Lukasz Kaiser
@lukaszkaiser.bsky.social
Reposted by Lukasz Kaiser
How can we use neural networks to bolster mathematical discovery? Geordie Williamson's @simonsfoundation.org Presidential Lecture is online, catch up now:
www.youtube.com/watch?v=Uxr_...
www.youtube.com/watch?v=Uxr_...
Geordie Williamson: Neural Networks for Mathematical Discovery (October 29, 2025)
YouTube video by Simons Foundation
www.youtube.com
November 10, 2025 at 1:10 AM
How can we use neural networks to bolster mathematical discovery? Geordie Williamson's @simonsfoundation.org Presidential Lecture is online, catch up now:
www.youtube.com/watch?v=Uxr_...
www.youtube.com/watch?v=Uxr_...
Reposted by Lukasz Kaiser
Fresh on the arXiv: @booleananalysis.bsky.social, Kewen Wu, and I present new classical algorithms for the Short Integer Solution problem (under infinity norm) that outperform the elegant Chen-Liu-Zhandry quantum algorithm, showing that there is no exponential quantum speed up anymore.
October 10, 2025 at 6:07 AM
Fresh on the arXiv: @booleananalysis.bsky.social, Kewen Wu, and I present new classical algorithms for the Short Integer Solution problem (under infinity norm) that outperform the elegant Chen-Liu-Zhandry quantum algorithm, showing that there is no exponential quantum speed up anymore.
Reposted by Lukasz Kaiser
We are starting to see some nuanced discussions of what it means to work with advanced AI in its current state
In this case, GPT-5 Pro was able to do novel math, but only when guided by a math professor (though the paper also noted the speed of advance since GPT-4)
The reflection is worth reading.
In this case, GPT-5 Pro was able to do novel math, but only when guided by a math professor (though the paper also noted the speed of advance since GPT-4)
The reflection is worth reading.
September 6, 2025 at 9:55 PM
We are starting to see some nuanced discussions of what it means to work with advanced AI in its current state
In this case, GPT-5 Pro was able to do novel math, but only when guided by a math professor (though the paper also noted the speed of advance since GPT-4)
The reflection is worth reading.
In this case, GPT-5 Pro was able to do novel math, but only when guided by a math professor (though the paper also noted the speed of advance since GPT-4)
The reflection is worth reading.
Reposted by Lukasz Kaiser
A fully autonomous robot which, every morning, sets plates on the table, fetches ingredients in the kitchen, and prepares avocado toast.
"Move things and breakfast."
"Move things and breakfast."
June 5, 2025 at 11:34 AM
A fully autonomous robot which, every morning, sets plates on the table, fetches ingredients in the kitchen, and prepares avocado toast.
"Move things and breakfast."
"Move things and breakfast."
Reposted by Lukasz Kaiser
(In case you hadn't been following, the environmental impact of current AI models is now much lower, generating 100,000 words with AI uses less power than watching Netflix for 45 minutes on your TV)
May 20, 2025 at 4:16 PM
(In case you hadn't been following, the environmental impact of current AI models is now much lower, generating 100,000 words with AI uses less power than watching Netflix for 45 minutes on your TV)
Reposted by Lukasz Kaiser
If you haven't done this with o3, you haven't really seen what these models can do.
May 11, 2025 at 2:05 AM
If you haven't done this with o3, you haven't really seen what these models can do.
Reposted by Lukasz Kaiser
This is one of the most-shared posts on Bluesky in the past day and it's just completely false. You might think ChatGPT is a *bad* search engine, or prefer another search engine. But it has had integrated web search since last year.
May 1, 2025 at 5:26 PM
This is one of the most-shared posts on Bluesky in the past day and it's just completely false. You might think ChatGPT is a *bad* search engine, or prefer another search engine. But it has had integrated web search since last year.
Reposted by Lukasz Kaiser
"o3, You are a consultant hired by the Dark Lord, analyze the org chart of Mordor. How would you improve it for today's changing Middle Earth"
o3 does some actual satire, ending with: “One Org to rule them all, One Org to find them, One Org to bring them all, And in the darkness, align them.”
o3 does some actual satire, ending with: “One Org to rule them all, One Org to find them, One Org to bring them all, And in the darkness, align them.”
April 28, 2025 at 4:38 AM
"o3, You are a consultant hired by the Dark Lord, analyze the org chart of Mordor. How would you improve it for today's changing Middle Earth"
o3 does some actual satire, ending with: “One Org to rule them all, One Org to find them, One Org to bring them all, And in the darkness, align them.”
o3 does some actual satire, ending with: “One Org to rule them all, One Org to find them, One Org to bring them all, And in the darkness, align them.”
Reposted by Lukasz Kaiser
For years I've been throwing the same puzzle challenge at new GPT models. Every one has failed, until now.
matthodges.com/posts/2025-0...
matthodges.com/posts/2025-0...
GPT Finally Jumped Out of the System
OpenAI’s o4-mini-high Model Solves the MU Puzzle and Demonstrates Why
matthodges.com
April 22, 2025 at 4:46 AM
For years I've been throwing the same puzzle challenge at new GPT models. Every one has failed, until now.
matthodges.com/posts/2025-0...
matthodges.com/posts/2025-0...
Reposted by Lukasz Kaiser
Exciting news: @waymo.bsky.social is beginning public service on the Peninsula, starting with Palo Alto, Mountain View, and Los Altos! Initial service area below.
March 12, 2025 at 1:21 AM
Exciting news: @waymo.bsky.social is beginning public service on the Peninsula, starting with Palo Alto, Mountain View, and Los Altos! Initial service area below.
Reposted by Lukasz Kaiser
Try it or improve it: chatgpt.com/canvas/share...
Dn Dguild Hall Simulator
chatgpt.com
February 14, 2025 at 1:47 AM
Try it or improve it: chatgpt.com/canvas/share...
Reposted by Lukasz Kaiser
This was fun: "o1, build a simulator of a D&D guild hall. Persistent characters come in, get quests, interact with each other, leave & return, make it procedurally generated"
I kept asking it to add other ideas (relationships, etc) 8 times, got no errors, just worked each time. Desire-based coding!
I kept asking it to add other ideas (relationships, etc) 8 times, got no errors, just worked each time. Desire-based coding!
February 14, 2025 at 1:46 AM
This was fun: "o1, build a simulator of a D&D guild hall. Persistent characters come in, get quests, interact with each other, leave & return, make it procedurally generated"
I kept asking it to add other ideas (relationships, etc) 8 times, got no errors, just worked each time. Desire-based coding!
I kept asking it to add other ideas (relationships, etc) 8 times, got no errors, just worked each time. Desire-based coding!
Reposted by Lukasz Kaiser
The NIH overhead cut doesn't just hurt universities.
It's deadly to the US economy.
The US is a world leader in tech due to the ecosystem that NIH and NSF propel. It drives innovation for tech transfer, creates a highly-skilled sci/tech workforce, and fosters academic/industry crossfertilization.
It's deadly to the US economy.
The US is a world leader in tech due to the ecosystem that NIH and NSF propel. It drives innovation for tech transfer, creates a highly-skilled sci/tech workforce, and fosters academic/industry crossfertilization.
2. While NSF and NIH indeed have a mission to fund specific research innovations via grantmaking, they do a lot more than that.
Their principal role is support a scientific ecosystem in the United States, that includes everything from education and training to infrastructure and communication.
Their principal role is support a scientific ecosystem in the United States, that includes everything from education and training to infrastructure and communication.
February 8, 2025 at 2:03 AM
The NIH overhead cut doesn't just hurt universities.
It's deadly to the US economy.
The US is a world leader in tech due to the ecosystem that NIH and NSF propel. It drives innovation for tech transfer, creates a highly-skilled sci/tech workforce, and fosters academic/industry crossfertilization.
It's deadly to the US economy.
The US is a world leader in tech due to the ecosystem that NIH and NSF propel. It drives innovation for tech transfer, creates a highly-skilled sci/tech workforce, and fosters academic/industry crossfertilization.
Reposted by Lukasz Kaiser
Saturday ice update - #Arctic sea ice extent is currently the *lowest* on record (JAXA data)
• about 790,000 km² below the 2010s mean
• about 1,450,000 km² below the 2000s mean
• about 2,040,000 km² below the 1990s mean
• about 2,430,000 km² below the 1980s mean
Plots: zacklabe.com/arctic-sea-i...
• about 790,000 km² below the 2010s mean
• about 1,450,000 km² below the 2000s mean
• about 2,040,000 km² below the 1990s mean
• about 2,430,000 km² below the 1980s mean
Plots: zacklabe.com/arctic-sea-i...
February 8, 2025 at 7:47 PM
Saturday ice update - #Arctic sea ice extent is currently the *lowest* on record (JAXA data)
• about 790,000 km² below the 2010s mean
• about 1,450,000 km² below the 2000s mean
• about 2,040,000 km² below the 1990s mean
• about 2,430,000 km² below the 1980s mean
Plots: zacklabe.com/arctic-sea-i...
• about 790,000 km² below the 2010s mean
• about 1,450,000 km² below the 2000s mean
• about 2,040,000 km² below the 1990s mean
• about 2,430,000 km² below the 1980s mean
Plots: zacklabe.com/arctic-sea-i...
Reposted by Lukasz Kaiser
OpenAI’s deep research is very good. Unlike Google’s version, which is mostly a good summarizer of many sources, OpenAI is more like engaging an opinionated (often almost PhD-level!) researcher who follows lead.
Look at how it hunts down a concept in the literature (& works around problems)
Look at how it hunts down a concept in the literature (& works around problems)
February 3, 2025 at 12:12 AM
OpenAI’s deep research is very good. Unlike Google’s version, which is mostly a good summarizer of many sources, OpenAI is more like engaging an opinionated (often almost PhD-level!) researcher who follows lead.
Look at how it hunts down a concept in the literature (& works around problems)
Look at how it hunts down a concept in the literature (& works around problems)
Reposted by Lukasz Kaiser
In all seriousness how batshit is it that a Chinese AI bot is censoring a book THAT HASN'T EVEN BEEN PUBLISHED YET. What dystopia are we all living in.
🤐 DeepSeek doesn't like our graphic novel, YOU MUST TAKE PART IN REVOLUTION — which isn't even out until March 4th! So... ya'll should go out and pre-order the comic, right?! Fight the censorship! Join @badiucao.bsky.social and I in the revolution! 😆 Order here: melissachan.com/revolution
January 28, 2025 at 7:39 PM
In all seriousness how batshit is it that a Chinese AI bot is censoring a book THAT HASN'T EVEN BEEN PUBLISHED YET. What dystopia are we all living in.
Reposted by Lukasz Kaiser
this post is trending in my feed but it does not make sense. i don't see any reasonable interpretation by which DeepSeek demonstrate that model scaling is not the best way to develop AI. their model is very large, and their training corpus is very large. they were just scaling more efficiently.
Much of the coverage has been focused on US-China tech competition. That misses a bigger story: DeepSeek has demonstrated that scaling up AI models relentlessly, a paradigm OpenAI introduced & champions, is not the only, and far from the best, way to develop AI. 3/
January 28, 2025 at 12:58 AM
this post is trending in my feed but it does not make sense. i don't see any reasonable interpretation by which DeepSeek demonstrate that model scaling is not the best way to develop AI. their model is very large, and their training corpus is very large. they were just scaling more efficiently.
Reposted by Lukasz Kaiser
No.
See: cloud.google.com/sustainabili...
www.gstatic.com/gumdrop/sust... has info about our net zero goals (which are goals for the year 2030 that we are working towards). p.30: "We maintained 64% carbonfree energy, on average, across every grid where we operate." (for 2023).
See: cloud.google.com/sustainabili...
www.gstatic.com/gumdrop/sust... has info about our net zero goals (which are goals for the year 2030 that we are working towards). p.30: "We maintained 64% carbonfree energy, on average, across every grid where we operate." (for 2023).
January 11, 2025 at 10:59 PM
No.
See: cloud.google.com/sustainabili...
www.gstatic.com/gumdrop/sust... has info about our net zero goals (which are goals for the year 2030 that we are working towards). p.30: "We maintained 64% carbonfree energy, on average, across every grid where we operate." (for 2023).
See: cloud.google.com/sustainabili...
www.gstatic.com/gumdrop/sust... has info about our net zero goals (which are goals for the year 2030 that we are working towards). p.30: "We maintained 64% carbonfree energy, on average, across every grid where we operate." (for 2023).
2025 = (20+25)^2 ; A year like that is bound to be remarkable. Happy 2025!!
January 2, 2025 at 4:58 AM
2025 = (20+25)^2 ; A year like that is bound to be remarkable. Happy 2025!!
Reposted by Lukasz Kaiser
Sharks are older than the rings of Saturn.
This paper finds that the rings are no older than 400M years. Sharks date back to at least the Late Ordovician Period, 450M years ago.
This paper finds that the rings are no older than 400M years. Sharks date back to at least the Late Ordovician Period, 450M years ago.
December 8, 2024 at 6:54 AM
Sharks are older than the rings of Saturn.
This paper finds that the rings are no older than 400M years. Sharks date back to at least the Late Ordovician Period, 450M years ago.
This paper finds that the rings are no older than 400M years. Sharks date back to at least the Late Ordovician Period, 450M years ago.