IEEE Spectrum
Video Friday: Robot Dog Handstand
Video Friday is your weekly selection of awesome robotics videos, collected by your friends at IEEE...
a month ago
Video Friday is your weekly selection of awesome robotics videos, collected by your friends at IEEE Spectrum robotics. We also post a weekly calendar of upcoming robotics events for the next few months. Please send us your events for inclusion.
Humanoids 2024: 22–24 November...
Andrej Karpathy blog
Short Story on AI: Forward Pass
p {
text-align: justify;
}
.post pre, .post code {
border: none;
background-color:...
over a year ago
p {
text-align: justify;
}
.post pre, .post code {
border: none;
background-color: #eee;
}
The inspiration for this short story came to me while reading Kevin Lacker’s Giving GPT-3 a Turing Test. It is probably worth it (though not required) to skim this post to get...
Rozado’s Visual...
What is the IQ of ChatGPT?
Making an AI model take an IQ test
over a year ago
Making an AI model take an IQ test
Society's Backend
A Privacy Review: TikTok
You should stick to Instagram Reels and YouTube Shorts
a year ago
You should stick to Instagram Reels and YouTube Shorts
Society's Backend
Backend Biweekly #3: 97 Updates and Resources
Huge Nvidia Updates, Gemini Hackathon for Money, and more
8 months ago
Huge Nvidia Updates, Gemini Hackathon for Money, and more
Matt Mazur
Running Mistral 7B Instruct on a Macbook
Similar to yesterday’s post on running Mistral 8x7Bs Mixture of Experts (MOE) model, I wanted to...
a year ago
Similar to yesterday’s post on running Mistral 8x7Bs Mixture of Experts (MOE) model, I wanted to document the steps I took to run Mistral’s 7B-Instruct-v0.2 model on a Mac for anyone else interested in playing around with it. Unlike yesterday’s post though, this 7B Instruct...
Artificial Ignorance
The Industrial Content Revolution
A fundamental change in the structure the internet.
a year ago
A fundamental change in the structure the internet.
Sam Altman
Reinforcement Learning Progress
Today, OpenAI released a new result. We used PPO (Proximal Policy Optimization), a general...
over a year ago
Today, OpenAI released a new result. We used PPO (Proximal Policy Optimization), a general reinforcement learning algorithm invented by OpenAI, to train a team of 5 agents to play Dota and beat semi-pros.
This is the game that to me feels closest to the real world and complex...
Daniel Miessler
GPT and Search
There’s a lot of talk about how GPT is going to take over search. Meaning, compete with or take down...
over a year ago
There’s a lot of talk about how GPT is going to take over search. Meaning, compete with or take down Google. I get the excitement there, but there are some pretty serious barriers to having this happen immediately. First, GPT is non-deterministic, meaning you can ask it the same...
The Berkeley...
Designing Societally Beneficial Reinforcement Learning Systems
Deep reinforcement learning (DRL) is transitioning from a research field focused on game playing to...
over a year ago
Deep reinforcement learning (DRL) is transitioning from a research field focused on game playing to a technology with real-world applications. Notable examples include DeepMind’s work on controlling a nuclear reactor or on improving Youtube video compression, or Tesla attempting...
Artificial Ignorance
GPTs won't make you rich
But they'll make you more productive.
11 months ago
But they'll make you more productive.
One Useful Thing
It is starting to get strange.
Let's talk about ChatGPT with Code Interpreter & Microsoft Copilot
a year ago
Let's talk about ChatGPT with Code Interpreter & Microsoft Copilot
Andrej Karpathy blog
Biohacking Lite
Throughout my life I never paid too much attention to health, exercise, diet or nutrition. I knew...
over a year ago
Throughout my life I never paid too much attention to health, exercise, diet or nutrition. I knew that you’re supposed to get some exercise and eat vegetables or something, but it stopped at that (“mom said”-) level of abstraction. I also knew that I can probably get away with...
Sam Altman
GPT-4o
There are two things from our announcement today I wanted to highlight.
First, a key part of our...
7 months ago
There are two things from our announcement today I wanted to highlight.
First, a key part of our mission is to put very capable AI tools in the hands of people for free (or at a great price). I am very proud that we’ve made the best model in the world available for free in...
AI Snake Oil
Introducing the REFORMS checklist for ML-based science
ML-based science is in trouble. Clear reporting standards for researchers could help.
a year ago
ML-based science is in trouble. Clear reporting standards for researchers could help.
Made by Ollin
Mac Replacement Icons
A standardized, simplified, and beautiful icon set
over a year ago
A standardized, simplified, and beautiful icon set
Society's Backend
Meet Devansh: Lessons Learned While Building a Community of 193,000 Subscribers
Content creation and networking lessons from a wildly successful freelance tech writer
a year ago
Content creation and networking lessons from a wildly successful freelance tech writer
Society's Backend
The Metrics Machine Learning Engineers Care About That Modelers Don't
And a brief overview of TPUs in Google data centers
9 months ago
And a brief overview of TPUs in Google data centers
Andrej Karpathy blog
Deep Neural Nets: 33 years ago and 33 years from now
.post-header h1 {
font-size: 35px;
}
.post pre,
.post code {
background-color: #fcfcfc;
...
over a year ago
.post-header h1 {
font-size: 35px;
}
.post pre,
.post code {
background-color: #fcfcfc;
font-size: 13px; /* make code smaller for this post... */
}
The Yann LeCun et al. (1989) paper Backpropagation Applied to Handwritten Zip Code Recognition is I believe of some...
Rozado’s Visual...
Prevalence of Prejudice-Denoting Words in News Media Discourse
Summary of manuscript “Prevalence of Prejudice-Denoting Words in News Media Discourse: A...
over a year ago
Summary of manuscript “Prevalence of Prejudice-Denoting Words in News Media Discourse: A Chronological Analysis”
Artificial Ignorance
The SearchGPT Paradigm
What people (still) fundamentally misunderstand about AI search.
4 months ago
What people (still) fundamentally misunderstand about AI search.
Strange Loop Canon
No, LLMs are not "scheming"
3 days ago
IEEE Spectrum
Video Friday: Multiple MagicBots
Video Friday is your weekly selection of awesome robotics videos, collected by your friends at IEEE...
2 weeks ago
Video Friday is your weekly selection of awesome robotics videos, collected by your friends at IEEE Spectrum robotics. We also post a weekly calendar of upcoming robotics events for the next few months. Please send us your events for inclusion.
Humanoids Summit: 11–12 December...
Weighty Thoughts
The Fall of Data Moats
AI actually makes most data moats weaker
a year ago
AI actually makes most data moats weaker
Made by Ollin
Emojin
An infinite kaomoji generator
over a year ago
An infinite kaomoji generator
Matt Mazur
Going Full Time on My SaaS After 13 Years
In January 2010 I soft-launched launched Preceden, a web-based timeline maker tool, followed a few...
a year ago
In January 2010 I soft-launched launched Preceden, a web-based timeline maker tool, followed a few weeks later by a larger launch on HackerNews: Today – almost 13 years to the day since the initial launch – I’m going full time on it and I couldn’t be more excited. A brief history...
Society's Backend
Top 10 Machine Learning Resources and Updates 06/14/2024
A new AI employee, a four hour video on recreating GPT-2, meritocracy at Scale, and more
6 months ago
A new AI employee, a four hour video on recreating GPT-2, meritocracy at Scale, and more
One Useful Thing
The Best Available Human Standard
What are the imperatives of the upside?
a year ago
What are the imperatives of the upside?
Marcus on AI
Don’t Ride This Bike! Generative AI’s persistent trouble with compositionality and parts
When the text-to-image AI generation system DALL-E2 was released in April 2022, the two of us,...
a week ago
When the text-to-image AI generation system DALL-E2 was released in April 2022, the two of us, together with Scott Aaronson, ran some informal experiments to probe its abilities.
Strange Loop Canon
The age of access
the great polarisation theory, and the death of professions
a year ago
the great polarisation theory, and the death of professions
Matt Mazur
LearnGPT is now EmergentMind
Naming things is hard. Shortly after ChatGPT launched at the end of November, I decided to build a...
a year ago
Naming things is hard. Shortly after ChatGPT launched at the end of November, I decided to build a site to share ChatGPT examples and eventually be the home to educational content to help people learn more about GPT. To my surprise, there were a lot of relevant available .com...
The Berkeley...
TinyAgent: Function Calling at the Edge
The ability of LLMs to execute commands through plain language (e.g. English) has enabled agentic...
6 months ago
The ability of LLMs to execute commands through plain language (e.g. English) has enabled agentic systems that can complete a user query by orchestrating the right set of tools (e.g. ToolFormer, Gorilla). This, along with the recent multi-modal efforts such as the GPT-4o or...
Weighty Thoughts
Who’s Winning the AI War?
All of us, except the AI startups and VCs—unless a real war breaks out
11 months ago
All of us, except the AI startups and VCs—unless a real war breaks out
Stories by Andrej...
Virtual Reality: still not quite there, again.
over a year ago
Andrej Karpathy blog
A Survival Guide to a PhD
This guide is patterned after my “Doing well in your courses”, a post I wrote a long time ago on...
over a year ago
This guide is patterned after my “Doing well in your courses”, a post I wrote a long time ago on some of the tips/tricks I’ve developed during my undergrad. I’ve received nice comments about that guide, so in the same spirit, now that my PhD has come to an end I wanted to compile...
Rozado’s Visual...
Pessimism in News Media Headlines
In previous work, I documented the growing emotional negativity (anger, fear, sadness, etc) of...
a year ago
In previous work, I documented the growing emotional negativity (anger, fear, sadness, etc) of American news media headlines between the years 2000 and 2019. Here, I extend that work by examining the attitudinal tone (pessimism, optimism or neutrality
Rozado’s Visual...
Excess mortality during COVID pandemic, % of population vaccinated and Stringency Index of...
Please be mindful of potential confounds: population density, population behavior, virus virulence...
over a year ago
Please be mindful of potential confounds: population density, population behavior, virus virulence drift, herd immunity, etc.
Daniel Miessler
NO. 362 | Dependency Scanner, Citrix Attacks, AI Analysis…
SECURITY NEWS Google released an open-source scanner for vulnerabilities in project dependencies....
over a year ago
SECURITY NEWS Google released an open-source scanner for vulnerabilities in project dependencies. It's a front-end to the OSV database that links a dependency list to its vulnerabilities. MORE The latest updates for Apple software fixed a new zero-day that could be used to hack...
Frank’s Ramblings
My Experience Living and Working in China, Part I
In this four-part article, I’ll go over some of the lessons I learned living and doing business in...
over a year ago
In this four-part article, I’ll go over some of the lessons I learned living and doing business in China’s tech industry. During my time in China, I’ve led a team of 10+ engineers to develop a location-based IoT and sensing platform, co-founded an open-source project called...
The Gradient
Financial Market Applications of LLMs
The AI revolution drove frenzied investment in both private and public companies and captured the...
8 months ago
The AI revolution drove frenzied investment in both private and public companies and captured the public’s imagination in 2023. Transformational consumer products like ChatGPT are powered by Large Language Models (LLMs) that excel at modeling sequences of tokens that represent...
The Berkeley...
Are We Ready for Multi-Image Reasoning? Launching VHs: The Visual Haystacks Benchmark!
Humans excel at processing vast arrays of visual information, a skill that is crucial for achieving...
5 months ago
Humans excel at processing vast arrays of visual information, a skill that is crucial for achieving artificial general intelligence (AGI). Over the decades, AI researchers have developed Visual Question Answering (VQA) systems to interpret scenes within single images and answer...
Artificial Ignorance
AI Roundup 074: Amazon's Adept acquisition
July 5, 2024.
5 months ago
Artificial Ignorance
Who in their right mind would do an AI hardware startup
From megaFLOPS to mega flops.
8 months ago
From megaFLOPS to mega flops.
Artificial Ignorance
How to reverse engineer the Substack (or any!) web API
Building an unauthorized Substack client with Claude.
4 months ago
Building an unauthorized Substack client with Claude.
The Berkeley...
On the Stepwise Nature of <br> Self-Supervised Learning
Figure 1: stepwise behavior in self-supervised learning. When training common SSL algorithms, we...
a year ago
Figure 1: stepwise behavior in self-supervised learning. When training common SSL algorithms, we find that the loss descends in a stepwise fashion (top left) and the learned embeddings iteratively increase in dimensionality (bottom left). Direct visualization of embeddings...
Andrej Karpathy blog
A from-scratch tour of Bitcoin in Python
.wrap {
max-width: 900px;
}
p {
font-family: sans-serif;
font-size: 15px;
...
over a year ago
.wrap {
max-width: 900px;
}
p {
font-family: sans-serif;
font-size: 15px;
font-weight: 300;
overflow-wrap: break-word; /* allow wrapping of very very long strings, like txids */
}
.post pre,
.post code {
background-color: #fafafa;
font-size: 13px; /*...
Artificial Ignorance
How to leverage long form content with AI
Specific tools and tactics for authors, podcasters, and videographers.
6 months ago
Specific tools and tactics for authors, podcasters, and videographers.
Frank’s Ramblings
My Experience Living and Working in China, Part II: COVID Stories
In this four-part article, I’ll go over some of the lessons I learned living and doing business in...
over a year ago
In this four-part article, I’ll go over some of the lessons I learned living and doing business in China’s tech industry. During my time in China, I’ve led a team of 10+ engineers to develop a location-based IoT and sensing platform, co-founded an open-source project called...
fast.ai
AI Safety and the Age of Dislightenment
Model licensing & surveillance will likely be counterproductive by concentrating power in...
a year ago
Model licensing & surveillance will likely be counterproductive by concentrating power in unsustainable ways
Society's Backend
How to Know if Your Data is Being Used Maliciously
The information you should know about a company before gifting them your data
a year ago
The information you should know about a company before gifting them your data
Artificial Ignorance
AI Roundup 029: Hug all the faces
August 25, 2023
a year ago
fast.ai
AI Harms are Societal, Not Just Individual
In the west, our ideas of harm are largely anchored to an individual being harmed by a particular...
over a year ago
In the west, our ideas of harm are largely anchored to an individual being harmed by a particular action at a discrete moment in time. Yet the harms caused by algorithmic systems are often collective and communal.
Rozado’s Visual...
Northern Awokening: Social-justice and prejudice-signifying language in Canadian news media
I have recently published a report with Aaron Wudrick from the Macdonald-Laurier Institute about...
a year ago
I have recently published a report with Aaron Wudrick from the Macdonald-Laurier Institute about changes in the language that the news media in Canada use. I have documented previously how in American news media mentions of terms that signify distinct forms of prejudice have...
Strange Loop Canon
Strange Loop Canon Commencement Address
5 months ago
The Gradient
Mamba Explained
Is Attention all you need? Mamba, a novel AI model based on State Space Models (SSMs), emerges as a...
8 months ago
Is Attention all you need? Mamba, a novel AI model based on State Space Models (SSMs), emerges as a formidable alternative to the widely used Transformer models, addressing their inefficiency in processing long sequences.
One Useful Thing
Working with AI: Two paths to prompting
Don't overcomplicate things
a year ago
Don't overcomplicate things
Society's Backend
The Step-by-Step Guide to Becoming a Machine Learning Engineer
And other practical guides to understand machine learning
11 months ago
And other practical guides to understand machine learning
Society's Backend
LLM Research Recap of 2024, Fine-tuning LLM Judges, Amazon's Nova Models, Google's Genie 2, and More
Society's Backend Reading List 12-10-2024
a week ago
Society's Backend Reading List 12-10-2024
Matt Mazur
Indie Hacking Week 2 Recap – TimelineGPT Progress, LearnGPT Decision Time
For Martin Luther King weekend my family and I headed to Boone, North Carolina and stayed in a cabin...
a year ago
For Martin Luther King weekend my family and I headed to Boone, North Carolina and stayed in a cabin for a few days. We went snow tubing at Beech Mountain while we were there, one of the many ski resorts in the area. As is common on vacations, I spent way too much time thinking...
Rozado’s Visual...
Mentions of Prejudice in Academic Papers: A Declining Trend Amidst Ongoing DEI Growth?
Prejudice-denoting terms in academic research have recently decreased while some DEI-related terms...
3 months ago
Prejudice-denoting terms in academic research have recently decreased while some DEI-related terms continue to rise—what does this shift reveal?
Sam Altman
Please Fund More Science
Experts on the COVID-19 pandemic seem to think there are three ways out—that is, for life, health,...
over a year ago
Experts on the COVID-19 pandemic seem to think there are three ways out—that is, for life, health, and the economy to return roughly to normal.
Either we get a vaccine good enough that R0 for the world goes below 1, a good enough treatment that people no longer need to be...
Society's Backend
Why Rust Isn't Killing C++
And a consideration for choosing a language
8 months ago
And a consideration for choosing a language
Strange Loop Canon
When we become cogs
on motivation and technology
a month ago
on motivation and technology
Marcus on AI
Which CEO will be the last to see it?
What kept me going during the dark years of 2022 and 2023 when Generative AI was wildly overhyped...
2 weeks ago
What kept me going during the dark years of 2022 and 2023 when Generative AI was wildly overhyped and I was constantly ridiculed was the secure knowledge that the truth would eventually come out.
Weighty Thoughts
News Roundup: May 15, 2024
Open AI announcements, DeepSeek-v2, and TSMC Arizona
7 months ago
Open AI announcements, DeepSeek-v2, and TSMC Arizona
Rozado’s Visual...
Out-of-office Donald Trump still more prominent in news media content than the current U.S....
In 2019, I helped my colleague Musa al-Gharbi document the extraordinary prominence of Donald Trump...
a year ago
In 2019, I helped my colleague Musa al-Gharbi document the extraordinary prominence of Donald Trump in news media content (see here). I have recently updated that previous analysis. Briefly stated, no U.S. president in recent history has received a similar amount of media...
Society's Backend
You Are What You Eat: Digital Edition
Why who you follow is more important than who follows you
a year ago
Why who you follow is more important than who follows you
Artificial Ignorance
AI Roundup 034: Meta Connect
September 29, 2023
a year ago
Weighty Thoughts
The Mystical Q
OpenAI Q*, Primer on Reinforcement Learning, and Implications
a year ago
OpenAI Q*, Primer on Reinforcement Learning, and Implications
Artificial Ignorance
How a $2000/hour escort uses AI to automate sex work
Listen now | A conversation with Adelyn Moore, an independent escort and adult content creator.
4 months ago
Listen now | A conversation with Adelyn Moore, an independent escort and adult content creator.
Society's Backend
AI Reading List 1: Understand Transformers, Reflection-70B Update, and LLMs Still Cannot Reason
Society's Backend Reading List 10-07-2024
2 months ago
Society's Backend Reading List 10-07-2024
Marcus on AI
Sora still appears to have trouble with physics
Exactly as I warned in February
a week ago
Exactly as I warned in February
Society's Backend
The Unfortunate Truth Regarding AI Regulation
And the impact it'll have for decades to come
6 months ago
And the impact it'll have for decades to come
AI Snake Oil
Four more things we worked on in 2022
We had a busy 2022. Here are a few things we worked on but didn’t cover here.
a year ago
We had a busy 2022. Here are a few things we worked on but didn’t cover here.
Rozado’s Visual...
The political preferences of Grok
Elon Musk's response to ChatGPT
a year ago
Elon Musk's response to ChatGPT
Artificial Ignorance
10 of the most impactful AI stories of 2023
A quick look back on a very busy year in AI.
a year ago
A quick look back on a very busy year in AI.
One Useful Thing
The Homework Apocalypse
Fall is going to be very different this year. Educators need to be ready.
a year ago
Fall is going to be very different this year. Educators need to be ready.
Strange Loop Canon
We all live on a knife's edge, and it's fine
Or: How we weaponised serendipity
8 months ago
Or: How we weaponised serendipity
One Useful Thing
Superhuman: What can AI do in 30 minutes?
AI multiplies your efforts. I found out by how much...
a year ago
AI multiplies your efforts. I found out by how much...
Artificial Ignorance
AI Roundup 049: Down the Rabbit hole
January 12, 2023.
11 months ago
fast.ai
Can LLMs learn from a single example?
We’ve noticed an unusual training pattern in fine-tuning LLMs. At first we thought it’s a bug, but...
a year ago
We’ve noticed an unusual training pattern in fine-tuning LLMs. At first we thought it’s a bug, but now we think it shows LLMs can learn effectively from a single example.
Daniel Miessler
AI is About to Feel Like AGI, and You Need to Get Ready
I just wrote a piece similar to this last week, but this one drives the point home even more....
over a year ago
I just wrote a piece similar to this last week, but this one drives the point home even more. Basically, the current trajectory of AI, with all the art generation, the language models, etc., are about to become a whole lot more instruction and response based. What does that mean?...
One Useful Thing
A quick and sobering guide to cloning yourself
It took me a few minutes to create a fake me giving a fake lecture.
a year ago
It took me a few minutes to create a fake me giving a fake lecture.
AI Snake Oil
OpenAI’s policies hinder reproducible research on language models
LLMs have become privately-controlled research infrastructure
a year ago
LLMs have become privately-controlled research infrastructure
AI Snake Oil
People keep anthropomorphizing AI. Here’s why
Companies and journalists both contribute to the confusion
a year ago
Companies and journalists both contribute to the confusion
Strange Loop Canon
Evaluations are all we need
On analysing talent in LLMs
11 months ago
On analysing talent in LLMs
Strange Loop Canon
Progress
when benefits are worth tallying the human cost of technological leaps
a year ago
when benefits are worth tallying the human cost of technological leaps
Rozado’s Visual...
DepolarizingGPT
A Political Chatbot that Gives 3 Politically Diverse Answers to Every Prompt
a year ago
A Political Chatbot that Gives 3 Politically Diverse Answers to Every Prompt
Strange Loop Canon
Slouching Towards Utopia
by Brad DeLong
a year ago
AI Snake Oil
Quantifying ChatGPT’s gender bias
Benchmarks allow us to dig deeper into what causes biases and what can be done about it
a year ago
Benchmarks allow us to dig deeper into what causes biases and what can be done about it
IEEE Spectrum
SwitchBot S10 Review: “This Is the Future of Home Robots”
I’ve been reviewing robot vacuums for more than a decade, and robot mops for just as long. It’s been...
2 months ago
I’ve been reviewing robot vacuums for more than a decade, and robot mops for just as long. It’s been astonishing how the technology has evolved, from the original iRobot Roomba bouncing off of walls and furniture to robots that use lidar and vision to map your entire house and...
Society's Backend
Taking the AI to Consumers
The Google Pixel 8 Pro is a much bigger deal than you think
a year ago
The Google Pixel 8 Pro is a much bigger deal than you think
Made by Ollin
HintBot
Hinting icons automatically with a neural network
over a year ago
Hinting icons automatically with a neural network
Sam Altman
How To Invest In Startups
There is a lot of advice about how to be a good startup founder. But there isn’t very much about...
over a year ago
There is a lot of advice about how to be a good startup founder. But there isn’t very much about how to be a good startup investor.
Before going any further, I should point out that this is a particularly hard time to invest in startups—it’s easier right now to be a...
Artificial Ignorance
AI and the workplace
How employees and CEOs alike can plan for the future.
7 months ago
How employees and CEOs alike can plan for the future.
Made by Ollin
NVIDIA Internship (2017)
Notes on my internship at NVIDIA Redmond.
over a year ago
Notes on my internship at NVIDIA Redmond.
One Useful Thing
How to Get an AI to Lie to You in Three Simple Steps
I keep getting fooled by AI, and it seems like others are, too.
a year ago
I keep getting fooled by AI, and it seems like others are, too.
Artificial Ignorance
Tutorial: How to chat with your documents
A step-by-step guide to doing Q&A with your data, using LlamaIndex and OpenAI.
a year ago
A step-by-step guide to doing Q&A with your data, using LlamaIndex and OpenAI.
One Useful Thing
Everyone is above average
Is AI a Leveler, King Maker, or Escalator?
a year ago
Is AI a Leveler, King Maker, or Escalator?
Society's Backend
I Beat Newsletter Fatigue With AI
And why direct forms of communication will always be super valuable
10 months ago
And why direct forms of communication will always be super valuable
The Gradient
What's Missing From LLM Chatbots: A Sense of Purpose
LLM-based chatbots’ capabilities have been advancing every month. These improvements are mostly...
3 months ago
LLM-based chatbots’ capabilities have been advancing every month. These improvements are mostly measured by benchmarks like MMLU, HumanEval, and MATH (e.g. sonnet 3.5, gpt-4o). However, as these measures get more and more saturated, is user experience increasing in proportion to...
Daniel Miessler
News & Analysis | NO. 349
over a year ago
Matt Mazur
Is the ChatGPT API Refusing to Summarize Academic Papers? Not so fast.
Yesterday on X, I shared a post about some responses I was getting from the ChatGPT 3.5 API...
11 months ago
Yesterday on X, I shared a post about some responses I was getting from the ChatGPT 3.5 API indicating that it was refusing to summarize arXiv papers: There has been a lot of discussion recently about the perceived decrease in the quality of ChatGPT’s responses and seeing...
Marcus on AI
CONFIRMED: LLMs have indeed reached a point of diminishing returns
Science, sociology, and the likely financial collapse of the Generative AI bubble
a month ago
Science, sociology, and the likely financial collapse of the Generative AI bubble
Artificial Ignorance
AI Roundup 069: Project Greymatter
May 31, 2024.
6 months ago
Daniel Miessler
My Answer to the “ChatGPT Isn’t Really Creative” Argument
There’s a river of argument about ChatGPT that goes something like this: I know GPT can do some cool...
a year ago
There’s a river of argument about ChatGPT that goes something like this: I know GPT can do some cool stuff, but it’s really just a next-word-completion trick. It’s not doing real creativity like humans do. I get that argument, and it’s attractive to me as well, but I think it’s...
One Useful Thing
The future of education in a world of AI
A positive vision for the transformation to come
a year ago
A positive vision for the transformation to come
Artificial Ignorance
AI Roundup 026: AudioCraft
August 4, 2023
a year ago
Artificial Ignorance
Stop begging for JSON
How OpenAI's Structured Outputs makes building with AI much more reliable.
a week ago
How OpenAI's Structured Outputs makes building with AI much more reliable.
AI Snake Oil
Can AI automate computational reproducibility?
A new benchmark to measure the impact of AI on improving science
3 months ago
A new benchmark to measure the impact of AI on improving science
The Berkeley...
Rethinking Human-in-the-Loop for Artificial Augmented Intelligence
How do we build and evaluate an AI system for real-world applications? In most AI research, the...
over a year ago
How do we build and evaluate an AI system for real-world applications? In most AI research, the evaluation of AI methods involves a training-validation-testing process. The experiments usually stop when the models have good testing performance on the reported datasets because...
Society's Backend
I Enjoy Technical Interviews and You Can Too
A simple change in mindset that completely changed the way I interview
a year ago
A simple change in mindset that completely changed the way I interview
Sam Altman
DALL•E 2
Today we did a research launch of DALL•E 2, a new AI tool that can create and edit images from...
over a year ago
Today we did a research launch of DALL•E 2, a new AI tool that can create and edit images from natural language instructions.
Most importantly, we hope people love the tool and find it useful. For me, it’s the most delightful thing to play with we’ve created so far. I find it to...
AI Snake Oil
AI leaderboards are no longer useful. It's time to switch to Pareto curves.
What spending $2,000 can tell us about evaluating AI agents
7 months ago
What spending $2,000 can tell us about evaluating AI agents
Artificial Ignorance
AI Roundup 088: Movie magic
October 11, 2024.
2 months ago
One Useful Thing
Captain's log: the irreducible weirdness of prompting AIs
Also, we have a prompt library!
9 months ago
Also, we have a prompt library!
Artificial Ignorance
Small newsletters, big ideas
A subscriber showcase full of hidden gems on Substack.
6 days ago
A subscriber showcase full of hidden gems on Substack.
Daniel Miessler
How to Survive and Thrive in a World Where AI Can Do Almost Everything
Click for printable size. Here’s a quick list of things we can do to get ready for AI’s ascendance....
over a year ago
Click for printable size. Here’s a quick list of things we can do to get ready for AI’s ascendance. You can click it to get the full size to print out. This is UL Member Content Subscribe Already a member? Login
Daniel Miessler
NO. 358 | NEWS, ANALYSIS & DISCOVERY
🦃 We're doing our second-ever discount on UL Membership starting the day after Thanksgiving. But...
over a year ago
🦃 We're doing our second-ever discount on UL Membership starting the day after Thanksgiving. But that's a Friday, so I'm going to enable the discount link earlier. How early, and how much of a discount? You'll have to find out. If the link works before the date, then it's...
IEEE Spectrum
Video Friday: Mobile Robot Upgrades
Video Friday is your weekly selection of awesome robotics videos, collected by your friends at IEEE...
2 months ago
Video Friday is your weekly selection of awesome robotics videos, collected by your friends at IEEE Spectrum robotics. We also post a weekly calendar of upcoming robotics events for the next few months. Please send us your events for inclusion.
ROSCon 2024: 21–23 October 2024,...
Artificial Ignorance
AI Roundup 092: Watermarking the AI wave
November 8, 2024.
a month ago
Daniel Miessler
Unsupervised Learning NO. 365 | China’s Decline, MicrosoftAI, Creativity Ratio…
🎙️If you're not subscribed to the podcast version of the newsletter, please add it using with your...
a year ago
🎙️If you're not subscribed to the podcast version of the newsletter, please add it using with your favorite client! APPLE | SPOTIFY | OTHER SECURITY NEWS NYC Surveillance Amnesty International has revealed new research showing that the NYPD has over 15,000 cameras that can do...
Strange Loop Canon
Google had a very bad week
Gemini's paradox
9 months ago
Artificial Ignorance
AI Roundup 075: Levels of AGI
July 12, 2024.
5 months ago
Rozado’s Visual...
The Increasing Prominence of Prejudice and Social Justice Rhetoric in UK News Media
I have recently published a report with Matthew Goodwin about the increasing prominence of prejudice...
over a year ago
I have recently published a report with Matthew Goodwin about the increasing prominence of prejudice and social justice rhetoric in UK news media. Recent years have seen considerable debate about the rise of political polarization in British society. Specifically, over the last...
IEEE Spectrum
Video Friday: Reachy 2
IEEE Spectrum robotics. We also post a weekly calendar of upcoming robotics events for the next few...
2 months ago
IEEE Spectrum robotics. We also post a weekly calendar of upcoming robotics events for the next few months. Please send us your events for inclusion.
IROS 2024: 14–18 October 2024, ABU DHABI, UAE
ICSR 2024: 23–26 October 2024, ODENSE, DENMARK
Cybathlon 2024: 25–27 October 2024,...
fast.ai
The Jupyter+git problem is now solved
Previously, using git with Jupyter could create conflicts and break notebooks. With nbdev2, the...
over a year ago
Previously, using git with Jupyter could create conflicts and break notebooks. With nbdev2, the problem has been totally solved.
IEEE Spectrum
Unlock the Power of Agile for Hardware Teams
The Modified Agile for Hardware Development (MAHD) Framework is the ultimate solution for hardware...
6 days ago
The Modified Agile for Hardware Development (MAHD) Framework is the ultimate solution for hardware teams seeking the benefits of Agile without the pitfalls of applying software-centric methods. Traditional development approaches, like waterfall, often result in delayed timelines,...
One Useful Thing
I, Cyborg: Using Co-Intelligence
How I used AI in my book about AI
9 months ago
How I used AI in my book about AI
Artificial Ignorance
AI Roundup 035: Me or your lying eyes
October 6, 2023.
a year ago
The Berkeley...
Modeling Extremely Large Images with $x$T
As computer vision researchers, we believe that every pixel can tell a story. However, there seems...
9 months ago
As computer vision researchers, we believe that every pixel can tell a story. However, there seems to be a writer’s block settling into the field when it comes to dealing with large images. Large images are no longer rare—the cameras we carry in our pockets and those orbiting our...
Daniel Miessler
NO. 363 | FrontView Mirror: 2023 Edition
New Content 💡FRONTVIEW MIRROR: 2023 EDITION 💡MY PHILOSOPHY AND RECOMMENDATIONS ON THE LASTPASS...
a year ago
New Content 💡FRONTVIEW MIRROR: 2023 EDITION 💡MY PHILOSOPHY AND RECOMMENDATIONS ON THE LASTPASS BREACHES SECURITY NEWS In a bit of Deja Vu from LastPass, Okta has now revealed that attackers have stolen source code from its GitHub repositories. This comes after it was hit by...
Artificial Ignorance
10 AI predictions for 2024
Hey Siri, set a reminder for 365 days.
12 months ago
Hey Siri, set a reminder for 365 days.
Artificial Ignorance
AI Roundup 085: You win some, you Newsom
September 20, 2024.
3 months ago
Matt Mazur
The Kenya Quick Answer Goes Viral, Again
On Thursday evening Chris Ingraham, a journalist with 100k followers on Twitter, shared a screenshot...
a year ago
On Thursday evening Chris Ingraham, a journalist with 100k followers on Twitter, shared a screenshot of the now-famous “african country that starts with k” Google Quick Answer, which quickly went viral, garnering over 82k likes and 3 million views as of the time of this writing...
Artificial Ignorance
AI Roundup 084: Strawberry / o1
September 13, 2024.
3 months ago
AI Snake Oil
What is algorithmic amplification and why should we care?
A symposium and a primer on social media recommendation algorithms
a year ago
A symposium and a primer on social media recommendation algorithms
One Useful Thing
How to Use AI to Do Stuff: An Opinionated Guide
Covering the state of play as of Summer, 2023
a year ago
Covering the state of play as of Summer, 2023
Society's Backend
Things Everyone Should Understand About the Stanford AI Index Report
And my notes on why they’re important
8 months ago
And my notes on why they’re important
Marcus on AI
AI influencer hype gone wild
Reality is quite the opposite
a month ago
Reality is quite the opposite
Artificial Ignorance
AI Roundup 081: Creative differences
August 23, 2024.
4 months ago
Artificial Ignorance
AI Roundup 027: Challenge accepted
August 11, 2023
a year ago
AI Snake Oil
On the Societal Impact of Open Foundation Models
Adding precision to the debate on openness in AI
9 months ago
Adding precision to the debate on openness in AI
Artificial Ignorance
How to talk to your family about AI this Thanksgiving
A handy guide for your uncle's burning questions.
a year ago
A handy guide for your uncle's burning questions.
The Gradient
Car-GPT: Could LLMs finally make self-driving cars happen?
Exploring the utility of large language models in autonomous driving: Can they be trusted for...
9 months ago
Exploring the utility of large language models in autonomous driving: Can they be trusted for self-driving cars, and what are the key challenges?
Weighty Thoughts
News Roundup: May 27, 2014 🧴
Why does Google suck so much, Microsoft Co-Pilot Everywhere, and Sam Altman
6 months ago
Why does Google suck so much, Microsoft Co-Pilot Everywhere, and Sam Altman
Rozado’s Visual...
Artificial Intelligence and Portraits of 17th Century Physicists
The case for customizable AI systems as an alternative to one-size-fits-all AI systems
10 months ago
The case for customizable AI systems as an alternative to one-size-fits-all AI systems
Artificial Ignorance
States are racing ahead of Congress to regulate deepfakes
Several states just banned deepfakes in political ads and porn.
9 months ago
Several states just banned deepfakes in political ads and porn.
Rozado’s Visual...
El sesgo político de ChatGPT
A la izquierda del centro político
a year ago
A la izquierda del centro político
Society's Backend
Stop Obsessing Over the Product and Start Thinking About the Bigger Picture
The actual takeaways from the iPhone 15 event
a year ago
The actual takeaways from the iPhone 15 event
Society's Backend
Why MLX is Important for the ML Community
And a step-by-step guide to train a machine learning model on your Mac
7 months ago
And a step-by-step guide to train a machine learning model on your Mac
The Berkeley...
The Berkeley Crossword Solver
We recently published the Berkeley Crossword Solver (BCS), the current state of the art for solving...
over a year ago
We recently published the Berkeley Crossword Solver (BCS), the current state of the art for solving American-style crossword puzzles. The BCS combines neural question answering and probabilistic inference to achieve near-perfect performance on most American-style crossword...
Rozado’s Visual...
The ‘Great Awokening' Preceded Trump. It May Continue Beyond Him
Newsweek opinion piece Follow-up: Prevalence of Prejudice-Denoting Words in News Media Discourse...
over a year ago
Newsweek opinion piece Follow-up: Prevalence of Prejudice-Denoting Words in News Media Discourse Over the last five years, poll after poll has found that the GOP base has grown warmer towards Blacks, Hispanics, immigrants and Muslims. They’ve simultaneously become
Artificial Ignorance
AI Roundup 090: Computer Use
October 25, 2024.
a month ago
Matt Mazur
Experimenting with GPT-4 Turbo’s JSON mode
One of the many new features announced at yesterday’s OpenAI dev day is better support for...
a year ago
One of the many new features announced at yesterday’s OpenAI dev day is better support for generating valid JSON output. From the JSON mode docs: A common way to use Chat Completions is to instruct the model to always return JSON in some format that makes sense for your use case,...
Sam Altman
What I Heard From Trump Supporters
After the election, I decided to talk to 100 Trump voters from around
the country. I went to the...
over a year ago
After the election, I decided to talk to 100 Trump voters from around
the country. I went to the middle of the
country, the middle of the state, and talked to many online.
This was a surprisingly interesting and helpful experience—I highly
recommend it. With three exceptions,...
Matt Mazur
My Indie SaaS Revenue has Grown 37% per Year for 13 Years
Unlike many indie founders, I’ve never shared revenue numbers for Preceden, my SaaS timeline maker...
11 months ago
Unlike many indie founders, I’ve never shared revenue numbers for Preceden, my SaaS timeline maker tool. Even if they were remarkable – which they are not really – I just don’t think there are many good reasons to publicly share revenue numbers, and there are lots of downsides....
Artificial Ignorance
The subscriptionization of AI
Navigating the paid AI landscape.
9 months ago
Navigating the paid AI landscape.
IEEE Spectrum
Boston Dynamics’ Latest Vids Show Atlas Going Hands On
Boston Dynamics is the master of dropping amazing robot videos with no warning, and last week, we...
a month ago
Boston Dynamics is the master of dropping amazing robot videos with no warning, and last week, we got a surprise look at the new electric Atlas going “hands on” with a practical factory task.
This video is notable because it’s the first real look we’ve had at the new Atlas doing...
AI Snake Oil
Artists can now opt out of generative AI. It’s not enough.
Opting out is the latest example of generative AI developers externalizing costs.
a year ago
Opting out is the latest example of generative AI developers externalizing costs.
One Useful Thing
What just happened
A transformative month rewrites the capabilities of AI
3 days ago
A transformative month rewrites the capabilities of AI
The Gradient
An Introduction to the Problems of AI Consciousness
Once considered a forbidden topic in the AI community, discussions around the concept of AI...
a year ago
Once considered a forbidden topic in the AI community, discussions around the concept of AI consciousness are now taking center stage, marking a significant shift since the current AI resurgence began over a decade ago.
Weighty Thoughts
One Definite Sign of a Bad Startup Idea
If you can’t tell anyone about your startup idea, it's not a good idea
over a year ago
If you can’t tell anyone about your startup idea, it's not a good idea
One Useful Thing
The shape of the shadow of The Thing
We can start to see, dimly, what the near future of AI looks like.
a year ago
We can start to see, dimly, what the near future of AI looks like.
Society's Backend
AI and Software Reading List 4: State of the Job Market, Apple's Private Cloud Compute Released for...
Society's Backend Reading List 10-28-2024
a month ago
Society's Backend Reading List 10-28-2024
fast.ai
nbdev+Quarto: A new secret weapon for productivity
Our favorite tool for software engineering productivity–nbdev, now re-written with Quarto
over a year ago
Our favorite tool for software engineering productivity–nbdev, now re-written with Quarto
Weighty Thoughts
The AI Executive Order
AKA: “We’ll do something about it later”
a year ago
AKA: “We’ll do something about it later”
Matt Mazur
Screw it, I’m Keeping Emergent Mind
A few months ago I announced I was going to try to sell Emergent Mind, my AI news aggregator, so I...
a year ago
A few months ago I announced I was going to try to sell Emergent Mind, my AI news aggregator, so I could focus on Preceden, my SaaS timeline maker. I wound up having a lot of discussions with potential buyers, but in the end the offers I received were either too low to be worth...
Marcus on AI
o3, AGI, the art of the demo, and what you can expect in 2025
OpenAI’s new model was revealed yesterday; its most fervent believers think AGI has already arrived....
yesterday
OpenAI’s new model was revealed yesterday; its most fervent believers think AGI has already arrived. Here’s what you should pay attention to in the coming year.
Sam Altman
The Strength of Being Misunderstood
A founder recently asked me how to stop caring what other people think. I didn’t have an answer, and...
over a year ago
A founder recently asked me how to stop caring what other people think. I didn’t have an answer, and after reflecting on it more, I think it's the wrong question.
Almost everyone cares what someone thinks (though caring what everyone thinks is definitely a mistake), and it's...
One Useful Thing
Which AI should I use? Superpowers and the State of Play
And then there were three...
9 months ago
And then there were three...
One Useful Thing
Freeing the chatbot
Intelligence, of a sort, is going to be all around us
7 months ago
Intelligence, of a sort, is going to be all around us
One Useful Thing
In Praise of Boring AI
Automation has always been about killing tedious work. AI can do the same.
a year ago
Automation has always been about killing tedious work. AI can do the same.
The Berkeley...
How to Evaluate Jailbreak Methods: A Case Study with the StrongREJECT Benchmark
When we began studying jailbreak evaluations, we found a fascinating paper claiming that you could...
3 months ago
When we began studying jailbreak evaluations, we found a fascinating paper claiming that you could jailbreak frontier LLMs simply by translating forbidden prompts into obscure languages. Excited by this result, we attempted to reproduce it and found something unexpected.
The...
AI Snake Oil
Does the UK’s liver transplant matching algorithm systematically exclude younger patients?
Seemingly minor technical decisions can have life-or-death effects
a month ago
Seemingly minor technical decisions can have life-or-death effects
Daniel Miessler
AI Art Just Opened The Threat to Human Work We Were Expecting from AGI
Let me start with the punchline: Something like 80% of most “knowledge work” is about to get...
over a year ago
Let me start with the punchline: Something like 80% of most “knowledge work” is about to get replaced by artificial intelligence. I’m not professionally educated or trained in AI, but I’ve read probably 30 books and spent thousands of hours thinking about it. I am not talking...
Sam Altman
American Equity
I’d like feedback on the following idea.
I think that every adult US citizen should get an annual...
over a year ago
I’d like feedback on the following idea.
I think that every adult US citizen should get an annual share of the US GDP.
I believe that owning something like a share in America would align all of us in making the country as successful as possible—the better the country does, the...
Marcus on AI
Satya Nadella and the three stages of scientific truth
A textbook case in how good ideas are often initially throttled
4 weeks ago
A textbook case in how good ideas are often initially throttled
Artificial Ignorance
AI Roundup 048: The Robot Constitution
January 5, 2024.
11 months ago
Rozado’s Visual...
Which is the Wokest AI?
A Wokeness Ranking of LLMs
9 months ago
A Wokeness Ranking of LLMs
Andrej Karpathy blog
(started posting on Medium instead)
The current state of this blog (with the last post 2 years ago) makes it look like I’ve disappeared....
over a year ago
The current state of this blog (with the last post 2 years ago) makes it look like I’ve disappeared. I’ve certainly become less active on blogs since I’ve joined Tesla, but
whenever I do get a chance to post something I have recently been defaulting
to doing it on Medium because...
Society's Backend
The Attention Thief: Machine Learning as a Monetization Weapon
How your greatest asset is silently stolen by technology every day
a year ago
How your greatest asset is silently stolen by technology every day
Matt Mazur
Introducing Preceden’s new AI-Powered Timeline Generator
For the past few months I’ve been heads down building an AI-powered timeline generator tool for...
a year ago
For the past few months I’ve been heads down building an AI-powered timeline generator tool for Preceden, my SaaS timeline maker software: The tool – which is free to use and available on Preceden’s homepage – lets you type in a topic or detailed description of a timeline and it...
AI Snake Oil
Are open foundation models actually more risky than closed ones?
A policy brief on open foundation models
a year ago
A policy brief on open foundation models
Jascha’s blog
Brain dump on the diversity of AI risk
window.dataLayer = window.dataLayer || [];
function gtag(){dataLayer.push(arguments);}
...
a year ago
window.dataLayer = window.dataLayer || [];
function gtag(){dataLayer.push(arguments);}
gtag('js', new Date());
gtag('config', 'G-1XJMTJ5KCK');
.md h2 {
font-size: 20px;
}
AI has the power to change the world in both wonderful and terrible ways. We should try to...
Andrej Karpathy blog
What a Deep Neural Network thinks about your #selfie
Convolutional Neural Networks are great: they recognize things, places and people in your personal...
over a year ago
Convolutional Neural Networks are great: they recognize things, places and people in your personal photos, signs, people and lights in self-driving cars, crops, forests and traffic in aerial imagery, various anomalies in medical images and all kinds of other useful things. But...
Daniel Miessler
NO. 361 | GPT++, Apple Security, CISA Cuba…
SECURITY NEWS South Korean authorities are warning that North Koreans are disguising themselves and...
over a year ago
SECURITY NEWS South Korean authorities are warning that North Koreans are disguising themselves and getting jobs in South Korea. The saddest part is that it appears to be just another income generation scheme, meaning they use the salaries to fund the North Korean nuclear...
Weighty Thoughts
Today’s AI critics don’t understand the history of technology
But is AI different than other technologies?
9 months ago
But is AI different than other technologies?
One Useful Thing
Catastrophe / Eucatastrophe
We have more agency over the future of AI than we think.
a year ago
We have more agency over the future of AI than we think.
Artificial Ignorance
AI Roundup 086: Llama 3.2
September 27, 2024.
2 months ago
Society's Backend
No One Should Be GPU Poor
For everyone to have access to AGI, everyone must also have access to the compute to use it
7 months ago
For everyone to have access to AGI, everyone must also have access to the compute to use it
Weighty Thoughts
Startup Valuation Made Simple(r) - Working Backwards
Counterintuitively, start from how much money you need
over a year ago
Counterintuitively, start from how much money you need
Weighty Thoughts
Let's talk about AI power costs
It's important, but a lot of recent attention has been concern-trolling
4 months ago
It's important, but a lot of recent attention has been concern-trolling
Rozado’s Visual...
Define Wokeness! Or how you shall know a word by the company it keeps
Visualizing what words often appear in the vicinity of woke/wokeness in news media content...
a year ago
Visualizing what words often appear in the vicinity of woke/wokeness in news media content illustrates why communication is almost impossible between red and blue America
Artificial Ignorance
AI Roundup 044: Google Gemini
December 8, 2023.
a year ago
Artificial Ignorance
A stroll through Google's Model Garden
What generative AI capabilities does Google offer to developers?
a year ago
What generative AI capabilities does Google offer to developers?
IEEE Spectrum
This Inventor Is Molding Tomorrow’s Inventors
This article is part of our special report, “Reinventing Invention: Stories from Innovation’s...
a month ago
This article is part of our special report, “Reinventing Invention: Stories from Innovation’s Edge.”
Marina Umaschi Bers has long been at the forefront of technological innovation for kids. In the 2010s, while teaching at Tufts University, in Massachusetts, she codeveloped the...
Society's Backend
Run Your Own Race
What Bluey can teach us about machine learning
8 months ago
What Bluey can teach us about machine learning
Daniel Miessler
NO. 360 | NEWS, ANALYSIS & DISCOVERY SERIES
SECURITY NEWS Security researchers found that Chinese electronics company Eufy (part of Anker) has...
over a year ago
SECURITY NEWS Security researchers found that Chinese electronics company Eufy (part of Anker) has major vulnerabilities in its security cameras. The issues include uploading data to the cloud when they said they weren't, and the existence of a URL endpoint that allows an...
Artificial Ignorance
Looking a gift llama in the mouth
How Llama 3.1 uniquely leverages Meta's business model (and why we should be a little bit cynical...
5 months ago
How Llama 3.1 uniquely leverages Meta's business model (and why we should be a little bit cynical about it)
Daniel Miessler
What Made the 90’s So Awesome?
I just read a brilliant essay about the 90’s by Freddie de Boer, and it got me thinking. What made...
a year ago
I just read a brilliant essay about the 90’s by Freddie de Boer, and it got me thinking. What made the 90’s so great? Here’s GPT’s answer: Give a 90’s lover’s view of what made the 90’s awesome. Include everything from parenting, art, entertainment, games, childhood, movies, TV,...
Artificial Ignorance
AI Roundup 042: sama drama
November 24, 2023.
a year ago
Daniel Miessler
Your Experience is Your Creativity
Creativity is usually described as an external force that graces you with inspiration. Something...
a year ago
Creativity is usually described as an external force that graces you with inspiration. Something that you have to open yourself to—that you have to allow in. But creativity is more like an inner forge of your past, perspectives, and passions. It’s not something you let in; it’s...
AI Snake Oil
AI scaling myths
Scaling will run out. The question is when.
5 months ago
Scaling will run out. The question is when.
Society's Backend
Meta's New Segmentation Model, A New Open-Source Image Generation Model, Apple Intelligence Model...
Machine learning resources and updates 8/5/2024
4 months ago
Machine learning resources and updates 8/5/2024
Society's Backend
Should You Get a PhD?
The one good reason to do a PhD and two others that fall flat
2 months ago
The one good reason to do a PhD and two others that fall flat
Rozado’s Visual...
The Political Preferences of LLMs
Substantial political homogeneity in Large Language Models (LLMs) responses to questions with...
10 months ago
Substantial political homogeneity in Large Language Models (LLMs) responses to questions with political connotations
Weighty Thoughts
Scaling is a Choice
Progress in tech is rarely “inevitable.” Looking at semiconductors and AI.
3 days ago
Progress in tech is rarely “inevitable.” Looking at semiconductors and AI.
Artificial Ignorance
AI Roundup 050: Synthetic Geometry
January 19, 2024.
11 months ago
Society's Backend
Discussions around OpenAI's o1, Superhuman AI, When AI Should Be An App, and More
Discussions from the past week: 09/16/2024
3 months ago
Discussions from the past week: 09/16/2024
AI Snake Oil
AI companies are pivoting from creating gods to building products. Good.
Turning models into products runs into five challenges
4 months ago
Turning models into products runs into five challenges
One Useful Thing
The Present Future: AI's Impact Long Before Superintelligence
You can start to see the outlines of an AI future, for better and worse
a month ago
You can start to see the outlines of an AI future, for better and worse
PromptArmor Blog
Data exfiltration from Writer.com with indirect prompt injection
Authors: PromptArmor and Kai Greshake
a year ago
Authors: PromptArmor and Kai Greshake
Weighty Thoughts
Compute is Overrated as AI’s Bottleneck
You can’t just blindly extrapolate compute requirements
a year ago
You can’t just blindly extrapolate compute requirements
IEEE Spectrum
Video Friday: Extreme Off-Road
Video Friday is your weekly selection of awesome robotics videos, collected by your friends at IEEE...
a month ago
Video Friday is your weekly selection of awesome robotics videos, collected by your friends at IEEE Spectrum robotics. We also post a weekly calendar of upcoming robotics events for the next few months. Please send us your events for inclusion.
Humanoids 2024: 22–24 November...
The Gradient
Do text embeddings perfectly encode text?
'Vec2text' can serve as a solution for accurately reverting embeddings back into text, thus...
9 months ago
'Vec2text' can serve as a solution for accurately reverting embeddings back into text, thus highlighting the urgent need for revisiting security protocols around embedded data.
One Useful Thing
Post-apocalyptic education
What comes after the Homework Apocalypse
3 months ago
What comes after the Homework Apocalypse
Marcus on AI
PSA: Microsoft may be training on your private data without your knowledge
The Great Data Heist continues.
3 weeks ago
The Great Data Heist continues.
Artificial Ignorance
AI Roundup 083: Bioacoustics
September 6, 2024.
3 months ago
Daniel Miessler
News & Analysis | NO. 350
over a year ago
One Useful Thing
What people ask me most. Also, some answers.
A FAQ of sorts
a year ago
Artificial Ignorance
AI Roundup 047: NYT v. OpenAI
Friday, December 29.
11 months ago
Sam Altman
Greg
A lot of people ask me what the ideal cofounder looks
like. I now have an answer: Greg...
over a year ago
A lot of people ask me what the ideal cofounder looks
like. I now have an answer: Greg Brockman.
Every successful startup I know has at least one person who
provides the force of will to make the startup happen. I’d thought a lot
about this in the abstract while advising YC...
The Berkeley...
Virtual Personas for Language Models via an Anthology of Backstories
Anthology, a method for conditioning LLMs to representative, consistent, and diverse virtual...
a month ago
Anthology, a method for conditioning LLMs to representative, consistent, and diverse virtual personas by generating and utilizing naturalistic backstories with rich details of individual values and experience.
-->
We introduce Anthology, a method for conditioning LLMs to...
AI Snake Oil
AI safety is not a model property
Trying to make an AI model that can’t be misused is like trying to make a computer that can’t be...
9 months ago
Trying to make an AI model that can’t be misused is like trying to make a computer that can’t be used for bad things
Society's Backend
Backend Biweekly #2: 112 Total Updates and Resources
Details on Metas training clusters, analysis of ML competitions in 2023, AI's impact on water...
9 months ago
Details on Metas training clusters, analysis of ML competitions in 2023, AI's impact on water resources, and more
Society's Backend
1000+ AI Agents Make a Civilization, xAI Creates the Largest Supercomputer in 4 Months, Possible AI...
Machine learning resources and updates 2024-09-09
3 months ago
Machine learning resources and updates 2024-09-09
The Berkeley...
Goal Representations for Instruction Following
Goal Representations for Instruction Following
Figure title. Figure caption. This image is...
a year ago
Goal Representations for Instruction Following
Figure title. Figure caption. This image is centered and set to 50%
page width. -->
A longstanding goal of the field of robot learning has been to create generalist agents that can perform tasks for humans. Natural language has...
The Gradient
Text-to-CAD: Risks and Opportunities
In the realm of AI-powered text-to-CAD, there's promise, but also a surge in subpar designs. Can we...
a year ago
In the realm of AI-powered text-to-CAD, there's promise, but also a surge in subpar designs. Can we steer this technology towards better outcomes?
Jascha’s blog
The hot mess theory of AI misalignment: More intelligent agents behave less coherently
window.dataLayer = window.dataLayer || [];
function gtag(){dataLayer.push(arguments);}
...
a year ago
window.dataLayer = window.dataLayer || [];
function gtag(){dataLayer.push(arguments);}
gtag('js', new Date());
gtag('config', 'G-1XJMTJ5KCK');
.md h2 {
font-size: 20px;
}
Many machine learning researchers worry about risks from building artificial intelligence...
Society's Backend
Pelosi opposes SB 1047, New LLM Training Paradigms, Prompt Caching to Save 90% of API Costs, and...
Machine learning resources and updates 8/19/2024
4 months ago
Machine learning resources and updates 8/19/2024
Artificial Ignorance
AI Roundup 055: Race conditions
February 23, 2023.
10 months ago
Society's Backend
MatMul-free Machine Learning Can Improve Efficiency, Excellent Foundational Learning Resources, SSMs...
The top 10 most important updates of last week: 7/8/24
5 months ago
The top 10 most important updates of last week: 7/8/24
Daniel Miessler
UL NO. 354 | THE NEWS & ANALYSIS SERIES
SECURITY NEWS The US has implemented a number of aggressive export controls to stop China from...
over a year ago
SECURITY NEWS The US has implemented a number of aggressive export controls to stop China from attaining advanced semiconductors. And now it looks like the bans will be expanded to quantum computing and AI as well. NYTIMES | MY ANALYSIS BELOW It appears Bytedance had a plan to...
Society's Backend
Anti-Social Media
How our technology habituates the paradoxical relationship between social media and social isolation
a year ago
How our technology habituates the paradoxical relationship between social media and social isolation
Marcus on AI
An AI rumor you won’t want to miss
Converging evidence that the core hypothesis driving generative AI may be wrong
a month ago
Converging evidence that the core hypothesis driving generative AI may be wrong
One Useful Thing
What happens when AI reads a book 🤖📖
And some prompts that might be useful when it does.
a year ago
And some prompts that might be useful when it does.
Weighty Thoughts
What I got Right or Wrong in 2020
Reflecting on the Past Year
over a year ago
Reflecting on the Past Year
IEEE Spectrum
Drones with Legs Can Walk, Hop, and Jump into the Air
“Whenever I encountered crows on the EPFL campus, I would observe how they walked, hopped over or...
2 weeks ago
“Whenever I encountered crows on the EPFL campus, I would observe how they walked, hopped over or jumped on obstacles, and jumped for take-offs,” says Won Dong Shin, a doctoral student at EPFL’s Laboratory of Intelligent Systems. “What I consistently observed was that they always...
One Useful Thing
What Apple's AI Tells Us: Experimental Models⁴
Siri versus the machine god?
6 months ago
Siri versus the machine god?
IEEE Spectrum
Fish-Inspired Sensor "Touches" Using Electric Fields
The ability to detect a nearby presence without seeing or touching it may sound fantastical—but it’s...
a week ago
The ability to detect a nearby presence without seeing or touching it may sound fantastical—but it’s a real ability that some creatures have. A family of African fish known as Mormyrids are weakly electric, and have special organs that can locate a nearby prey, whether it’s in...
Artificial Ignorance
The AI research tool that saves me hours every week
And why it might revolutionize the search industry.
11 months ago
And why it might revolutionize the search industry.
Strange Loop Canon
Disruption starts at the margins, but doesn't stop there
Contra Hoel on AI and its supply paradox
a year ago
Contra Hoel on AI and its supply paradox
Matt Mazur
AOL Underground Podcast Interview about AOL-Files.com
Back in 1998 when I was 13 years old I got heavily involved in the AOL hacking scene, originally...
over a year ago
Back in 1998 when I was 13 years old I got heavily involved in the AOL hacking scene, originally building add-on software called progs (Revolution, Meridian), publishing code libraries called bas files (Alpha32), and later co-founding AOL-Files.com (where I went by the hacker...
Artificial Ignorance
AI Zoology: Why are there so many animal models?
Inside the emerging model menagerie.
a year ago
Inside the emerging model menagerie.
Society's Backend
AI Reading List 3: Meta's Self-Taught Evaluator, AI for Cancer Diagnosis, and a Technical...
Society's Backend Reading List 10-21-2024
2 months ago
Society's Backend Reading List 10-21-2024
Marcus on AI
Update re: Microsoft and training data
Relevant to the post I sent earlier today,...
3 weeks ago
Relevant to the post I sent earlier today, https://www.howtogeek.com/is-microsoft-using-your-word-documents-to-train-ai/says that an unnamed spokesperson at Microsoft claims that “Microsoft does not use customer data from Microsoft 365 consumer and commercial applications to...
Artificial Ignorance
Getting to the top of the GPT Store (and building an AI-native search engine, too)
Listen now | A conversation with Christian Salem, founder and CPO of Consensus.
8 months ago
Listen now | A conversation with Christian Salem, founder and CPO of Consensus.
Daniel Miessler
NO. 366 | T-Breach, Siri++, Conception Ages…
🎙️If you're not subscribed to the podcast version of the newsletter, please add it using with your...
a year ago
🎙️If you're not subscribed to the podcast version of the newsletter, please add it using with your favorite client! APPLE | SPOTIFY | OTHER SECURITY NEWS Another T-Mobile Breach T-Mobile has had another security breach, this one affecting at least 37 million accounts. They...
Daniel Miessler
OpenAI’s Purpose is to Build AGI, and What That Means
Sam Altman, the CEO of OpenAI, has said multiple times that, He says it in this video as well. >...
a year ago
Sam Altman, the CEO of OpenAI, has said multiple times that, He says it in this video as well. > We’re very much here to build AGI. Sam Altman I am not sure how many people realize this about the company. They’re not like playing with other AI-related tech and AGI might come out...
Weighty Thoughts
Commentary on Technology, Startups, and Investing
Welcome to Weighty Thoughts by me, James Wang. General Partner at Creative Ventures. Ex-Bridgewater...
over a year ago
Welcome to Weighty Thoughts by me, James Wang. General Partner at Creative Ventures. Ex-Bridgewater and Google X. Co-Founder Lioness. Sign up now so you don’t miss the first issue. In the meantime, tell your friends!
Artificial Ignorance
AI Roundup 078: Voice mode
August 2, 2024.
4 months ago
fast.ai
From Deep Learning Foundations to Stable Diffusion
Practical Deep Learning for Coders part 2, 2022
over a year ago
Practical Deep Learning for Coders part 2, 2022
Daniel Miessler
Napkin Ideas Around What Changes to Expect Post-ChatGPT
Work Replacement Talent Magnification Solopreneuers AI Specialists Idea Dominance Use Cases Random...
over a year ago
Work Replacement Talent Magnification Solopreneuers AI Specialists Idea Dominance Use Cases Random Thoughts If you’re reading this you already know the internet is on fire over the new GPTChatBot from OpenAI. There are people using it to create full virtual machines, to be their...
Andrej Karpathy blog
Short Story on AI: A Cognitive Discontinuity.
p {
text-align: justify;
}
The idea of writing a collection of short stories has been on my mind...
over a year ago
p {
text-align: justify;
}
The idea of writing a collection of short stories has been on my mind for a while. This post is my first ever half-serious attempt at a story, and what better way to kick things off than with a story on AI and what that might look like if you...
Society's Backend
Coming soon!
Weekly publications to teach you about the technology that governs our society and how it impacts...
a year ago
Weekly publications to teach you about the technology that governs our society and how it impacts you. Coming in August!
Artificial Ignorance
GPT-4o and the illusion of AGI
Why speed and multimodality is becoming the name of the game.
7 months ago
Why speed and multimodality is becoming the name of the game.
Society's Backend
Positioning Myself: The Greatest Piece of Career Advice I've Ever Received
And how it changed my personal life too
12 months ago
And how it changed my personal life too
Strange Loop Canon
Discovery is the original sin of the modern age
technology, living post scarcity, and how to find what you want
a year ago
technology, living post scarcity, and how to find what you want
AI Snake Oil
Starting reading the AI Snake Oil book online today
The book will be published on September 24
3 months ago
The book will be published on September 24
The Berkeley...
Ghostbuster: Detecting Text Ghostwritten by Large Language Models
The structure of Ghostbuster, our new state-of-the-art method for detecting AI-generated...
a year ago
The structure of Ghostbuster, our new state-of-the-art method for detecting AI-generated text.
Large language models like ChatGPT write impressively well—so well, in fact, that they’ve become a problem. Students have begun using these models to ghostwrite assignments, leading...
One Useful Thing
On holding back the strange AI tide
There is no way to stop the disruption. We need to channel it instead
a year ago
There is no way to stop the disruption. We need to channel it instead
Matt Mazur
Full Time Indie Hacking: First 3 Months in Review
At the end of 2022 I stopped contracting at Help Scout to focus full time on Preceden, a SaaS...
a year ago
At the end of 2022 I stopped contracting at Help Scout to focus full time on Preceden, a SaaS timeline maker tool that I had been running mostly as a side project since 2010. It’s April now so I figured I’d share an update on how things are going. My periodic Friday updates cover...
AI Snake Oil
A misleading open letter about sci-fi AI dangers ignores the real risks
Misinformation, labor impact, and safety are all risks. But not in the way the letter implies.
a year ago
Misinformation, labor impact, and safety are all risks. But not in the way the letter implies.
Marcus on AI
𝗼𝟯 “𝗔𝗥𝗖 𝗔𝗚𝗜” 𝗽𝗼𝘀𝘁𝗺𝗼𝗿𝘁𝗲𝗺 𝗺𝗲𝗴𝗮𝘁𝗵𝗿𝗲𝗮𝗱: 𝘄𝗵𝘆 𝘁𝗵𝗶𝗻𝗴𝘀 𝗴𝗼𝘁...
Kevin Roose, of Hard Fork and NYT, was so impressed with OpenAI’s rollout that he joked “of course...
4 hours ago
Kevin Roose, of Hard Fork and NYT, was so impressed with OpenAI’s rollout that he joked “of course they have to announce AGI the day my vacation starts”.
The Berkeley...
Linguistic Bias in ChatGPT: Language Models Reinforce Dialect Discrimination
Sample language model responses to different varieties of English and native speaker...
3 months ago
Sample language model responses to different varieties of English and native speaker reactions.
ChatGPT does amazingly well at communicating with people in English. But whose English?
Only 15% of ChatGPT users are from the US, where Standard American English is the default. But...
Marcus on AI
8 signs that Donald Trump has a progressive form of dementia
And how our media is failing, wildly, in one of the most important stories in US history
a month ago
And how our media is failing, wildly, in one of the most important stories in US history
fast.ai
GPT 4 and the Uncharted Territories of Language
Language is a source of limitation and liberation. GPT 4 pushes this idea to the extreme by giving...
a year ago
Language is a source of limitation and liberation. GPT 4 pushes this idea to the extreme by giving us access to unlimited language.
Matt Mazur
When LTD Purchasers Meet an Inactive User Policy
Last year I participated in a Lifetime Deal (LTD) promotion to offer Preceden to the AppSumo...
a year ago
Last year I participated in a Lifetime Deal (LTD) promotion to offer Preceden to the AppSumo community. Maybe I’ll dive into my experience there in another post, but I wanted to share an interesting thing that’s happening now, a year after the deal ended. AppSumo has a policy...
Made by Ollin
A Video History of Self-Driving Cars
over a year ago
Sam Altman
Join the YC Software Team
If you want to get funded by YC as a founder in the future, but you don't have a startup that's...
over a year ago
If you want to get funded by YC as a founder in the future, but you don't have a startup that's ready for that yet, joining the YC software team is a great hack to get there.
The YC software team is a small group of hackers in SF that write the software that makes all the parts...
The Berkeley...
Fully Autonomous Real-World Reinforcement Learning with Applications to Mobile Manipulation
Reinforcement learning provides a conceptual framework for autonomous agents to learn from...
a year ago
Reinforcement learning provides a conceptual framework for autonomous agents to learn from experience, analogously to how one might train a pet with treats. But practical applications of reinforcement learning are often far from natural: instead of using RL to learn through trial...
Sam Altman
The Merge
A popular topic in Silicon Valley is talking about what year humans and machines will merge (or, if...
over a year ago
A popular topic in Silicon Valley is talking about what year humans and machines will merge (or, if not, what year humans will get surpassed by rapidly improving AI or a genetically enhanced species). Most guesses seem to be between 2025 and 2075.
People used to call this the...
AI Snake Oil
The bait and switch behind AI risk prediction tools
Toronto recently used an AI tool to predict when a public beach will be safe. It went horribly awry....
over a year ago
Toronto recently used an AI tool to predict when a public beach will be safe. It went horribly awry. The developer claimed the tool achieved over 90% accuracy in predicting when beaches would be safe to swim in. But the tool did much worse: on a majority of the days when the...
Society's Backend
All Machine Learning Resources and Updates 7/8/24
Everything I've been reading
5 months ago
Everything I've been reading
Artificial Ignorance
AI Roundup 032: The Insight Forum
September 15, 2023.
a year ago
The Gradient
We Need Positive Visions for AI Grounded in Wellbeing
Introduction
Imagine yourself a decade ago, jumping directly into the present shock of conversing...
4 months ago
Introduction
Imagine yourself a decade ago, jumping directly into the present shock of conversing naturally with an encyclopedic AI that crafts images, writes code, and debates philosophy. Won’t this technology almost certainly transform society — and hasn’t AI’s impact on us so...
Artificial Ignorance
The hidden side of Apple Intelligence
More than just another keynote recap.
6 months ago
More than just another keynote recap.
Society's Backend
Backend Biweekly #1: Important AI Developments and ML Learning Resources
Updates on Apple, Mistral, and Microsoft and resources to build the GPT tokenizer, use MLX to train...
9 months ago
Updates on Apple, Mistral, and Microsoft and resources to build the GPT tokenizer, use MLX to train models, and benchmark LLMs
The Berkeley...
Why do Policy Gradient Methods work so well in Cooperative MARL? Evidence from Policy Representation
In cooperative multi-agent reinforcement learning (MARL), due to its on-policy nature, policy...
over a year ago
In cooperative multi-agent reinforcement learning (MARL), due to its on-policy nature, policy gradient (PG) methods are typically believed to be less sample efficient than value decomposition (VD) methods, which are off-policy. However, some recent empirical studies demonstrate...
One Useful Thing
Strategies for an Accelerating Future
Four questions to ask your organization.
10 months ago
Four questions to ask your organization.
Artificial Ignorance
How to build an AI search engine (Part 1)
Working with Claude, Brave, and streaming responses.
2 months ago
Working with Claude, Brave, and streaming responses.
Weighty Thoughts
Most AI startups are doomed
Just because it matters doesn’t mean it’s defensible or profitable
a year ago
Just because it matters doesn’t mean it’s defensible or profitable
AI Snake Oil
Is Avoiding Extinction from AI Really an Urgent Priority?
The history of technology suggests that the greatest risks come not from the tech, but from the...
a year ago
The history of technology suggests that the greatest risks come not from the tech, but from the people who control it
Daniel Miessler
Companies as Alaskan Fishing Boats
What if companies are supposed to be like Alaskan fishing boats? You know, the kind on The Deadliest...
over a year ago
What if companies are supposed to be like Alaskan fishing boats? You know, the kind on The Deadliest Catch. So you have this tiny crew of total badasses. Everyone is a superhero at their particular role because the crew needs to stay extremely small to protect profits. The...
Society's Backend
How to Use Benchmarks to Build Successful Machine Learning Systems
And what benchmarks mean for real-world applications
4 months ago
And what benchmarks mean for real-world applications
Artificial Ignorance
Tutorial: How to narrate video with Sora, GPT-Vision, and ElevenLabs
The future of entertainment is going to be a wild ride.
9 months ago
The future of entertainment is going to be a wild ride.
Artificial Ignorance
YCombinator's AI boom is still going strong (W24)
Combing through all 158 YC AI startups (65% of the batch).
8 months ago
Combing through all 158 YC AI startups (65% of the batch).
Daniel Miessler
NO. 357 | NEWS, ANALYSIS, & DISCOVERY SERIES
SECURITY NEWS Attackers have dumped nearly 8 million Australian health records on the dark web after...
over a year ago
SECURITY NEWS Attackers have dumped nearly 8 million Australian health records on the dark web after breaching a health insurance company with almost 10 million customers. MORE NSA has released guidance asking companies to switch to memory-safe languages like Rust, C#, Go, and...
The Gradient
Salmon in the Loop
On fish counting – a complex sociotechnical problem in a field that is going through the process of...
a year ago
On fish counting – a complex sociotechnical problem in a field that is going through the process of digital transformation.
One Useful Thing
Secret Cyborgs: The Present Disruption in Three Papers
The future is already here, we just need to figure out a few details.
a year ago
The future is already here, we just need to figure out a few details.
Society's Backend
OpenAI’s Blunder is a Loss for the ML Community
The timeline and why OpenAI’s actions are a big deal
7 months ago
The timeline and why OpenAI’s actions are a big deal
AI Snake Oil
A safe harbor for AI evaluation and red teaming
An argument for legal and technical safe harbors for AI safety and trustworthiness research
9 months ago
An argument for legal and technical safe harbors for AI safety and trustworthiness research
Strange Loop Canon
LLMs breach a threshold
Open Source models get more powerful, and an AI system scores silver in Maths Olympiad
4 months ago
Open Source models get more powerful, and an AI system scores silver in Maths Olympiad
Artificial Ignorance
The State of AI Engineering
Notes from the first AI Engineer Summit.
a year ago
Notes from the first AI Engineer Summit.
The Berkeley...
FIGS: Attaining XGBoost-level performance with the interpretability and speed of CART
FIGS (Fast Interpretable Greedy-tree Sums): A method for building interpretable models by...
over a year ago
FIGS (Fast Interpretable Greedy-tree Sums): A method for building interpretable models by simultaneously growing an ensemble of decision trees in competition with one another.
Recent machine-learning advances have led to increasingly complex predictive models, often at the cost...
Society's Backend
Why everyone loves Spider-Man
Lessons anyone can learn from our friendly neighborhood web-crawler
a year ago
Lessons anyone can learn from our friendly neighborhood web-crawler
Artificial Ignorance
Who's winning the AI arms race?
It depends whether AI is a "disruptive" or "sustaining" technology.
a year ago
It depends whether AI is a "disruptive" or "sustaining" technology.
Strange Loop Canon
Life in India is a series of bilateral negotiations
finding the new equilibrium for the next superpower
2 months ago
finding the new equilibrium for the next superpower
Full Width 
Shortcuts 
Sign Up 
Log In 
Articles 
Reading List 