Andy Masley

Here's the paper www.nature.com/articles/s44...

AI-driven productivity gains enable more CO₂ emissions than they avoid in a global energy–economy model - npj Climate Action nature.com

I’m incredibly honored and grateful to be named one of @time.com's 100 people in AI this year! When I started writing I felt a little crazy as just another guy with an AI blog trying to budge a global debate. It's been amazing to actually move how people think. time.com/collec…

The 100 Most Influential People in AI 2026 time.com

A lot of people will say they don't believe my numbers on individual AI prompts not being bad. You can use this collection of estimates from outside experts and just see for yourself how many times you'd have to send each prompt to raise your carbon footprint by 1%. huggingfac…

EcoLogits Calculator - a Hugging Face Space by genai-impact huggingface.co

www.andymasley.com/visuals/ai-p...

What does your chatbot use actually cost? andymasley.com

I think I figured out exactly how the “AI used a bottle of water per prompt” miscalculation happened open.substack.com/pub/andymasl...

I think I figured out exactly how the “AI used a bottle of water per prompt” miscalculation happened blog.andymasley.com

More and more previous stats are starting to be thrown around about how China's influencing the data center debate. Most of these don't hold up once you know what they were measuring. If you skim my post here you'll be able to recognize when they're coming up blog.andymasley.c…

I'm very skeptical that China has played a meaningful role in the US data center backlash blog.andymasley.com

Wrote a rant about how other large industries reveal how impoverished the data center debate is based on the image of the steel mill I posted yesterday blog.andymasley.com/p/other-larg...

Other large industries show how impoverished the data center debate is blog.andymasley.com

This article was basically exactly what I'd like all reporting on data centers to look like. A solid grappling with costs and benefits + interviewing locals who represent the full spectrum of opinion www.sfchronicle.com/opinion/arti...

Client Challenge sfchronicle.com

Just posted the first of a few posts on what I see as the history of the data center panic, this first one is on what was happening pre-ChatGPT blog.andymasley.com/p/a-history-...

A history of the data center panic - part 1 blog.andymasley.com

I wrote out why I don't think the data center backlash seems especially helpful for AI safety (though obviously it's a good chance to recruit people to separately think about AI more) and why I and others into EA don't want to just lie to people about them blog.andymasley.com/…

AI safety and the data center backlash blog.andymasley.com

The AI labs really need to improve their image. They could form some sort of political pac, maybe call it the "AIPAC." I'm sure it'd be really popular

Feel like a lot of AI conversations are at about the level of someone saying "I like Wikipedia and YouTube" and then someone else saying "Oh so you're pro internet, you're an internet booster? You don't see any potential problems with the internet?"

If I avoid the word "alignment" and just ask "If we keep scaling AI models up, will we reliably avoid situations where they do something catastrophic pursuing a normal goal?" the answer seems like no right now

Wow this farmer has a message for AI companies who want to buy his land

My big main objection to how people talk about chatbots is they often imply the majority of people are so tasteless and stupid that they need to be protected from their own use of a talking computer

Had two separate people I knew from college and hadn't talked to in a while mention that AI had helped them find a life changing medical diagnosis breakthrough, it was a little wild and I wasn't fishing for that at all.

It's been almost a year since I posted The AI Water Issue is Fake. In that time, I'm not aware of an instance where the normal operation of a data center harmed people's water access.

If the AI water issue were real, it'd be easy to go to people's homes that were actually affected by the normal operational water draw of a data center. Instead, MPU always chooses: 1) Places polluted by construction 2) Protestors without data centers nearby 3) Pecan farmers It's been over a year

I've had a very nice back-and-forth email exchange with Shaolei Ren, who provided the original "AI uses a bottle of water per prompt" estimate with the Washington Post. (1/5)

I don't really get how you can believe: 1) Superintelligence is possible in principle within a decade 2) We should be pretty chill about it and just march forward