When magazines like 'Scientific American' are run by ideologues producing biased dreck, it only makes it more difficult to defend the institution of science itself.

‘Scientific American’ Departing Editor Laura Helmuth Helped Degrade Science

“It’s a Budget to batter businesses, with big hikes to payroll costs coming from National Insurance increases and a higher minimum wage. Put together it means

"A budget to batter business owners" says AJ Bell's Laura Suter as she crunches the numbers on impact of NICs & minimum wage hikes - IFA Magazine

Laura Bratton is a reporter for Yahoo Finance.

Nvidia stock soars to fresh record as Wall Street stays bullish ahead of earnings

Nvidia stock eyes record high, looks to unseat Apple as world's most valuable company

Nvidia stock steadies with record high in sight as AI boom continues

Pro-Trump influencers including close Trump ally Laura Loomer were fooled by an AI-generated image of an alleged victim of Hurricane Helene.

Gullible Trump Cronies Losing Their Minds Over Fake AI Slop on Twitter

Nvidia, AMD stocks lead tech rally after Fed rate cut

Laura

laura

Many people have discovered ChatGPT’s competencies in code generation. But, how well does it actually perform, and are there alternatives? In this post, I take a look at some scenarios where AI can and the degree to which it can be useful. Finally, I do my best to evaluate its performance. help with coding Intro ChatGPT has been everywhere in the last few months. Reading and hearing about how people use it and ‘force’ it to make mistakes has been fun for a week or two, but it's getting tiring. For some reason, my LinkedIn feed still has at least one "Look how ChatGPT failed"-post every day, and I don't know what people are trying to prove with that. At the same time, I am perplexed by software developers claiming that they are using ChatGPT to do their work. Statements like "This week, I didn't write a single line of code and only used ChatGPT" even made it to podcasts where people discussed which jobs AI might replace in the future. For everyone who has seriously worked as a software developer, it is evident that AI cannot replace software developers soon. Yes, ChatGPT can generate code for a given problem description, but it is not good at solving problems, which is the main task of a software engineer. AI can support specific tasks, though, and I am looking at some examples in this post. AI Pair Programming Several services implement smart code completion, also called AI pair programming. The term is quite misleading and was probably invented by someone without much experience with pair programming. While pair programming is specifically beneficial when two people work together to discuss how something should be implemented, AI [pair programming](https://Pair programming - Wikipedia  Wikipedia https://en.wikipedia.org › wiki › Pair_programming) tools instead help with the next step. As soon as you decide what to implement and where to put the code, it suggests a solution. The two most popular ones are and GitHub Copilot is based on , a model for translating natural language to code based on GPT-3. According to a paper by OpenAI researchers, the model generated working code for given Python docstrings for more than 70% of the attempts. That's quite impressive but still also emphasizes that it requires basic coding skills (i.e., writing a docstring) to retrieve good results. Like ChatGPT, CodeX can also be used for free. GitHub Copilot Tabnine. OpenAI Codex on the OpenAI website I have used Copilot for private projects for quite a while and liked it a lot. In the beginning (when it was still in technical preview), it mainly copied lines from GitHub code. Since then, it has improved and evolved a lot. Especially for code that implements generic tasks (like calculations or conversions), there is a high chance that Copilot suggests a complete function based on a doc string. For example, when you write it would suggest the complete function, including a return value. def reverse(self, x:int) -> int... Copilot also works well for writing tests. Just add the name "test" to a given function name and it does the job. In contrast, Tabnine was a massive disappointment for me, though, and felt like a dumb auto-completion suggesting mostly useless or wrong code. While Copilot uses one large model, Tabnine is based on several smaller models that solve individual problems. This is probably also why it is more convenient to use Tabnine in local mode and train it with your own code (which is not possible with Copilot). Even with a lot of context and well-named variables, the suggestions I received from Tabnine were often far from what you would expect. Using local mode, it worked well in large repositories with a lot of similar code. But this is far too easy compared to the problems I would expect such a model to solve. It was not smart but felt more like a tool for automatic copy and pasting. There might be other use cases and circumstances where Tabnine works well (for example, in huge company code bases where everything is written in a similar style and language) but as someone who is working on many smaller projects, it just didn't work for me. Asking AI to write Code As mentioned in the intro, one of the primary use cases of ChatGPT seems to be code generation. Of course, that was already possible with CodeX, but the boom of ChatGPT made many more people aware of what is possible. Compared to the plugins mentioned before, this approach aims to translate textual problem descriptions into code instead of being integrated into an existing code base. Doing this task is challenging for a model as it involves understanding a problem and generating suitable, functional code that solves this problem. In the past, many models failed to do the former reliably, and even with more specialized models, it still is a considerable challenge that is not solved by ChatGPT. However, let's have a look at the code ChatGPT generates. Asking ChatGPT to generate a weather app I asked ChatGPT to write a Streamlit application that fetches weather data for Dortmund from a free API and displays it in a chart. This is what I got: import streamlit as st
import requests
import pandas as pd
import plotly.express as px

# Set up OpenWeatherMap API key
api_key = "YOUR_API_KEY_HERE"

# Define function to get weather data from OpenWeatherMap API
def get_weather_data():
    url = f"https://api.openweathermap.org/data/2.5/onecall?lat=51.5134&lon=7.4814&exclude=hourly,minutely&appid={api_key}&units=metric"
    response = requests.get(url)
    data = response.json()
    return data

# Get weather data
data = get_weather_data()

# Convert weather data to a Pandas DataFrame
df = pd.DataFrame(data["daily"])

# Format the date column
df["dt"] = pd.to_datetime(df["dt"], unit="s")
df["date"] = df["dt"].dt.date
df = df.set_index("date")

# Create a line chart using Plotly Express
chart = px.line(df, x=df.index, y="temp.day", title="Dortmund Weather Forecast")
chart.update_xaxes(title="Date")
chart.update_yaxes(title="Temperature (°C)")

# Display the chart in the Streamlit app
st.plotly_chart(chart) In general, that's not too bad. ChatGPT also gave some instructions on how to get the API key. Unfortunately, it didn't tell me I also have to install , but as I asked for a Streamlit app, it's OK to assume I know how to install libraries. plotly Nevertheless, there are two issues: The temperature value is not parsed correctly. It is saved as a dictionary and accessing it with will throw an error. A quick and easy fix would be to replace this with . I assume that's a good example of how the model has learned a lot of different ways how to handle the reply of the API but isn't successful in drawing a connection between parsing the data and accessing it for the chart again. y="temp.day" y=df['temp'].str['day'] The app only works if there is a valid API key and throws an error if there is none. The missing error handling is unfortunate as it displays error messages in the front end. However, it could quickly be fixed by either catching the error and displaying a more user-friendly error message (preferred) or implementing the API call so that the error is not displayed in the UI. Apart from these issues, the code looks fine and runs after fixing the first issue. Also, while the number of comments might be too ambitious in general, it's helpful to understand the generated code. Writing tests with ChatGPT Next, I asked ChatGPT to write a unit test for the application. import unittest
from app import get_weather_data

class TestWeatherApp(unittest.TestCase):

    def test_weather_data(self):
        # Call the function to get weather data
        data = get_weather_data()

        # Check that the data contains the expected keys
        self.assertIn("current", data)
        self.assertIn("hourly", data)
        self.assertIn("daily", data)

        # Check that the daily data contains the expected keys
        daily_data = data["daily"][0]
        self.assertIn("dt", daily_data)
        self.assertIn("temp", daily_data)
        self.assertIn("feels_like", daily_data)
        self.assertIn("pressure", daily_data)
        self.assertIn("humidity", daily_data)
        self.assertIn("dew_point", daily_data)
        self.assertIn("wind_speed", daily_data)
        self.assertIn("wind_deg", daily_data)
        self.assertIn("weather", daily_data)
        self.assertIn("clouds", daily_data)
        self.assertIn("pop", daily_data)
        self.assertIn("uvi", daily_data) This one is interesting. First of all, the test calls get_weather_data(), but the API key is defined outside the function. Therefore, it will run without an API key and will always fail. So the test is useless. Apart from that, this is not a unit test but an integration test. It tests if the (external) API returns the correct data. To implement a unit test, expected data should be mocked, so I asked ChatGPT to fix this. This is not a unit test but an integration test. Can you fix this? import unittest
from unittest.mock import patch
from app import get_weather_data

class TestWeatherApp(unittest.TestCase):

    @patch('requests.get')
    def test_get_weather_data(self, mock_get):
        # Set up test data
        expected_data = {'daily': [
            {'dt': 1645808400,
            'temp': {'day': 5.5, 'min': 2.15, 'max': 7.9, 'night': 3.4, 'eve': 7.9, 'morn': 2.15},
            'feels_like': {'day': 0.3, 'night': -3.04, 'eve': 4.06, 'morn': -4.01},
            'pressure': 1028, 'humidity': 51, 'dew_point': -2.57, 'wind_speed': 2.81, 'wind_deg': 194,
            'weather': [{'id': 801, 'main': 'Clouds', 'description': 'few clouds', 'icon': '02d'}],
            'clouds': 16, 'pop': 0, 'uvi': 2.21}],
            'timezone_offset': 3600}

        # Set up mock response from requests.get
        mock_response = type(
                'MockResponse',
                (object,),
                {'json': lambda self: expected_data}
            )
        mock_get.return_value = mock_response

        # Call the function to get weather data
        data = get_weather_data()

        # Check that the function returns the expected data
        self.assertEqual(data, expected_data) This looks much better, but still, I am missing testing for potential error cases, so I asked again: Shouldn't the unittest also check if the app crashes in case something unusual happens. For example, if the API is down? ChatGPT added this part to the code generated earlier: @patch('requests.get')
    def test_get_weather_data_error(self, mock_get):
        # Set up mock response from requests.get to raise an exception
        mock_get.side_effect = Exception("Connection error")

        # Call the function to get weather data
        with self.assertRaises(Exception):
            data = get_weather_data() This addition looks reasonable. However, as the original app does not handle any error cases, this test doesn't make much sense. Testing if the app doesn't crash when we know it will crash leads nowhere, so I stopped asking. Overall, the results presented by ChatGPT were OK. They certainly help with prototyping or giving a template to start from. However, the code still needs to be made production-ready. ChatGPT would certainly help with that, but improving the code step by step by asking the right questions is not an efficient way of working. Outlook Last year, Deepmind's AlphaCode took part in competitive programming contests and The code and some comments by a human competitor can be found . This has been quite a breakthrough as it was the first time an AI coding model proved problem-solving skills. It will be interesting to see how these models improve and find their way into software development. performed like a median human competitor. here Also published . here

The Current Reality and Future Possibilities of AI-Powered Coding

Discovering Web3 Gaming with Long Do and Laura Cullell

The tech workforce in the US is not growing at an optimal pace. The number of schools offering computer science is not enough which is impacting the access of computer science to young women and students from marginalized communities. The truth is that there are over in the US but only graduated into the tech workforce last year. 410,020 open computing jobs 71,226 computer science students This is because only in the US offer computer science which deeply impacts its access to young women and students from marginalized communities. 47% of public high schools To address this pressing issue, we have launched #MyStartInTech, a campaign dedicated to drawing attention to this cause and in turn, help widen access to computer science in schools. As part of this campaign, we present the #MyStartInTech interview series where some of the esteemed tech entrepreneurs and professionals in the industry share details about their journey in tech. With this, we hope to draw attention to the infinite opportunities that will lay open if young women and students are given the opportunity to study computer science. In this interview, Laura Fernández, CEO and Co-Founder, , the first campus for Data Science, Product Management, and UX/UI Design for women, by women - talks about how she got her start in the tech world. allWomen Laura is an expert in tech and innovative learning, as well as community building. She also dedicates time to mentoring young entrepreneurs and enjoys curating tech content for brands. Let’s find out what she has to say! What is the importance of technology to you now and how does it impact your life? I think we all see the importance of technology to our daily lives from picking up our phones in the morning to check the news and social media, contacting our friends, doing a guided workout or meditation through an app, or playing some new music while we get ready for work. Then, of course, there's all of the technology that helps us facilitate our work-life like team organization and collaboration tools. Moreover, there's all the technology that's built specifically for women like period tracking apps, female-based online communities, and more that support the female experience in meaningful ways. This technology is everywhere and it's only becoming more inclusive as we make the tech industry more diverse. How important is it to increase access to computer science and technology to underrepresented communities and young women today? We know that our world is becoming increasingly more digital; we feel the effects of it every day. But only the people who know how to use these technologies will be able to benefit from the prosperity of them. That means having a leg-up in terms of job opportunities across all sectors as technology increasingly penetrates every industry. It's also really important for underrepresented communities and young women to have access to technology and computer science now so that they can help create it in the future. If not, our algorithms will continue to reflect our biased society by being biased themselves. Why do you think access to computer literacy in school is important? Computer literacy from a young age is really important because it sets people up for success. If we allow more people to have access to computers and technology earlier on in life, we'll see this reflected back in our future technology. It will be more meaningful and intuitive, built by people who really understand how it works and its potential. Moreover, the more we give diverse groups of people this opportunity to become computer literate at a younger age, the brighter our technology will be in terms of helping more people in valuable ways. What do you think about .Tech Domains taking up this cause? We think this is a super meaningful cause, and we really applaud your mission and your work. What we say here at allWomen is that it all starts with culture. Our culture tells us that women, particularly women of color, don't build technology - men do. But if we start changing who has access to technology at a younger age, we believe this culture will change along with it! Learn more about #MyStartInTech campaign and lend your support at . https://startin.tech/

#MyStartInTech: An Interview With Laura Fernández, allWomen.Tech

Uni-OVSeg: A Step Towards Efficient and Bias-Resilient Vision Systems

Apple's big bet that consumers would splurge on artificial intelligence doesn't seem to be paying off. Earlier this month, the world's most valuable company announced the latest iteration of its iPhone, and the reaction was largely... mixed. Report after report on memes and reactions, and how the new iPhone was essentially just a slightly better version of what came before it. Investors weren't all that impressed either, with the company's stock barely reacting to the news. Apple's iPhone 16 announcement was being closely watched for a number of reasons, particularly because this was the phone the company was supposed to flex its AI muscles with. Despite being built from the "ground up" for artificial intelligence, the iPhone 16 failed to blow away consumers. Worse yet, Apple's competitors seemed to have beat the company to the punch. Hours after the iPhone 16 lineup was revealed, China's Huawei lifted the curtain on the Mate XT, a new tri-foldable smartphone that went on sale the same day as Apple's new device. Despite being twice the price of the iPhone 16 Pro Max, the Mate XT went on to rack up 3 million in pre-orders in China, with many walk-in customers unable to get their hands on the device on the day of the launch. So, what went wrong? Some in the industry view the new iPhone as a soft-launch of what will ultimately define Apple's AI push in the future. "With lots of words like "later this year" and "early next year", the core Apple message for iPhone 16 was: Next year will be better," Needham analyst Laura Martin said in a note, according to Reuters. Consumers really aren't ready to upgrade devices on a mere commitment that Apple will ultimately deliver on its promise of AI. As it stands, the company's Apple Intelligence is yet to launch, and there's some way to go. Compare that to Google, which announced AI-ready phones at its annual 'Made by Google' event that took place recently. Apple seems to have its work cut out for it, and I'll be keeping an eye out on whether the company's sales have increased following the launch of the iPhone 16 come next earnings season. Apple ranked #22 on HackerNoon's Tech Company Rankings this week. GET PUBLISHED ON HACKERNOON.
Join a global network of  45,000+ published devs, builders, founders, makers, VCs, hodlrs, and  hackers. Start submitting your tech stories and tutorials to get  published FREE on HackerNoon — no pop-ups, no paywalls. Switching gears, Sony announced the PlayStation 5 Pro, a beefier version of its current-gen console that comes at a whopping price of $700. Consumers seem to have taken exception to both the price, and the fact that the better console does not include a disc drive by default. I own a PS5 with a disc drive and personally think it's a waste of money since I don't buy games on discs. I wanted to get an all-digital version, but was unable to because it was out of stock, thus the reason why I own a version with a disc drive. Still, there might be folks out there who game using discs, and maybe that's a functionality that's important to them. Anyway, while the Ps5 Pro's design is not very different from the Ps5 (an extra stripe on the sides), Sony says its features include a GPU upgrade, advanced ray tracing, and AI-driven upscaling. Cool. All indicators point to the console being dead on arrival, specifically because of its high price. This is reminiscent of the PS3 days, when the console was panned for its high price tag, but Sony insists the new console will come with must-have bells and whistles that would greatly enhance the gaming experience. Whether it does or not, I'll be keeping an eye out. Sony ranked #42 on HackerNoon's Tech Company Rankings this week. In Other News.. 📰 Ex-Alameda Research CEO Caroline Ellison Sentenced to Two Years in Prison for Her Role in FTX Fraud — via CoinDesk
AI2’s Molmo shows open source can meet, and beat, closed multimodal models — via TechCrunch
TikTok purges accounts tied to Russian media over ‘covert influence’ efforts ahead of US election — via CNN
Meta readies AR glasses reveal at Connect event — via Reuters
Half of U.S. states seek to crack down on AI in elections — via Axios
How Apple and Microsoft’s trusted brands are being used to scam you online — via CNBC And that's a wrap! Don't forget to share this newsletter with your family and friends! See y'all next week. PEACE! ☮️ — Sheharyar Khan, Editor, Business Tech @ HackerNoon *All rankings are current as of Monday. To see how the rankings have changed, please visit HackerNoon's Tech Company Rankings page. Tech, What the Heck!? is a once-weekly newsletter written by HackerNoon editors that combine HackerNoon's proprietary data with news-worthy tech stories from around the internet. Humorous and insightful, the newsletter recaps trending events that are shaping the world of tech. Subscribe here. Apple 's big bet that consumers would splurge on artificial intelligence doesn't seem to be paying off. Apple Apple Earlier this month, the world's most valuable company announced the latest iteration of its iPhone, and the reaction was largely... mixed. Report after report on memes and reactions, and how the new iPhone was essentially just a slightly better version of what came before it. Investors weren't all that impressed either, with the company's stock barely reacting to the news. barely reacting barely reacting Apple's iPhone 16 announcement was being closely watched for a number of reasons, particularly because this was the phone the company was supposed to flex its AI muscles with. Despite being built from the "ground up" for artificial intelligence, the iPhone 16 failed to blow away consumers. this blow away blow away Worse yet, Apple's competitors seemed to have beat the company to the punch. Hours after the iPhone 16 lineup was revealed, China's Huawei lifted the curtain on the Mate XT, a new tri-foldable smartphone that went on sale the same day as Apple's new device. Despite being twice the price of the iPhone 16 Pro Max, the Mate XT went on to rack up 3 million in pre-orders in China, with many walk-in customers unable to get their hands on the device on the day of the launch . pre-orders pre-orders day of the launch day of the launch So, what went wrong? Some in the industry view the new iPhone as a soft-launch of what will ultimately define Apple's AI push in the future. "With lots of words like "later this year" and "early next year", the core Apple message for iPhone 16 was: Next year will be better," Needham analyst Laura Martin said in a note, according to Reuters. Consumers really aren't ready to upgrade devices on a mere commitment that Apple will ultimately deliver on its promise of AI. As it stands, the company's Apple Intelligence is yet to launch, and there's some way to go. Compare that to Google , which announced AI-ready phones at its annual 'Made by Google' event that took place recently. Google Apple seems to have its work cut out for it, and I'll be keeping an eye out on whether the company's sales have increased following the launch of the iPhone 16 come next earnings season. Apple ranked #22 on HackerNoon's Tech Company Rankings this week. Tech Company Rankings Tech Company Rankings GET PUBLISHED ON HACKERNOON. Join a global network of  45,000+ published devs, builders, founders, makers, VCs, hodlrs, and  hackers. Start submitting your tech stories and tutorials to get  published FREE on HackerNoon — no pop-ups, no paywalls. GET PUBLISHED ON HACKERNOON. GET PUBLISHED ON HACKERNOON . GET PUBLISHED ON HACKERNOON Join a global network of  45,000+ published devs, builders, founders, makers, VCs, hodlrs, and  hackers. Start submitting your tech stories and tutorials to get  published FREE on HackerNoon — no pop-ups, no paywalls. Join a global network of  45,000+ published devs, builders, founders, makers, VCs, hodlrs, and  hackers. Start submitting your tech stories and tutorials to get  published FREE on HackerNoon — no pop-ups, no paywalls. Switching gears, Sony announced the PlayStation 5 Pro, a beefier version of its current-gen console that comes at a whopping price of $700. Sony Consumers seem to have taken exception to both the price, and the fact that the better console does not include a disc drive by default. I own a PS5 with a disc drive and personally think it's a waste of money since I don't buy games on discs. I wanted to get an all-digital version, but was unable to because it was out of stock, thus the reason why I own a version with a disc drive. Still, there might be folks out there who game using discs, and maybe that's a functionality that's important to them. Anyway, while the Ps5 Pro's design is not very different from the Ps5 (an extra stripe on the sides), Sony says its features include a GPU upgrade, advanced ray tracing, and AI-driven upscaling. says says Cool . Cool All indicators point to the console being dead on arrival, specifically because of its high price. This is reminiscent of the PS3 days, when the console was panned for its high price tag, but Sony insists the new console will come with must-have bells and whistles that would greatly enhance the gaming experience. Whether it does or not, I'll be keeping an eye out. Sony ranked #42 on HackerNoon's Tech Company Rankings this week. In Other News.. 📰 In Other News.. 📰 Ex-Alameda Research CEO Caroline Ellison Sentenced to Two Years in Prison for Her Role in FTX Fraud — via CoinDesk AI2’s Molmo shows open source can meet, and beat, closed multimodal models — via TechCrunch TikTok purges accounts tied to Russian media over ‘covert influence’ efforts ahead of US election — via CNN Meta readies AR glasses reveal at Connect event — via Reuters Half of U.S. states seek to crack down on AI in elections — via Axios How Apple and Microsoft’s trusted brands are being used to scam you online — via CNBC Ex-Alameda Research CEO Caroline Ellison Sentenced to Two Years in Prison for Her Role in FTX Fraud — via CoinDesk CoinDesk CoinDesk AI2’s Molmo shows open source can meet, and beat, closed multimodal models — via TechCrunch TechCrunch TechCrunch TikTok purges accounts tied to Russian media over ‘covert influence’ efforts ahead of US election — via CNN CNN CNN Meta readies AR glasses reveal at Connect event — via Reuters Reuters Reuters Half of U.S. states seek to crack down on AI in elections — via Axios Axios Axios How Apple and Microsoft’s trusted brands are being used to scam you online — via CNBC CNBC CNBC And that's a wrap! Don't forget to share this newsletter with your family and friends! See y'all next week. PEACE! ☮️ — Sheharyar Khan , Editor, Business Tech @ HackerNoon Sheharyar Khan Sheharyar Khan *All rankings are current as of Monday. To see how the rankings have changed, please visit HackerNoon's Tech Company Rankings page. *All rankings are current as of Monday. To see how the rankings have changed, please visit HackerNoon's Tech Company Rankings page. *All rankings are current as of Monday. To see how the rankings have changed, please visit HackerNoon's Tech Company Rankings page. Tech Company Rankings Tech Company Rankings Tech, What the Heck!? is a once-weekly newsletter written by HackerNoon editors that combine HackerNoon's proprietary data with news-worthy tech stories from around the internet. Humorous and insightful, the newsletter recaps trending events that are shaping the world of tech. Subscribe here. Tech, What the Heck!? is a once-weekly newsletter written by HackerNoon editors that combine HackerNoon's proprietary data with news-worthy tech stories from around the internet. Humorous and insightful, the newsletter recaps trending events that are shaping the world of tech. Subscribe here. Tech, What the Heck!? is a once-weekly newsletter written by HackerNoon editors that combine HackerNoon's proprietary data with news-worthy tech stories from around the internet. Humorous and insightful, the newsletter recaps trending events that are shaping the world of tech. Subscribe here . here here

Stories, News & Mentions about LAURA

Laura’s Stories on HackerNoon

Laura's News & Mentions