1261
Fighting the New York Times' invasion of user privacy
OpenAI is fighting the New York Times’ demand for 20 million private ChatGPT conversations and accelerating new security and privacy protections to protect your data.
Topics /
AI & Machine Learning
The most popular blog posts about ai & machine learning on Hacker News. 6,306 posts.
1262
Defamed by ChatGPT
Yesterday, President Joe Biden declared that “it remains to be seen” whether Artificial Intelligence (AI) is “dangerous.” I would beg to differ. I have been writing about th…
1263
Non-determinism in GPT-4 is caused by Sparse MoE
It’s well-known at this point that GPT-4/GPT-3.5-turbo is non-deterministic, even at temperature=0.0. This is an odd behavior if you’re used to dense decoder-only models, where temp=0 should imply greedy sampling which should imply full determinism, because the logits for the next token should be a pure function of…
1264
Run CLIP on iPhone to search photos
How I run CLIP (ViT-B-32) on iPhone to search the Photos album offline with Queryable: exporting Core ML text and image encoders for CLIP photo search
1265
1266
$100 AI Music Video: Claude Fable 5 vs. GPT-5.6 Sol
We gave Claude Fable 5 and GPT-5.6 Sol the same song, a budget, web search, and local ffmpeg, then let each autonomously direct a music video.
1267
Feds Target AI Critics as "Foreign Agents"
Trump admin sees China behind opposition to AI data centers
1268
There are no "rogue" AI agents
Identifying them as such only lets companies like OpenAI off the hook.
1269
GPT-5-Codex
Codex just got faster, more reliable, and better at real-time collaboration and tackling tasks independently anywhere you develop—whether via the terminal, IDE, web, or even your phone.
1270
Neural Networks in JavaScript with Deeplearn.js
An extensive tutorial on how to implement a neural network in JavaScript with deeplearn.js to perform machine learning in JavaScript. You can bring neural networks to your browser by using the performance benefits of WebGL. Deeplearn.js makes it possible to implement a WebGL powered neural network ...
1271
1272
GPT-3.5 Turbo fine-tuning and API updates
Developers can now bring their own data to customize GPT-3.5 Turbo for their use cases.
1273
1274
How to Think About GPUs
We love TPUs at Google, but GPUs are great too. This chapter takes a deep dive into the world of GPUs – how each chip works, how they're networked together, and what that means for LLMs, especially compared to TPUs. While there are a multitude of GPU architectures from NVIDIA, AMD, Intel, and others, here we will…
1275
Getting 50% (SoTA) on Arc-AGI with GPT-4o
You can just draw more samples
1276
Muse Spark: Scaling towards personal superintelligence
Today, we’re excited to introduce Muse Spark, the first in the Muse family of models developed by Meta Superintelligence Labs. Muse Spark is a natively multimodal reasoning model with support for tool-use, visual chain of thought, and multi-agent orchestration.
1277
ARCore: Augmented reality at Android scale
Today, we’re releasing a preview of a new SDK called ARCore. It brings augmented reality capabilities to existing and future Android phones. Developers can start experim…
1278
Decisions API is in public beta
Use the Decisions API to check conditions, select from fixed options, and score text, images, and JSON against a rubric.
1279
Time Series Prediction – A short introduction for pragmatists
Are you trying to predict time series but don't know where to start? This blog post will provide a comparison of the most prominent techniques and show you how …
1280
My experience with Claude Code after two weeks of adventures
Hatching... Cursor Shenanigans Cursor, my beloved, started rate limiting shenanigans a few days back. For a good 2 weeks after June 16, 2025, we had almo...
1281
Agents don't need memory, they need documentation
Why agents need project documentation instead of retrieval from old conversations.
1282
LLMs are steroids for your Dunning-Kruger
ChatGPT and LLMs amplify both insight and illusion. They are not knowledge engines, but confidence engines shaping how we think and learn.
1283
PoisonGPT: We hid a lobotomized LLM on Hugging Face to spread fake news
We will show in this article how one can surgically modify an open-source model, GPT-J-6B, and upload it to Hugging Face to make it spread misinformation while being undetected by standard benchmarks.
1284
The latest chapter for the self-driving car: mastering city street driving
Over the last year we’ve shifted the focus of the Google self-driving car project onto mastering city street driving.
1285
Waymo and Toyota outline partnership to advance autonomous driving deployment
Toyota Motor Corporation (“Toyota”) and Waymo reached a preliminary agreement to explore a collaboration focused on accelerating the development and deployment of autonomous driving technologies.
1286
Remastering Star Trek: Deep Space Nine with Machine Learning
The site for the Remako Mod, a Final Fantasy VII mod that uses AI to improve the graphics of this classic JRPG.
1287
YOLOv5: State-of-the-art object detection at 140 FPS
YOLOv5 license and commercial-use guide: AGPL-3.0 terms, specs, resources, and alternatives. Compare YOLOv5 against other vision models free.
1288
Can AI do maths yet? Thoughts from a mathematician
So the big news this week is that o3, OpenAI’s new language model, got 25% on FrontierMath. Let’s start by explaining what this means.
1289
When everyone has AI and the company still learns nothing
Are people using AI, or is the organization learning from it? What changed because we spent those tokens? And who moves discoveries from individuals to teams to organizational capabilities?
1290
Why I Left Google DeepMind
I fought against Google’s Pentagon AI deal from the inside. Powerful people and institutions failed to keep their AI ethics promises under pressure.