Short posts
Just tried out Atlas (OpenAI’s new browser). Asked it to find me some cheap ram.
3 mins later, it told me Microcenter’s best price was $299. I checked manually (took 10s) and found one at $183. None are $299.
🥸
(Video sped up 5x)
My lukewarm take is that this might work sometimes for some tasks. But woof are the privacy and security implications bad. The risk/reward is hopelessly unbalanced.
Trump has posted at least 62 AI-generated images or videos on his socials, which NYT has rounded up here https://www.nytimes.com/interactive/2025/10/21/business/media/trump-ai-truth-social-no-kings.html
New from me: Last year, the best open-weight AI models were made in the U.S. Now, they are all made in China.
More data and what it means -> 🎁 wapo.st/4nPUBud
My colleagues obtained recordings of a private lecture series by tech investor Peter Thiel where he:
- calls AI critics are “the Antichrist”
- said wealth gives the “illusion of power and autonomy” (he’s worth $27 billion)
- pitches a religious vision of Silicon Valley
Just saw the end of the Phillies-Dodgers, absolutely brutal. Bartman has nothing on that.
We’re about to get flooded with deepfakes.
Here’s an AI-generated clip of me “asking” Sam Altman what they train their systems on (made in 10 seconds with Sora 2):
Here are a couple real answers, if you’re still into that whole truth thing:
Just ran some evals on Claude Sonnet 4.5. It’s better than 4 on some but worse on a lot. LLM progress is so weird. You really gotta test this stuff on what you care about.
Worth a read: OpenAI released an eval for real work tasks across a bunch of industries. They didn’t release the individual results (lame), but you can replicate them from the prompts and files.
A usable nugget: If you’re outputting pdfs, xlsx or pptx, use Claude.
https://openai.com/index/gdpval/
New by me: OpenAI won’t say whose content trained its video tool. We found some clues.
Google’s blog post on launching Gemini in Chrome does not include the word “privacy” or “security.” Am I missing something or are they not addressing the very real threat of prompt injections?
https://blog.google/products/chrome/new-ai-features-for-chrome/
