bluebirdconst times = [];for (let i = 0; i < n; i++) {times.push(i);}Bluebird.map(times, iterator).then(result => console.log(result)); // [0, 2, 4, ..., 18] // AigleAigle.times(n, iterator).then(result => console.log(result)); // [0, 2, 4, ..., 18] Conclusion is a promise library that provides safer, less verbose, and maintainable code that boasts impressive benchmarks. The documented examples above are only a subset of 's functions. If you are already familiar with or , you will be able to easily transition into writing even more readable code with . With that said, let’s enjoy writing faster and more efficient Javascript using . Aigle Aigle Lodash Async Aigle Aigle Reference Aigle Bluebird

Aigle vs Bluebird

Generative Artificial Intelligence for Software Engineering: Background

Introduction Late in 2023, I was considering writing an article about Acorn Labs' work simplifying Kubernetes containers. I even started it and wrote a few paragraphs, but then I got busy with other things. At the time, I was going to revisit it in February 2024, they made a couple of announcements; one was for Rubra, and the other was more of a personal project by one of the co-founders, Darren Shepherd, called GPTScript. This project was so exciting that a month later, Acorn announced they were changing their company focus to GPTScript (not sure what would happen to Rubra). I’ve been following all the developments on this project, closer than I typically do, because I find it absolutely amazing as enablement technology. It reminds me of the philosophy behind a product I worked with back in the 80s and 90s called MPEX from VESOFT, written by Eugen Volokh. What Darren has done with GPTScript is kind of hard to wrap your head around at first, especially for a programmer. I haven’t spent enough time with it yet to fully understand all the implications, but I have a lot of thoughts about it. In this article, I’ll walk through some basic introductions and examples and give you some thoughts about where you could take it. Getting Started I worked off the docs and various videos and examples; I gave this a go on my Windows 11 laptop. Let’s get this sucker installed: winget install gptscript-ai.gptscript
$env:OPENAI_API_KEY = 'your-api-key' The docs suggest running a “hello world” example at this point, but that isn’t very interesting, so let’s try the “bob” example. Also, note that there are many built-in examples you can get a list of by issuing the gptscript—-list-tools command and an online ‘examples’ directory. What does “bob.gpt” look like? This presents us with some obvious questions, like what this all means and how a value of ‘How are you doing?’ ended up in the ‘question’ argument. GPTScript operates around the notion of tools where a tool is described in a GPT file. Each file can be made up of more than one named tool, separated by three dashes. Borrowing from the docs, Tool parameters are key-value pairs defined at the beginning of a tool block before any instructional text. They are specified in the format key: value. The parser recognizes the following keys (case-insensitive and spaces are ignored): A tool starts with a preamble that defines the tool's name, description, args, available tools, and additional parameters. The preamble is followed by the tool's body, which contains the instructions for the tool. Comments in the preamble are lines starting with # and are ignored by the parser. Comments are not really encouraged as the text is typically more useful in the description, argument descriptions, or instructions. I’m still not sure why Bob was giving a default value for the argument. It would change if I gave it an argument, so that was a bit of a mystery. I decided to try the samples-readme.gpt script, which looks like this: Reading this, I understood that it would give me a summary of each gpt file in the examples directory. When I first tried it without an argument, it failed to find any input, so I tried with an example file as an argument, and it worked. Then I tried it with their larger example script for hacker-news-headlines, and instead, it did what I thought it should have done the first time. It created a readme.md file in the examples directory with a summary of each example. I’ve seen people in the community starting to generate some very cool tools that are being added to the GPTScript tools web page. I had a little bit of mixed results with my tests. Most things worked as expected, but I had a few things that didn’t, and I didn’t have the opportunity to dig deeper. That said, this is clearly powerful, and it is just scratching the surface. Summary As a lifelong programmer, this feels weird; it involves doing things in ways that seem like something is missing. This makes GPTScript fascinating and scary as if black magic is involved somehow. The natural language and loose feel of it take a little getting used to, but as you think about it, you start to realize how amazing this can be as an enablement technology. One of the first things that came to mind, but I didn’t test, was could this be used to drive OpenFoam? OpenFoam is open-source CFD software that is rather notorious for how difficult it is to use, which is why I never learned how to use it. What if I could just describe what I wanted to do with it, and it worked? Imagine doing this with any software with an API you could talk to. My mind starts to explode thinking about it. Acorn has created an online “tools” catalog that you can utilize, and they are iterating at an insane pace. This is really exciting. Check out my other What the Heck is… articles at the links below: What The Heck Is DuckDB?
What the Heck Is Malloy?
What the Heck is PRQL?
What the Heck is GlareDB?
What the Heck is SeaTunnel?
What the Heck is LanceDB?
What the heck is SDF?
What the Heck is Paimon?
What the Heck is Proton?
What the Heck is PuppyGraph? Introduction Introduction Late in 2023, I was considering writing an article about Acorn Labs ' work simplifying Kubernetes containers. I even started it and wrote a few paragraphs, but then I got busy with other things. At the time, I was going to revisit it in February 2024, they made a couple of announcements; one was for Rubra, and the other was more of a personal project by one of the co-founders, Darren Shepherd , called GPTScript . Acorn Labs Acorn Labs Darren Shepherd Darren Shepherd GPTScript GPTScript This project was so exciting that a month later, Acorn announced they were changing their company focus to GPTScript (not sure what would happen to Rubra). announced announced I’ve been following all the developments on this project, closer than I typically do, because I find it absolutely amazing as enablement technology. It reminds me of the philosophy behind a product I worked with back in the 80s and 90s called MPEX from VESOFT, written by Eugen Volokh. What Darren has done with GPTScript is kind of hard to wrap your head around at first, especially for a programmer. I haven’t spent enough time with it yet to fully understand all the implications, but I have a lot of thoughts about it. In this article, I’ll walk through some basic introductions and examples and give you some thoughts about where you could take it. Getting Started I worked off the docs and various videos and examples; I gave this a go on my Windows 11 laptop. Let’s get this sucker installed: docs docs winget install gptscript-ai.gptscript
$env:OPENAI_API_KEY = 'your-api-key' winget install gptscript-ai.gptscript
$env:OPENAI_API_KEY = 'your-api-key' The docs suggest running a “hello world” example at this point, but that isn’t very interesting, so let’s try the “bob” example. Also, note that there are many built-in examples you can get a list of by issuing the gptscript—-list-tools command and an online ‘ examples ’ directory. examples examples What does “bob.gpt” look like? This presents us with some obvious questions, like what this all means and how a value of ‘How are you doing?’ ended up in the ‘question’ argument. GPTScript operates around the notion of tools where a tool is described in a GPT file. Each file can be made up of more than one named tool, separated by three dashes. tools Borrowing from the docs, Tool parameters are key-value pairs defined at the beginning of a tool block before any instructional text. They are specified in the format key: value. The parser recognizes the following keys (case-insensitive and spaces are ignored): Tool parameters are key-value pairs defined at the beginning of a tool block before any instructional text. They are specified in the format key: value. The parser recognizes the following keys (case-insensitive and spaces are ignored): A tool starts with a preamble that defines the tool's name, description, args, available tools, and additional parameters. The preamble is followed by the tool's body, which contains the instructions for the tool. Comments in the preamble are lines starting with # and are ignored by the parser. Comments are not really encouraged as the text is typically more useful in the description, argument descriptions, or instructions. A tool starts with a preamble that defines the tool's name, description, args, available tools, and additional parameters. The preamble is followed by the tool's body, which contains the instructions for the tool. Comments in the preamble are lines starting with # and are ignored by the parser. Comments are not really encouraged as the text is typically more useful in the description, argument descriptions, or instructions. I’m still not sure why Bob was giving a default value for the argument. It would change if I gave it an argument, so that was a bit of a mystery. I decided to try the samples-readme.gpt script, which looks like this: Bob samples-readme.gpt Reading this, I understood that it would give me a summary of each gpt file in the examples directory. When I first tried it without an argument, it failed to find any input, so I tried with an example file as an argument, and it worked. Then I tried it with their larger example script for hacker-news-headlines , and instead, it did what I thought it should have done the first time. It created a readme.md file in the examples directory with a summary of each example. hacker-news-headlines readme.md I’ve seen people in the community starting to generate some very cool tools that are being added to the GPTScript tools web page. I had a little bit of mixed results with my tests. Most things worked as expected, but I had a few things that didn’t, and I didn’t have the opportunity to dig deeper. That said, this is clearly powerful, and it is just scratching the surface. Summary As a lifelong programmer, this feels weird; it involves doing things in ways that seem like something is missing. This makes GPTScript fascinating and scary as if black magic is involved somehow. The natural language and loose feel of it take a little getting used to, but as you think about it, you start to realize how amazing this can be as an enablement technology. One of the first things that came to mind, but I didn’t test, was could this be used to drive OpenFoam ? OpenFoam is open-source CFD software that is rather notorious for how difficult it is to use, which is why I never learned how to use it. What if I could just describe what I wanted to do with it, and it worked? Imagine doing this with any software with an API you could talk to. OpenFoam OpenFoam My mind starts to explode thinking about it. Acorn has created an online “ tools ” catalog that you can utilize, and they are iterating at an insane pace. This is really exciting. tools tools Check out my other What the Heck is… articles at the links below: What The Heck Is DuckDB? What the Heck Is Malloy? What the Heck is PRQL? What the Heck is GlareDB? What the Heck is SeaTunnel? What the Heck is LanceDB? What the heck is SDF? What the Heck is Paimon? What the Heck is Proton? What the Heck is PuppyGraph? What The Heck Is DuckDB? What The Heck Is DuckDB? What the Heck Is Malloy? What the Heck Is Malloy? What the Heck is PRQL? What the Heck is PRQL? What the Heck is GlareDB? What the Heck is GlareDB? What the Heck is SeaTunnel? What the Heck is SeaTunnel? What the Heck is LanceDB? What the Heck is LanceDB? What the heck is SDF? What the heck is SDF? What the Heck is Paimon? What the Heck is Paimon? What the Heck is Proton? What the Heck is Proton? What the Heck is PuppyGraph? What the Heck is PuppyGraph?

What the Heck Is GPTScript?

This paper is under CC 4.0 license. available on arxiv Authors: (1) Prajwal Niraula, Department of Earth, Atmospheric and Planetary Sciences; (2) Julien de Wit, Department of Earth, Atmospheric and Planetary Sciences; (3) Benjamin V. Rackham, Department of Earth, Atmospheric and Planetary Sciences; (4) Elsa Ducrot, Astrobiology Research Unit, University of Li`ege; (5) Artem Burdanov, Department of Earth, Atmospheric and Planetary Sciences; (6) Ian J. M. Crossfield, Kansas University Department of Physics and Astronomy; (7) Valerie Van Grootel´, Space Sciences, Technologies and Astrophysics Research (STAR) Institute, University of Li`ege; (8) Catriona Murray, 5Cavendish Laboratory; (9) Lionel J. Garcia, Astrobiology Research Unit, University of Li`ege; (10) Roi Alonso, Instituto de Astrof´ısica de Canarias & Dpto. de Astrof´ısica, Universidad de La Laguna; (11) Corey Beard, Department of Physics & Astronomy, The University of California; (12) Yilen Gomez Maqueo Chew, Instituto de Astronom´ıa, Universidad Nacional Aut´onoma de M´exico, Ciudad Universitaria; (13) Laetitia Delrez, Astrobiology Research Unit, University of Li`ege, Space Sciences, Technologies and Astrophysics Research (STAR) Institute, University of Li`ege & 0Observatoire de lUniversit´e de Gen`eve; (14) Brice-Olivier Demory, University of Bern, Center for Space and Habitability; (15) Benjamin J. Fulton, NASA Exoplanet Science Institute/Caltech-IPAC; (16) Michael Gillon, Astrobiology Research Unit, University of Li`ege; (17) Maximilian N. Gunther, Department of Physics, and Kavli Institute for Astrophysics and Space Research, Massachusetts Institute of Technology; (18) Andrew W. Howard, California Institute of Technology; (19) Howard Issacson, Department of Astronomy, University of California Berkeley; (20) Emmanuel Jehin, Space Sciences, Technologies and Astrophysics Research (STAR) Institute, University of Li`ege; (21) Peter P. Pedersen, Cavendish Laboratory; (22) Francisco J. Pozuelos, Astrobiology Research Unit, University of Li`ege & Space Sciences, Technologies and Astrophysics Research (STAR) Institute, University of Li`ege; (23) Didier Queloz, Cavendish Laboratory; (24) Rafael Rebolo-Lopez, Instituto de Astrof´ısica de Canarias & Dpto. de Astrof´ısica, Universidad de La Laguna'; (25) Sairam Lalitha, School of Physics & Astronomy, University of Birmingham; (26) Daniel Sebastian, Astrobiology Research Unit, University of Li`ege (27) Samantha Thompson, Cavendish Laboratory; (28) Amaury H.M.J. Triaud, School of Physics & Astronomy, University of Birmingham. Table of Links Abstract & Introduction Observation Analysis and Validation Future Prospects & References 4. FUTURE PROSPECTS The search for transiting planets around small stars has been motivated in large part by their potential for atmospheric characterization. Owing to the size and proximity of its host, K2-315b is thus one of the few known terrestrial exoplanets possibly amenable for atmospheric characterization in the next two decades. In order to quantify and contextualize its prospects for atmospheric study, we followed the same approach as for TRAPPIST-1 in Gillon et al. (2016) (see de Wit & Seager 2013), focusing here on all known terrestrial planets. We selected terrestrial planets as planets with a reported radius below 1.6 R⊕ in the NASA Exoplanet Archive[4] (Rogers 2015; Fulton et al. 2017). We thus derive the amplitude of the planets’ signals in transmission as: where Rp is the planetary radius, R∗ is the stellar radius, and heff is the effective atmospheric height, µ is the atmospheric mean molecular mass, T is the atmospheric temperature and g is the local gravity. We assume heff to cover seven atmospheric scale heights, µ the atmospheric mean molecular mass to be 20 amu, and the atmospheric temperature to be the equilibrium temperature for a Bond albedo of 0. For the planets with missing masses, we estimated g using the model of Chen & Kipping (2017). The signal amplitudes are reported in Figure 7 together with the SNR relative to TRAPPIST-1 b’s, calculated by scaling the signal amplitude with the hosts’ brightness in J band. We find that K2-315b fares closely to the outer planets of TRAPPIST-1 in terms of potential for atmospheric exploration with JWST— its warmer and thus larger atmosphere compensating for its larger star. In fact, its relative SNR for transmission spectroscopy is half those of TRAPPIST-1 f–h, meaning that assessing the presence of a µ ∼ 20 atmosphere around the planet would require of the order of 40 transits—four times the ∼10 transits required for a similar assessment for TRAPPIST-1 f–h (Lustig-Yaeger et al. 2019). K2-315b is thus at the very edge of JWST’s capability for atmospheric characterization, mostly due to its “large” host star. The derivation above allows us to rank planets in terms of relative potential for atmospheric characterization, assuming a similar atmospheric scenario. In practice, a significant difference in atmospheric mean molecular mass, surface pressure, and/or cloud/haze altitude will strongly affect the actual potential of a planet for characterization (Lustig-Yaeger et al. 2019). For example, in some cases clouds could render the characterization of the favorable TRAPPIST-1 e difficult even with the generation of instruments following the JWST (i.e., LUVOIR, Pidhorodetska et al. 2020). With an estimated radial velocity semi-amplitude of 1.3 m s−1 (assuming a mass comparable to that of Earth), the planet could be accessible for mass measurements using modern ultra-precise radial velocity instruments. Such possibilities and a ranking amongst the 10 best-suited Earth-sized planets for atmospheric study, EPIC 249631677 b will therefore play an important role in the upcoming era of comparative exoplanetology for terrestrial worlds. It will surely be a prime target for the generation of observatories to follow JWST and bring the field fully into this new era. This project makes use of publicly available K2 data. P.N. would like to acknowledge funding for Kerr Fellowship and Elliot Fellowship at MIT. J.d.W. and MIT gratefully acknowledge financial support from the Heising-Simons Foundation, Dr. and Mrs. Colin Masson and Dr. Peter A. Gilman for Artemis, the first telescope of the SPECULOOS network situated in Tenerife, Spain. B.V.R. thanks the Heising-Simons Foundation for support. V.V.G. is a F.R.S.-FNRS Research Associate. M.G. and E.J. are F.R.S.-FNRS Senior Research Associates. This work was also partially supported by a grant from the Simons Foundation (PI Queloz, grant number 327127), and funding from the European Research Council under the European Union’s Seventh Framework Programme (FP/2007-2013) ERC Grant Agreement Number 336480, from the ARC grant for Concerted Research Actions financed by the Wallonia-Brussels Federation, from the Balzan Prize Foundation, from F.R.S-FNRS (Research Project ID T010920F), from the European Research Council (ERC) under the European Unions Horizon 2020 research and innovation programme (grant agreement n◦ 803193/BEBOP), and from a Leverhulme Trust Research Project Grant n◦ RPG-2018-418. Acknowledgments: Keck:I (HIRES), Kepler, MAST (HLSP, K2), SNO, SSO Facilities: astropy (Astropy Collaboration et al. 2013, 2018), batman (Kreidberg 2015), emcee (Foreman-Mackey et al. 2013), isochrones (Morton 2015), prose (Garcia et al. in prep.), SPOCK (Sebastian et al. in prep.), transitleastsquares (Hippke & Heller 2019), w¯otan (Hippke et al. 2019) Software: REFERENCES Allard, F., Homeier, D., & Freytag, B. 2012, Philosophical Transactions of the Royal Society of London Series A, 370, 2765, doi: 10.1098/rsta.2011.0269 Astropy Collaboration, Robitaille, T. P., Tollerud, E. J., et al. 2013, A&A, 558, A33, doi: 10.1051/0004-6361/201322068 Astropy Collaboration, Price-Whelan, A. M., Sip˝ocz, B. M., et al. 2018, AJ, 156, 123, doi: 10.3847/1538-3881/aabc4f Bensby, T., Feltzing, S., & Oey, M. S. 2014, A&A, 562, A71, doi: 10.1051/0004-6361/201322631 Borucki, W. J., Koch, D., Basri, G., et al. 2010, Science, 327, 977, doi: 10.1126/science.1185402 Buchner, J., Georgakakis, A., Nandra, K., et al. 2014, A&A, 564, A125, doi: 10.1051/0004-6361/201322971 Burdanov, A., Delrez, L., Gillon, M., & Jehin, E. 2018, SPECULOOS Exoplanet Search and Its Prototype on TRAPPIST, 130, doi: 10.1007/978-3-319-55333-7 130 Chen, J., & Kipping, D. 2017, ApJ, 834, 17, doi: 10.3847/1538-4357/834/1/17 Choi, J., Dotter, A., Conroy, C., et al. 2016, ApJ, 823, 102, doi: 10.3847/0004-637X/823/2/102 Christiansen, J. L., Crossfield, I. J. M., Barentsen, G., et al. 2018, AJ, 155, 57, doi: 10.3847/1538-3881/aa9be0 Chubak, C., Marcy, G., Fischer, D. A., et al. 2012, arXiv e-prints. https://arxiv.org/abs/1207.6212 Cutri, R. M., Skrutskie, M. F., van Dyk, S., et al. 2003, VizieR Online Data Catalog, II/246 Cutri et al., R. M. 2013, VizieR Online Data Catalog, II/328 de Wit, J., & Seager, S. 2013, Science, 342, 1473, doi: 10.1126/science.1245450 Delrez, L., Gillon, M., Queloz, D., et al. 2018, in Society of Photo-Optical Instrumentation Engineers (SPIE) Conference Series, Vol. 10700, Proc. SPIE, 107001I, doi: 10.1117/12.2312475 Dotter, A. 2016, ApJS, 222, 8, doi: 10.3847/0067-0049/222/1/8 Dressing, C. D., & Charbonneau, D. 2013, ApJ, 767, 95, doi: 10.1088/0004-637X/767/1/95 Esselstein, R., Aigrain, S., Vanderburg, A., et al. 2018, ApJ, 859, 167, doi: 10.3847/1538-4357/aac20e Fernandes, C. S., Van Grootel, V., Salmon, S. J. A. J., et al. 2019, ApJ, 879, 94, doi: 10.3847/1538-4357/ab2333 Feroz, F., Hobson, M. P., & Bridges, M. 2009, MNRAS, 398, 1601, doi: 10.1111/j.1365-2966.2009.14548.x Foreman-Mackey, D., Hogg, D. W., Lang, D., & Goodman, J. 2013, PASP, 125, 306, doi: 10.1086/670067 Fressin, F., Torres, G., Charbonneau, D., et al. 2013, ApJ, 766, 81, doi: 10.1088/0004-637X/766/2/81 Fulton, B. J., Petigura, E. A., Howard, A. W., et al. 2017, AJ, 154, 109, doi: 10.3847/1538-3881/aa80eb Gaia Collaboration, Brown, A. G. A., Vallenari, A., et al. 2018, A&A, 616, A1, doi: 10.1051/0004-6361/201833051 Garcia, L. J., Gillon, M., & SPECULOOS. in prep. Gillon, M., Jehin, E., Fumel, A., Magain, P., & Queloz, D. 2013, in European Physical Journal Web of Conferences, Vol. 47, European Physical Journal Web of Conferences, 03001, doi: 10.1051/epjconf/20134703001 Gillon, M., Jehin, E., Lederer, S. M., et al. 2016, Nature, 533, 221, doi: 10.1038/nature17448 Gillon, M., Triaud, A. H. M. J., Demory, B.-O., et al. 2017, Nature, 542, 456, doi: 10.1038/nature21360 Gillon, M., Meadows, V., Agol, E., et al. 2020, arXiv e-prints, arXiv:2002.04798. https://arxiv.org/abs/2002.04798 Goldreich, P., & Soter, S. 1966, Icarus, 5, 375, doi: 10.1016/0019-1035(66)90051-0 G¨unther, M. N., Pozuelos, F. J., Dittmann, J. A., et al. 2019, Nature Astronomy, 3, 1099, doi: 10.1038/s41550-019-0845-5 G¨unther, M. N., Zhan, Z., Seager, S., et al. 2020, AJ, 159, 60, doi: 10.3847/1538-3881/ab5d3a Hardegree-Ullman, K. K., Cushing, M. C., Muirhead, P. S., & Christiansen, J. L. 2019, AJ, 158, 75, doi: 10.3847/1538-3881/ab21d2 Hippke, M., David, T. J., Mulders, G. D., & Heller, R. 2019, AJ, 158, 143, doi: 10.3847/1538-3881/ab3984 Hippke, M., & Heller, R. 2019, A&A, 623, A39, doi: 10.1051/0004-6361/201834672 Howard, A. W., Johnson, J. A., Marcy, G. W., et al. 2010, ApJ, 721, 1467, doi: 10.1088/0004-637X/721/2/1467 Howell, S. B., Sobeck, C., Haas, M., et al. 2014, PASP, 126, 398, doi: 10.1086/676406 Huber, D., Bryson, S. T., Haas, M. R., et al. 2016, ApJS, 224, 2, doi: 10.3847/0067-0049/224/1/2 Ilin, E., Schmidt, S. J., Davenport, J. R. A., & Strassmeier, K. G. 2019, A&A, 622, A133, doi: 10.1051/0004-6361/201834400 Jao, W.-C., Henry, T. J., Gies, D. R., & Hambly, N. C. 2018, ApJL, 861, L11, doi: 10.3847/2041-8213/aacdf6 Jehin, E., Gillon, M., Queloz, D., et al. 2018, The Messenger, 174, 2, doi: 10.18727/0722-6691/5105 Kass, R. E., & Raftery, A. E. 1995, Journal of the American Statistical Association, 90, 773 Kipping, D. M. 2013, MNRAS, 435, 2152, doi: 10.1093/mnras/stt1435 Kolbl, R., Marcy, G. W., Isaacson, H., & Howard, A. W. 2015, AJ, 149, 18, doi: 10.1088/0004-6256/149/1/18 Kostov, V. B., Schlieder, J. E., Barclay, T., et al. 2019, AJ, 158, 32, doi: 10.3847/1538-3881/ab2459 Kreidberg, L. 2015, PASP, 127, 1161, doi: 10.1086/683602 Lasker, B. M., Sturch, C. R., McLean, B. J., et al. 1990, AJ, 99, 2019, doi: 10.1086/115483 Luger, R., Agol, E., Kruse, E., et al. 2016, AJ, 152, 100, doi: 10.3847/0004-6256/152/4/100 Lustig-Yaeger, J., Meadows, V. S., & Lincowski, A. P. 2019, AJ, 158, 27, doi: 10.3847/1538-3881/ab21e0 Mann, A. W., Dupuy, T., Kraus, A. L., et al. 2019, ApJ, 871, 63, doi: 10.3847/1538-4357/aaf3bc Morris, S. L. 1985, ApJ, 295, 143, doi: 10.1086/163359 Morton, T. D. 2015, isochrones: Stellar model grid package. http://ascl.net/1503.010 Mulders, G. D., Pascucci, I., & Apai, D. 2015, ApJ, 814, 130, doi: 10.1088/0004-637X/814/2/130 Murray, C. A., Delrez, L., Pedersen, P. P., et al. 2020, MNRAS, 495, 2446, doi: 10.1093/mnras/staa1283 Niraula, P., Redfield, S., de Wit, J., et al. 2018, arXiv e-prints, arXiv:1812.09227. https://arxiv.org/abs/1812.09227 Niraula, P., Redfield, S., Dai, F., et al. 2017, AJ, 154, 266, doi: 10.3847/1538-3881/aa957c Ochsenbein, F., Bauer, P., & Marcout, J. 2000, A&AS, 143, 23, doi: 10.1051/aas:2000169 Patra, K. C., Winn, J. N., Holman, M. J., et al. 2017, AJ, 154, 4, doi: 10.3847/1538-3881/aa6d75 Paxton, B., Bildsten, L., Dotter, A., et al. 2011, ApJS, 192, 3, doi: 10.1088/0067-0049/192/1/3 Paxton, B., Cantiello, M., Arras, P., et al. 2013, ApJS, 208, 4, doi: 10.1088/0067-0049/208/1/4 Paxton, B., Marchant, P., Schwab, J., et al. 2015, ApJS, 220, 15, doi: 10.1088/0067-0049/220/1/15 Pidhorodetska, D., Fauchez, T., Villanueva, G., Domagal-Goldman, S., & Kopparapu, R. K. 2020, arXiv e-prints, arXiv:2001.01338. https://arxiv.org/abs/2001.01338 Quinn, S. N., Becker, J. C., Rodriguez, J. E., et al. 2019, AJ, 158, 177, doi: 10.3847/1538-3881/ab3f2b Rabus, M., Lachaume, R., Jord´an, A., et al. 2019, MNRAS, 484, 2674, doi: 10.1093/mnras/sty3430 Reid, I. N., Brewer, C., Brucato, R. J., et al. 1991, PASP, 103, 661, doi: 10.1086/132866 Reiners, A., Joshi, N., & Goldman, B. 2012, AJ, 143, 93, doi: 10.1088/0004-6256/143/4/93 Rodriguez, D. 2016, Dr-Rodriguez/Kinematics-App: Stellar Kinematics V1.0, v1.0, Zenodo, doi: 10.5281/zenodo.192159 Rodriguez, J. E., Vanderburg, A., Eastman, J. D., et al. 2018, AJ, 155, 72, doi: 10.3847/1538-3881/aaa292 Rogers, L. A. 2015, ApJ, 801, 41, doi: 10.1088/0004-637X/801/1/41 Seager, S., & Mall´en-Ornelas, G. 2003, ApJ, 585, 1038, doi: 10.1086/346105 Sebastian, D., Gillon, M., & SPECULOOS. in prep. Stassun, K. G., & Torres, G. 2018, ApJ, 862, 61, doi: 10.3847/1538-4357/aacafc Stassun, K. G., Oelkers, R. J., Paegert, M., et al. 2019, AJ, 158, 138, doi: 10.3847/1538-3881/ab3467 Van Grootel, V., Fernandes, C. S., Gillon, M., et al. 2018, ApJ, 853, 30, doi: 10.3847/1538-4357/aaa023 Vanderburg, A., & Johnson, J. A. 2014, PASP, 126, 948, doi: 10.1086/678764 Vanderburg, A., Johnson, J. A., Rappaport, S., et al. 2015, Nature, 526, 546, doi: 10.1038/nature15527 Vogt, S. S., Allen, S. L., Bigelow, B. C., et al. 1994, in Proc. SPIE, Vol. 2198, Instrumentation in Astronomy VIII, ed. D. L. Crawford & E. R. Craine, 362 Winn, J. N. 2010, Exoplanet Transits and Occultations, ed. S. Seager, 55–77 Yee, S. W., Petigura, E. A., & von Braun, K. 2017, ArXiv e-prints. https://arxiv.org/abs/1701.00922 [4] https://exoplanetarchive.ipac.caltech.edu

π Earth: a 3.14-day Earth-sized Planet from K2's Kitchen Served Warm: Future Prospects & References

320 Stories To Learn About Deep Learning

The news of The New York Times filing a lawsuit against Microsoft and OpenAI for copyright infringement struck the internet like a rogue comet, sparking a fierce debate about the ‘inherent dangers’ of AI. One thing is clear: there has been no shortage of lawsuits contesting the legality or fair use of a work of art in the digital space, with intellectual property being the subject matter. Meanwhile, concerns over copyright infringement in AI have recently increased, highlighting the need to protect the ownership of AI-generated content. Ringfence, a web3-based platform for Generative AI creators, is building the infrastructure for creators to protect and monetize various media, including photos, images, videos, documents, and music. In this interview, Ringfence CEO offered valuable insights into Artificial Intelligence Generated Content(AIGC)and the need to incentivise creators. Whitney Gibbs Mr. Gibbs, can you tell us about yourself and your background? My journey in web3 began in 2017, buying the top of the bull market.  Over the next year, I researched different projects and communities and built relationships in the space. In 2018, I found my way to GPU mining and all things Proof of Work. That year, I began what would evolve into the HASHR8 Podcast (later Compass Mining Podcast, now The Mining Pod). For the next two years, I focused on providing the crypto mining community with valuable insights, and in 2020, the podcast was recognized by Forbes as one of crypto’s most important podcasts. Around this time, I started Compass Mining with two friends. As CEO from 2020-2022, I lead Compass’s growth from zero to becoming the world’s largest marketplace for hardware and hosting. Bitcoin mining Compass was (and still is) an amazing business doing hundreds of millions of dollars in annual sales, and we built a great team of nearly 100 people around the world. The coolest thing about Compass was that we created the “retail mining” category, opening Bitcoin mining up to people who would have otherwise been excluded. In June 2022, I stepped down as CEO of Compass Mining to begin researching new market opportunities to see how I could best impact the global adoption of web3. It was clear to me that artificial intelligence was the most important technological advancement in our world since the internet, so all research was focused on where blockchain technologies and AI might intersect. The most immediate, clear intersection is the establishment of digital provenance and attribution in the training of neural networks. There are millions of people around the world currently providing their data to large companies like FOR FREE. While I don't think this is a malicious action on the part of those companies, there is a better way to do things. This is ultimately what led to the formation of Ringfence. OpenAI and Stable Diffusion Congratulations on your fundraising success. I would like to know what you think of the ongoing lawsuit against Microsoft and OpenAI. Thank you! The New York Times’ lawsuit is a landmark legal case—there’s little doubt its outcomes will shape the future landscape of generative AI. The way some see it, The New York Times is fighting on behalf of journalists and authors worldwide to get answers. The way we see it, it’s a major case among many where creators—both companies and individuals—are fighting for their creative rights. But you can’t ignore that it’s incredibly nuanced. In my opinion, the New York Times case has merit: These companies train their data on billions and billions of data points, including from media outlets, and these chatbots are perfectly capable of regenerating information strikingly similar to what a journalist produces daily—even articles behind paywalls. But NYT is also alleging that these chatbots can regurgitate articles word-for-word. It remains to be seen exactly how these tools, in this case, ChatGPT specifically, were able to do this and whether this process relied more on external sources via web searching rather than an internal training set. This will be an important point in the case—if it's the former, it might be fair to say it comes under Fair Use. And if it’s the latter? The case might just raise more questions than it answers. Ultimately, I think what’s most troubling to creators is that some of these AI companies have claimed that their work is transformative enough of the original training data that it comes under Fair Use. It seems unfair to creators to be able to take their work en masse, train an AI tool, and then use that tool for commercial purposes without giving the creators due credit or compensation. This is what Ringfence is built to solve. But whatever the case and its outcome, I think the New York Times, like other creators before and after it, has every right to fight to regain creative and financial control over its intellectual property. How did copyright infringement become a burning issue in Generative AI? Copyright has long been a burning issue when it comes to Intellectual Property rights and digital ownership. As Generative AI has only made it easier to generate digital content, it’s no surprise that this issue has been exacerbated. The digital output of these models relies on the immense amount of information used to train the Large Language Model behind the Generative AI without consent or compensation to the original creators. There are countless cases, really since the inception of publicly available , of artists and authors having their work scraped from the internet and used to train them. More recently, we now have AI creating music imitating certain artists and even AI movies in the style of different directors. Generative AI models As these tools have become more advanced, they’ve been able to reproduce artistically similar images, text, and now even video and audio. With them also being commercialized, the copyright vs. Fair Use debate has only become more foggy. How do you leverage AI and blockchain technology? Blockchain is fundamental to Ringfence, providing the infrastructure for digital authentication and monetization. In our case, creations uploaded to the cloud can be tokenized, with their metadata also stored on-chain. This serves a few purposes. Firstly, blockchain provides a transparent and immutable ledger for tracking the digital provenance of creation, providing a clear record of ownership. Secondly, the tokenization also offers digital ownership itself, facilitating peer-to-peer transfer. We also make use of smart contracts to facilitate enforceable end-user agreements, enabling fair and seamless monetization. With these, creators can set clear rules for how AI gets to use their work, whether it can be personalized or used to train models, and how they prefer to be compensated when their work is used to create AI derivatives. Generative AI can be a tool to enable hyperproductivity. Artists, authors, and creatives of any kind should be able to make use of it in a way that respects and rewards their creativity. Creators should be able to use their work to rapidly iterate and create derivatives, but they should also be able to license their work to train AI models—and get rewarded. Why is content monetization one of your key objectives? Data is the most valuable commodity in the world, yet rarely are those supplying it the same ones benefiting from it. At present, this is most true with AI. Companies like OpenAI, , and Stable Diffusion garner billion-dollar valuations and generate million-dollar monthly profits while paying out NOTHING to those who have supplied the key data points without which they would cease to function. Midjourney For Ringfence, we believe that content monetization is a critical first step toward the monetization of all personal data. Through content monetization and the neural network needed to support it, we intend to answer important questions around attribution, compensation, and true fair use. Can you go into more detail about your recent fundraising success? Raising capital in 2023 was no small task, even with a compelling idea and sound business plan. The end of the era of zero-interest rates, coupled with global economic uncertainty, changed the venture landscape and made raising capital hard for many. Ringfence was fortunate to secure funding from a great cohort of investors who believe in our vision for the future and were excited about backing our experienced team. Any parting words? AI is a tool or technology like any other—it can have an immensely positive impact, but it might also be incredibly detrimental to creators without us taking a step back and assessing it closely. Ultimately, creatives shouldn't have to compromise or buckle under the threat of AI taking their work. They should be able to create freely, working alongside AI to improve their productivity while being fairly rewarded. At the risk of sounding a little too sentimental, I hope that we can all work towards a shared future where creatives can collaborate with AI fairly, bringing in a new age of creativity—but where innovation can continue.

Ringfence CEO Whitney Gibbs On Copyright Infringement And Compensation For AI and Web3 Creators

Gemini - A Family of Highly Capable Multimodal Models: Discussion and Conclusion, References

ERC-6551: Powering the Re-rising of the NFT

The US Government Sets Plans in Motion to Attract AI Talent to the Country 

Today, many content creators on the internet have already begun using AIGC tools to significantly improve their content production efficiency. However, content creators only make up a small portion of all internet users. The majority, ordinary people, can also fully utilize the rapidly developing artificial intelligence technology, and the tools and products empowered by it, to gain advantages in fierce competition. For most people, career development is related to one’s life success and happiness and is an essential aspect throughout one’s life. Job hunting or job hopping is one of the most critical topics in people’s career development. Recently, I discovered that many tech-savvy individuals have been using some powerful artificial intelligence tools to aid them in job hunting or job hopping. Thus, I decided to personally try out several AI resume creation and modification tools that currently have the highest user numbers on the market, and share my experience and evaluation results with you. Can these tools genuinely assist us in better preparing our resumes and finding our dream jobs? What unique features and advantages do they possess? Let's explore the answers to these questions together. Principles and Evaluation Dimensions of AI Tools Principle of these AI tools: Most companies today use different suppliers' applicant tracking systems (ATS) to manage recruitment data and processes. Simply put, the ATS is similar to a CRM system or database, but it can also automate the recruitment process. Meaning, if your resume does not meet the requirements and does not reach a corresponding rating or score, it will be directly placed in the so-called talent pool, and will not appear in front of the recruiter. Conversely, if your resume is seen by the ATS system as a high match, you are likely to receive an interview invitation. The principle of AI tools is to scan your resume or cover letter, check job keywords, skills, work experience, etc., whether they match the target position, and whether the format meets the ATS reading requirements. By optimizing the score or rating of your resume, it tries to present your resume in front of recruiters as much as possible. 2. Tool Selection Basis and Objectivity: Considering that there are many such AI products on the market currently, but many are fleeting and cannot stand the long-term test of the market and users. Therefore, I found a few tools in this track with the most users, the best reputation, and relatively comprehensive functions. Then, I surveyed their traffic for the last three months through SimilarWeb and ranked them. I chose the three most popular AI tools on the market until September for the next step of evaluation. They are , and . I will put their links at the end of the article for your reference. ResumeWorded Jobscan, Kickresume 3. Evaluation Dimensions and Condition Settings: Considering the goals users hope to achieve with these tools and the costs they will incur, I mainly tested and evaluated them from the following dimensions: the core functions of the tools, whether the business logic and processes are clear, the tool's price and cost, the usability of the free version, and the main disadvantages of the tools. The full score is five stars. For evaluation materials, I assumed that I was a developer looking for a job or seeking job-hopping opportunities, then generated a common resume with GPT-4, and used its materials to create a test LinkedIn account. I will also put the resumes and LinkedIn accounts I used for testing in the link below for everyone to check. Performance and Ratings of AI Tools According to the data we previously collected, this tool has over 2.7 million monthly visits and is still growing. Based on experience, the number of users corresponding to this traffic is approximately 270,000 to 540,000. For an AI tool, this is quite an astonishing figure. Therefore, we had high expectations. 1. First, I tried the most popular tool on the market, resumeworded. After logging in, we could see that this tool has three core functions: resume optimization, resume content matching according to JD (Job Description), and LinkedIn account optimization. These features basically cover the process of job search, including network expansion, job hunting, and resume writing. First, I used the Score My Resume feature to upload my resume. The system scored my resume within 15 seconds and provided suggestions for improvement. I had two options at this point since the system analyzed my resume in terms of Impact, Brevity, Style, and Section, as well as the Soft Skills and Hard Skills presented in my resume. The first option was to improve the content corresponding to each aspect based on the system's suggestions to enhance the resume score. The second option was to use the Line Analysis feature in the Tools to analyze the content of the resume line by line and manually modify it based on the system’s suggestions, or use the Magic Write feature to let AI modify the content. Free users have two opportunities to use the Magic Write feature. I tried it, and the generated content was indeed more specific and persuasive than what I wrote myself. Then, I tried the LinkedIn Review feature. I found it not much different from the previous resume optimization feature. Note that the system cannot directly analyze the LinkedIn account links. Instead, you need to download the content as a PDF file from the LinkedIn Profile’s more options, and then upload this file to resumeworded. After uploading the PDF file, the AI will also score the profile and provide optimization suggestions based on Headline, Summary, Experience, Education, etc. Besides, it has two unique features: one is keyword optimization for LinkedIn searches, and the other is templates for emails and messages needed for network expansion on LinkedIn. However, these features are chargeable. Finally, we checked the subscription price of this AI tool. It costs $49 per month. Typically, the job search cycle requires at least a quarter. Therefore, I believe a quarterly subscription might be a more realistic choice for most users. Its quarterly price is $99. Considering that the free version of this tool offers quite limited features, serious job seekers almost have to pay the subscription fee, which objectively speaking, is not cheap. In consideration of the above evaluation content: It has relatively complete core functions, earning it 1 star. The business logic clarity is average, 0.9 stars; cost, 0.7 stars; usability of the free version, 0.3 stars; no significant drawbacks found in the tool, 1 star, making the final score 3.9 stars. 2. Next, we tried jobscan, currently ranked second by users. According to the data we previously collected, this tool has over 1.9 million monthly visits and is also continually growing. Based on experience, the number of users corresponding to this traffic is approximately 190,000 to 380,000. After logging in, we found this tool to be more like a personal version of ATS (Applicant Tracking System). Its core functions, besides the common resume scanning optimization and LinkedIn account optimization, also include a job and interview management system. It actually has another vital but free feature - Resume Builder. I don't know why it's not placed in the central core function area of the page but is located in the Quick Links area at the bottom of the page. I tried this feature, and it provides simple templates for you to fill in information and then generates an ATS-friendly resume. But it’s just format-friendly to ATS, so you must use the Resume Scan in the core functions to perfect the resume you generated. The Resume Scan of this tool is not merely scanning the resume and giving optimization suggestions. It is similar to the resume content matching according to JD feature of the previous tool. You need to input the JD of the target position while uploading the resume. So here, I uploaded that test resume of mine and found a random senior software engineer position JD on indeed.com to input. The scanning result was roughly as follows: The AI provides very detailed optimization suggestions, covering resume format, keyword matching, Hard Skills and Soft Skills matching, Recruiter tips, and other regular content. You can also input the company of the target position to achieve deep optimization for a specific ATS system, which is quite professional. Regrettably, only paying users can use its Power Edit feature and the Cover Letter optimization. Although I couldn’t try them this time, I specifically watched the demonstration videos on the official website. Both features are impressive. Then, I tried its account optimization feature. Like the previous tool, it also requires uploading the downloaded LinkedIn profile file. After scanning this file, I saw a very detailed report generated by the system. Unfortunately, like the first tool, most of the content can only be unlocked by paying users. LinkedIn Then, I tried its Job Tracker. Like the Resume Builder, it is a free but relatively simple tool and does not involve AI-related capabilities. We can add new positions in it, managing the entire process from resume submission to getting an offer. Finally, we still checked the subscription price of this AI tool, which is actually not much different from the first one. It subscribes quarterly, and the cost for a quarter is $90, about 10% cheaper than the first one. However, we found that, like the first tool, the free version can only be used for testing and experiencing. If you really want to use it to help find a job, you must use the paid version. In consideration of the above evaluation content: It has relatively complete core functions, earning it 1 star. The business logic and process are very clear, 0.9 stars; cost, 0.8 stars; usability of the free version, 0.4 stars; no significant drawbacks found in the tool, 1 star, making the final score 4.1 stars. . Based on the data we collected earlier, this tool receives over 1.8 million visits per month. Although this is a significant decrease compared to 3 million in July, it’s still quite impressive. 3. Lastly, we tested Kickresume, currently ranked third by users Traditionally, this traffic translates to approximately 180,000 to 360,000 users. Considering the high conversion and payment rates typical of these AI tools, it suggests a strong user base. In my personal experience, this tool offers the strongest AI experience among the three. Upon logging in, an AI assistant automatically starts, guiding you to input your information and use a template to create your resume. Even if you skip this step, you can always return by clicking the AI Resume button on the dashboard, creating a new resume. Alternatively, you can click the "Import from LinkedIn" button at the bottom left corner, allowing the AI to automatically import information from your LinkedIn account, subsequently creating a new resume based on this content. This process is more intelligent compared to the previous two tools. However, it was disappointing that after using it once, I was prompted to upgrade to a paid version for continued use. A unique feature of this tool is "Pyjama Jobs", prominently located in the top navigation bar. It’s a feature specifically aimed at freelancers or digital nomads like me, helping find remote work opportunities. Out of curiosity, I tried it. After loading the resume I created by importing my LinkedIn profile, the system indeed found many remote job opportunities for me. After clicking the "Matches" tab, the AI specifically filtered out four remote development jobs that best matched my skill set and level. The salary for these remote positions seemed quite satisfying, averaging over $100,000 annually! Overall, this tool has more distinctive features compared to the first two. The main differences are in two areas: firstly, its business logic doesn’t require you to upload a pre-written resume for optimization. Instead, it uses AI technology and relevant templates to directly help you generate an ATS-friendly resume. Secondly, it not only helps optimize resumes or LinkedIn accounts, but it also has a job database specifically for certain groups. After preparing the resume and cover letter, you can apply directly, making the process very smooth. Finally, we looked at its subscription price. Among the three most popular tools, its price is the most friendly. For a quarterly subscription, it costs only $39 for three months. Moreover, students or teachers can apply for free Premium access. Objectively speaking, this price is quite attractive. After comprehensive evaluation: it doesn’t have LinkedIn account optimization in its core features, but it has added attractive features for specific groups, earning it 1 star. Its business logic and processes are very clear, earning another star. Its cost-effectiveness is very competitive, earning another star. The usability of the free version gets 0.3 stars, and since we didn’t find any major flaws, it gets another star, making a total score of 4.3 stars. Analysis and Conclusion Although I only evaluated the three most popular AI tools currently used in job hunting or seeking job change opportunities, the results are quite representative. I believe we can draw the following conclusions: 1. Starting of AI Technology Application: AI technology has started to be applied in this field. Although it does not cover every aspect of job hunting or career development, it significantly improves our efficiency in certain aspects, giving us considerable advantages over competitors not using these tools. For instance, LinkedIn account enhancement, resume writing and optimization, bypassing the ATS system to increase interview opportunities, and accurately matching the best career opportunities based on the data in our resumes. 2. Tools Supplement, not Replace: While these tools do offer advantages in certain aspects, they cannot completely replace professional consultants or our efforts. For example, even if we successfully bypass the ATS system and get interview opportunities using these tools, aspects such as pre-interview company and position research, value proposition during the interview, and post-interview salary negotiation still require our personal skill development. 3. Subscription is Almost Necessary for Serious Users: The free usability of all tools is quite limited, mainly for trial purposes. However, I don’t see this as a major disadvantage of these tools. For those genuinely looking for a job or wanting to change jobs, investing a few dozen dollars per quarter is a worthwhile investment. After all, in most cases, the job hunting process can be completed within a quarter, rather than taking one to two years. 4. Tailored Tool Selection for Specific Groups: For certain groups of people, subscribing to the most popular tool might not be the best choice. The most popular tools are often designed for the broadest user base and might not necessarily suit you best. For instance, in my case, I would most likely choose Kickresume. Its price is the most attractive to me, and as a freelancer and remote worker, its features are exceptionally beneficial. That concludes this evaluation. Which AI tool do you prefer, or if you have any opinions on my analysis, please leave a comment. I look forward to exchanging ideas with everyone! Links to the tools used in the test: https://resumeworded.com/ https://www.jobscan.co/ https://www.kickresume.com/ Link to the resume used in the test: https://docs.google.com/document/d/1cl9PK3XsJzXBjwQ526rUXklz5L5EyUTOLbwwm5sy6dk/edit?usp=sharing Also published here.

People Are Competing for Your Dream Job With These AI Tools

The Promise and Potential of LLM in Crypto

Personal Recollections of Joan of Arc — Volume 1 by Mark Twain, is part of the HackerNoon Books Series. You can jump to any chapter in this book . Joan Makes Cowards Brave Victors here Joan Makes Cowards Brave Victors THE NEXT day Joan wanted to go against the enemy again, but it was the feast of the Ascension, and the holy council of bandit generals were too pious to be willing to profane it with bloodshed. But privately they profaned it with plottings, a sort of industry just in their line. They decided to do the only thing proper to do now in the new circumstances of the case—feign an attack on the most important bastille on the Orleans side, and then, if the English weakened the far more important fortresses on the other side of the river to come to its help, cross in force and capture those works. This would give them the bridge and free communication with the Sologne, which was French territory. They decided to keep this latter part of the program secret from Joan. Joan intruded and took them by surprise. She asked them what they were about and what they had resolved upon. They said they had resolved to attack the most important of the English bastilles on the Orleans side next morning—and there the spokesman stopped. Joan said: “Well, go on.” “There is nothing more. That is all.” “Am I to believe this? That is to say, am I to believe that you have lost your wits?” She turned to Dunois, and said, “Bastard, you have sense, answer me this: if this attack is made and the bastille taken, how much better off would we be than we are now?” The Bastard hesitated, and then began some rambling talk not quite germane to the question. Joan interrupted him and said: “That will not do, good Bastard, you have answered. Since the Bastard is not able to mention any advantage to be gained by taking that bastille and stopping there, it is not likely that any of you could better the matter. You waste much time here in inventing plans that lead to nothing, and making delays that are a damage. Are you concealing something from me? Bastard, this council has a general plan, I take it; without going into details, what is it?” “It is the same it was in the beginning, seven months ago—to get provisions for a long siege, then sit down and tire the English out.” “In the name of God! As if seven months was not enough, you want to provide for a year of it. Now ye shall drop these pusillanimous dreams—the English shall go in three days!” Several exclaimed: “Ah, General, General, be prudent!” “Be prudent and starve? Do ye call that war? I tell you this, if you do not already know it: The new circumstances have changed the face of matters. The true point of attack has shifted; it is on the other side of the river now. One must take the fortifications that command the bridge. The English know that if we are not fools and cowards we will try to do that. They are grateful for your piety in wasting this day. They will reinforce the bridge forts from this side to-night, knowing what ought to happen to-morrow. You have but lost a day and made our task harder, for we will cross and take the bridge forts. Bastard, tell me the truth—does not this council know that there is no other course for us than the one I am speaking of?” Dunois conceded that the council did know it to be the most desirable, but considered it impracticable; and he excused the council as well as he could by saying that inasmuch as nothing was really and rationally to be hoped for but a long continuance of the siege and wearying out of the English, they were naturally a little afraid of Joan’s impetuous notions. He said: “You see, we are sure that the waiting game is the best, whereas you would carry everything by storm.” “That I would!—and moreover that I will! You have my orders—here and now. We will move upon the forts of the south bank to-morrow at dawn.” “And carry them by storm?” “Yes, carry them by storm!” La Hire came clanking in, and heard the last remark. He cried out: “By my baton, that is the music I love to hear! Yes, that is the right time and the beautiful words, my General—we will carry them by storm!” He saluted in his large way and came up and shook Joan by the hand. Some member of the council was heard to say: “It follows, then, that we must begin with the bastille St. John, and that will give the English time to—” Joan turned and said: “Give yourselves no uneasiness about the bastille St. John. The English will know enough to retire from it and fall back on the bridge bastilles when they see us coming.” She added, with a touch of sarcasm, “Even a war-council would know enough to do that itself.” Then she took her leave. La Hire made this general remark to the council: “She is a child, and that is all ye seem to see. Keep to that superstition if you must, but you perceive that this child understands this complex game of war as well as any of you; and if you want my opinion without the trouble of asking for it, here you have it without ruffles or embroidery—by God, I think she can teach the best of you how to play it!” Joan had spoken truly; the sagacious English saw that the policy of the French had undergone a revolution; that the policy of paltering and dawdling was ended; that in place of taking blows, blows were ready to be struck now; therefore they made ready for the new state of things by transferring heavy reinforcements to the bastilles of the south bank from those of the north. The city learned the great news that once more in French history, after all these humiliating years, France was going to take the offensive; that France, so used to retreating, was going to advance; that France, so long accustomed to skulking, was going to face about and strike. The joy of the people passed all bounds. The city walls were black with them to see the army march out in the morning in that strange new position—its front, not its tail, toward an English camp. You shall imagine for yourselves what the excitement was like and how it expressed itself, when Joan rode out at the head of the host with her banner floating above her. We crossed the five in strong force, and a tedious long job it was, for the boats were small and not numerous. Our landing on the island of St. Aignan was not disputed. We threw a bridge of a few boats across the narrow channel thence to the south shore and took up our march in good order and unmolested; for although there was a fortress there—St. John—the English vacated and destroyed it and fell back on the bridge forts below as soon as our first boats were seen to leave the Orleans shore; which was what Joan had said would happen, when she was disputing with the council. We moved down the shore and Joan planted her standard before the bastille of the Augustins, the first of the formidable works that protected the end of the bridge. The trumpets sounded the assault, and two charges followed in handsome style; but we were too weak, as yet, for our main body was still lagging behind. Before we could gather for a third assault the garrison of St. Prive were seen coming up to reinforce the big bastille. They came on a run, and the Augustins sallied out, and both forces came against us with a rush, and sent our small army flying in a panic, and followed us, slashing and slaying, and shouting jeers and insults at us. Joan was doing her best to rally the men, but their wits were gone, their hearts were dominated for the moment by the old-time dread of the English. Joan’s temper flamed up, and she halted and commanded the trumpets to sound the advance. Then she wheeled about and cried out: “If there is but a dozen of you that are not cowards, it is enough—follow me!” Away she went, and after her a few dozen who had heard her words and been inspired by them. The pursuing force was astonished to see her sweeping down upon them with this handful of men, and it was their turn now to experience a grisly fright—surely this is a witch, this is a child of Satan! That was their thought—and without stopping to analyze the matter they turned and fled in a panic. Our flying squadrons heard the bugle and turned to look; and when they saw the Maid’s banner speeding in the other direction and the enemy scrambling ahead of it in disorder, their courage returned and they came scouring after us. La Hire heard it and hurried his force forward and caught up with us just as we were planting our banner again before the ramparts of the Augustins. We were strong enough now. We had a long and tough piece of work before us, but we carried it through before night, Joan keeping us hard at it, and she and La Hire saying we were able to take that big bastille, and must. The English fought like—well, they fought like the English; when that is said, there is no more to say. We made assault after assault, through the smoke and flame and the deafening cannon-blasts, and at last as the sun was sinking we carried the place with a rush, and planted our standard on its walls. The Augustins was ours. The Tourelles must be ours, too, if we would free the bridge and raise the siege. We had achieved one great undertaking, Joan was determined to accomplish the other. We must lie on our arms where we were, hold fast to what we had got, and be ready for business in the morning. So Joan was not minded to let the men be demoralized by pillage and riot and carousings; she had the Augustins burned, with all its stores in it, excepting the artillery and ammunition. Everybody was tired out with this long day’s hard work, and of course this was the case with Joan; still, she wanted to stay with the army before the Tourelles, to be ready for the assault in the morning. The chiefs argued with her, and at last persuaded her to go home and prepare for the great work by taking proper rest, and also by having a leech look to a wound which she had received in her foot. So we crossed with them and went home. Just as usual, we found the town in a fury of joy, all the bells clanging, everybody shouting, and several people drunk. We never went out or came in without furnishing good and sufficient reasons for one of these pleasant tempests, and so the tempest was always on hand. There had been a blank absence of reasons for this sort of upheavals for the past seven months, therefore the people too to the upheavals with all the more relish on that account. About HackerNoon Book Series: We bring you the most important technical, scientific, and insightful public domain books. This book is part of the public domain. Mark Twain (2004). Personal Recollections of Joan of Arc — Volume 1. Urbana, Illinois: Project Gutenberg. Retrieved October 2022 https://www.gutenberg.org/cache/epub/2874/pg2874-images.html This eBook is for the use of anyone anywhere at no cost and with almost no restrictions whatsoever. You may copy it, give it away or re-use it under the terms of the Project Gutenberg License included with this eBook or online at , located at . www.gutenberg.org https://www.gutenberg.org/policy/license.html

Joan Makes Cowards Brave Victors

Both Web3 and AI have taken the world by storm in recent years, disrupting traditional models and processes across industries. They bring immense potential for change, especially from the point of view of building community-oriented, value-driven systems. I recently spoke with Jet Liu, Product Director at Phoenix and VP of Product at Apex Technologies, who is working at the intersection of AI and Web3. With Jet, who has more than ten years of experience building enterprise-grade AI applications and data infrastructures, we discussed how a combination of AI, Web3, and innovations in computation technology can help shape the future. Hi Jet. It’s great to speak with you. Please tell us about your journey and how you got interested in AI. Hey, thanks for having me. Having worked in the marketing technology (Martech) space for many years, I eventually arrived at APEX Technologies, Shanghai, where I'm currently the VP of Product. I've helped them build an end-to-end product suite for enterprises needing AI-powered consumer data solutions. Over 400 large enterprises and multinationals have used the company's technology, including the likes of Starbucks, Maserati, PingAn Group, Walmart, McLaren, Michael Kors, Smart Automotive, and many more. While I've been interested in exploring AI deeply for several years, especially alongside other emerging technologies like blockchain and cryptography, the real chance and scope to do so arrived when I joined APEX. I haven't looked back ever since—I didn't need to. Deep down, I've always wanted to do this, and I am also a part of this exciting project, Phoenix. So it's all going great. What problems are you looking to solve with this latest project you mentioned? When it's about emerging tech, you have two options. One, you can sulk and whine about past glories, missing out on the future, clinging on to nostalgia—think of the 'AI will destroy humans' camp. Two, you can get your hands dirty building the future. That's what we're doing with Phoenix. It's an answer to the currently inaccessible AI computation landscape characterized by restrictively high costs and resource-intensive frameworks that most people (or even enterprises) can't use. AI's development has been highly centralized so far, dominated by capital-rich giants like OpenAI, Microsoft, and so on. High entry barriers are a key enabling factor here—something we aim to fix. So, overall, Phoenix is all about democratizing AI computation and making it widely accessible for innovation-driven enterprises. We're leveraging blockchain technology and cryptography to this end, fostering decentralization for a better future. Phoenix has an interesting organizational structure. It's not a typical DAO. Tell us more about this and why it matters. Sure. The Phoenix DAO is unique because it combines three industry leaders in AI data research and development: APEX Technologies, FLC, and Tensor Investment. Backed by NewMargin Ventures, Tencent, and Anker Investment, APEX Technologies is mainly responsible for building Phoenix's enterprise-grade Layer 1 blockchain. FLC is a decentralized AI research organization based in Hong Kong. Their specialization includes federated learning, multi-party computation, and decentralized edge computing. Both APEX and FLC are working on the Phoenix Computation Layer, which enables multi-party computation and AI on-demand decentralized compute. Besides this, FLC is also helping with Phoenix's blockchain integration and the upcoming AIGC NFT Portal dApp. Tensor, an AI-driven proprietary trading firm specializing in emerging asset classes, is the third pillar of Phoenix DAO. They're focusing on the development, modeling, and partnerships for AlphaNet. Plus, they're supporting the ecosystem with Deep Learning and Deep Reinforcement Learning tech and GPU computation resources. It should be clear why Phoenix DAO's structure matters. Something as powerful and broadly impactful as decentralized AI computation demands collaboration among the best of the best, and that's what we're trying to achieve. And we're widening the scope by involving a globally-distributed, engaged, and driven Web3 community. What are the main products offered by Phoenix, and how do they contribute to supporting Web3 communities and enterprises? Currently, we're building three main products for the Phoenix ecosystem. The first and most important one is the decentralized computation layer, of course. That's like the backbone for everything to come. It's a novel infrastructure combining decentralized AI and privacy-oriented multi-party computation (MPC). This'll help intelligent Web3-native applications scale easily and increase the speed/efficiency of deploying AI models. The two other products include AlphaNet and NYBL. AlphaNet is a robust AI-powered platform for high-frequency crypto traders and an integrated scaling infrastructure. NYBL is a metaverse by AIGC that'll utilize our computation layer's AI Node Network for scaling and GPU-based video/image processing. Besides these, Phoenix also has an L1 enterprise-grade blockchain and Oracle platform for max efficiency and scope. Altogether, we can enable decentralized deep learning, GPU resource scaling, codeless AI deployment, and multi-party computation across industries like health, retail, consumer IoT, financial services, supply chain, and Web3. Zooming out a little, what's the biggest advantage/strength of integrating AI with Web3? Thanks for this question. Naysayers spin up AI vs. Web3 debates, but I feel that's pretty naive. These two are the most complimentary technologies we've ever got to date. We're in the age of data, where everything from consumer behaviors to impressions on social media is quantified/quantifiable. So much so that it's now justified to say—data is power. But this also implies a greater need to break up silos and promote decentralization. While AI is the tool for humanity to leverage massive unstructured data sets, monetizing them in an equitable manner, Web3 unlocks infrastructure for decentralization. Web3 also introduces novel organizational structures—like DAOs—besides new asset classes and revenue streams. By efficiently adopting the AI-Web3 combinations, we can finally build the community-oriented and user-centric world that has been on the cards since the Internet's advent. That'll be a truly global world without arbitrary barriers, hierarchical domination, and resource misutilization. What are the challenges of adopting Web3 and AI? And where do you see them a few years down the line? The biggest challenge is misinformation and FUD. The noise-signal ratio is disturbingly high because Web3 and AI are still in the nascent stages of market development and technical design. People are ready to believe anything the self-proclaimed gurus say on Twitter or other social media platforms. There are technical limitations for us to overcome. No doubt. The scalability of L1 blockchains is one example. Then there's the question of AI's accuracy and overcoming bias. But none of these make the concerned technologies inherently bad or harmful. And surely AI isn't here to destroy us—let's get real. It's a powerful tool instead and can significantly boost our progress. Anyway, all this FUD will evaporate once we start seeing more groundbreaking innovations. It's happening already and will only take so much time. The Internet had its critics alright, and we know how things went for them—they became Internet users themselves because what else would they do? It'll be the same with Web3 and AI. Soon these emerging technologies will disrupt almost every aspect of our lives. And once we realize the benefits on a practical, day-to-day level, there'll be no turning back. It's not a question of if but when. The future is inevitable—how we embrace it is what matters.

Unraveling Decentralized AI Computation & Web3 with Jet Liu, Product Director at Phoenix

It was my usual day. I was swiping new interesting vacancies to see what is required for my and higher positions. Then I started using ChatGPT to generate a cover letter for me, so my applications will be more personalized, and my chances to have feedback will be higher, but I had to copy-paste more than three times, so I decided to automate this process. React, TypeScript, JavaScript, CSS, HTML Project's tech stack: Step 1: Obtain OpenAI API Key The first thing you need is to obtain the OpenAI API key, so you'll be able to make requests to the ChatGPT model. To do so, you need: Generate OpenAI API key here https://platform.openai.com/account/api-keys Set up a paid account here: . Optional https://platform.openai.com/account/billing/overview It might be required if your trial is expired. And don't forget to save the API key in a safe place, as it will be required in the next steps and during usage. Step 2: Use the WebExtension template I've developed a small WebExtension, which contains the most needed code to automate job applications. The source code can be found here: . I used as a target job board platform. If you're using djinni.co you can just go ahead and use the template extension as an app. https://github.com/mainarthur/open-djinni-ai.git https://djinni.co I'll show you how to adapt this template to other job boards. I'll use as a target in this tutorial. https://jobs.dou.ua/ Template's structure In the contentScript folder, you can find file, which will be injected into the target webpage where you can add DOM manipulations and other logic contentScript.ts In the popup folder, you can find files that will be used to render the extension's popup. In the static folder, you can update the extension's logo and manifest.json. Using manifest.json, you specify basic metadata about your extension, such as the name and version, and can also specify aspects of your extension's functionality (such as background scripts, content scripts, and browser actions). #1 #2 Step 3: Template modification and usage Dependencies: , , git node.js yarn So let's start with cloning the template repo: git clone https://github.com/mainarthur/open-djinni-ai.git open-dou-ai
cd open-dou-ai Enter fullscreen mode. Exit fullscreen mode You can change open-dou-ai to whatever name of the folder you want, or don't include it at all # Update remote URLYou can change the remote URL in case you want to push your changes to own git account and here is an easy command for it: git remote set-url origin git@github.com:yourUserName/YourRepoName.git Then we should install all dependencies: yarn install Enter fullscreen mode. Exit fullscreen mode and let's start analyzing the web page Let's take a look at vacancy URL It will help you find a pattern for the URL because you will need it for the manifest file: src/static/manifest.json "content_scripts": [
    {
      "matches": [
        "https://jobs.dou.ua/companies/*/vacancies/*"
      ],
      "js": [
        "contentScript.js"
      ]
    }
  ], Enter fullscreen mode. Exit fullscreen mode Now our contentScript will be executed only on dou.ua vacancy page Next, we will need an apply button. Locate it on the website. Then you should find the ID or class CSS selector for the element. The easiest way to do so is to use DevTools. and now we can update the content script with our new selector const allApplyButtons = document.querySelectorAll(
  "a#reply-btn-id"
); Enter fullscreen mode. Exit fullscreen mode When we click "apply," the text area for the cover letter shows up, then we repeat the same procedure as we did with apply button. Then we update contentScript again with this selector: setTimeout(async () => {
    const textarea = document.querySelector(
      "textarea#reply_descr"
    ) as HTMLTextAreaElement; Enter fullscreen mode. Exit fullscreen mode The extension also requires vacancy text for ChatGPT prompt generation. So you need to find a selector for vacancy text: const data = {
      model: "gpt-3.5-turbo",
      messages: [
        {
          role: "user",
          content:
            "Hello, my skills are " +
            (await get(SKILLS_LIST_KEY)) +
            ", my experience:" +
            (await get(EXPERIENCE_KEY)) +
            " generate me a personalized apply message for a vacancy: " +
            (
              document.querySelector(
                "div.vacancy-section"
              ) as HTMLDivElement
            ).innerText,
        },
      ],
    }; Enter fullscreen mode. Exit fullscreen mode And let's also don't forget to update the name of our new extension in popup.tsx return (
    <React.Fragment>
      <Nav>
        <Heading>OpenAI dou.ua Apply</Heading>
      </Nav>
      <Main> Enter fullscreen mode. Exit fullscreen mode and in file: manifest.json {
  "manifest_version": 3,
  "name": "Dou AI",
  "description": "Chrome extension for applying dou.ua jobs using OpenAI's ChatGPT",
  "version": "1.0.0",
  "action": { Enter fullscreen mode. Exit fullscreen mode And now we're ready to build and run the extension. To build your code, you can run this command: yarn build Enter fullscreen mode. Exit fullscreen mode and load the build in your browser Go to chrome://extensions/ Turn on developer mode Load the unpacked extension, and here you have it. Now we can move back to the vacancy page and refresh it. Open the popup and fill all data: We're ready to use it with any vacancy because now we have the 'Apply with ChatGPT' button. Let's click on it and wait for the result And we're finally done. The resulting code you can find here: . This project still has a lot of room for improvement: we can add animations, adapt it for multiple platforms at once, and include the ability to dynamically select CSS selectors. https://github.com/mainarthur/open-dou-ai Also published . here

Stories, News & Mentions about AIG

Aig’s Stories on HackerNoon

Aig's News & Mentions