🔥The 9th Round of Easy Loan, Earn $40 Reward is in progress❗️ ⏰ Promotion Period: January 15th - Feburary 15th, 2025 👉 Register now and check more details at gate.io/campaigns/358

The origin story of “AI as Normal Technology”, and lessons learned Many people have asked how the “AI as Normal Technology” paper came to be. This paper has been an (ongoing) journey for me and Sayash Kapoor in developing not just the substance of our arguments but also learning how

How to tell someone they are utterly wrong and probably stupid and likely immoral, but nicely, so that good things may yet come of it.

Can we uncover the list of topics a language model is censored on? Refused topics vary strongly among models. Claude-3.5 vs DeepSeek-R1 refusal patterns:

I just looked into "speckle interferometry" which is a way to correct for atmospheric distortion by processing many short exposure images from a telescope... and man, there is no end to the magic you can do with a Fourier transform. Holograms, MRI, SAR, and now this!

I used to be an The Associated Press editor. When everything happens all at once, it's AP you want. There's nothing on earth like it. apnews.com/article/israel…

I think it's awesome that X is airing the dirty laundry of platform moderation. Every platform does this, X does (some of) it in public.

Sure, people in any of these places might have posted the same information. Might even have made the same phone calls. It's possible. But sure, tell me again that social media will replace professional journalism.

Please send this link to every college-age person you know (including grad students) who might have the time and inclination to spend 10 minutes a day this summer helping bolster a culture of decency and democracy in our country. columbiaspectator.com/opinion/2025/0…

Yeah. A lot of academic criticism of AI uses snappy put-downs ("bullshit generator", "stochastic parrot") instead of really engaging with a very fast moving field. Remember: today's AI is the worst it's ever going to be. learningfromexamples.com/p/what-academi…

Excellent write-up of the conditions that lead to market failures and misalignment with human flourishing, and how AI-enabled market intermediaries could address this misalignment by scaling outcome-based contracting.

It's been 100+ years since "The quick brown fox..." so I've been messing about with AI + genetic algorithms to generate new pangrams. And it's not working at all. Turns out, GAs actually kinda suck. Why? Well, evolution had *way* more compute than we do. statmodeling.stat.columbia.edu/2025/04/17/wha…

Are you a researcher interested in how technology can build genuine connections between people? Prosocial Design Network is co-hosting a research<>practice convening with Roblox on "Fostering Connection in Social Gaming," October 17 linkedin.com/feed/update/ur…

One of the social media algorithms we tested was "surprising validators," which tried to show things you agreed with from sources you distrust. It was nicknamed "heartbreak" in honor of this famous Clickhole article. clickhole.com/heartbreaking-…

Rolling out an *Open API for AI Community Note Writers*!! Anyone can write their own AI note writer to propose notes, but they will be treated the same as any note proposal: notes need to be rated as helpful by (human) raters who typically disagree before they show publicly

two good projects emerged this week me and Jonathan Stray wrote a good piece in the better conflict bulletin on, well, it’s in the name i also created a microsite w/ a list of nearly 400 methods of non violent action inspired by the famous gene sharp 198 point list links below