Reposted by Abhinav Sharma
amazing q&a with paul krugman
June 25, 2025 at 1:22 PM
amazing q&a with paul krugman
When answers are cheap, how do you train to ask great questions?
May 26, 2025 at 4:26 PM
When answers are cheap, how do you train to ask great questions?
When I first read this, everyone around me had read this. Not sure how true that still is, so here we go calteches.library.caltech.edu/51/2/CargoCu...
March 27, 2025 at 6:00 AM
When I first read this, everyone around me had read this. Not sure how true that still is, so here we go calteches.library.caltech.edu/51/2/CargoCu...
So long as our approach to intelligence remains vectorizing and compressing all human-identified concepts and their relationships, there will always be more work to do.
March 23, 2025 at 7:02 PM
So long as our approach to intelligence remains vectorizing and compressing all human-identified concepts and their relationships, there will always be more work to do.
Curious how "thinking" in LLMs involves spitting out a bunch of extra words, but our best moments of thinking often involve a profound silence right before. They both "work," but hmm curious
March 23, 2025 at 12:22 AM
Curious how "thinking" in LLMs involves spitting out a bunch of extra words, but our best moments of thinking often involve a profound silence right before. They both "work," but hmm curious
Surprised at how much you can learn about model reasoning weaknesses from having them create mock data on a graph of entity types.
March 22, 2025 at 6:54 PM
Surprised at how much you can learn about model reasoning weaknesses from having them create mock data on a graph of entity types.
For private, desktop computing (each of those words chosen carefully) we'll move away from fancy overcomplicated interfaces back to basic file formats that map to fundamental tools for thought (document, video, table, flowchart, etc.) and AI doing the heavy lifting.
March 22, 2025 at 5:50 PM
For private, desktop computing (each of those words chosen carefully) we'll move away from fancy overcomplicated interfaces back to basic file formats that map to fundamental tools for thought (document, video, table, flowchart, etc.) and AI doing the heavy lifting.
Claude 3.7 (and probably the others) are so good at generating "good looking" design with terrible information hierarchy that I wonder how many teams are shipping stuff that unintentionally and subtly misdirects users.
March 22, 2025 at 4:39 PM
Claude 3.7 (and probably the others) are so good at generating "good looking" design with terrible information hierarchy that I wonder how many teams are shipping stuff that unintentionally and subtly misdirects users.
We need a markdown for spreadsheets for AI's sake. How close to possible it it even? Is CSV the best 80/20, or SQL?
March 22, 2025 at 3:37 PM
We need a markdown for spreadsheets for AI's sake. How close to possible it it even? Is CSV the best 80/20, or SQL?
why is it that everyone seems to know everyone, even though each of us has a small number of friends?
the emergence of the giant component is one of my favorite results in network theory, for how simple but profound it is.
the emergence of the giant component is one of my favorite results in network theory, for how simple but profound it is.
March 21, 2025 at 8:52 PM
why is it that everyone seems to know everyone, even though each of us has a small number of friends?
the emergence of the giant component is one of my favorite results in network theory, for how simple but profound it is.
the emergence of the giant component is one of my favorite results in network theory, for how simple but profound it is.
In some ways, the internet will be fine post LLM steady state. Some of its best content has been raw data and research put up for free from financially protected academic resources, which have pushed hard to open up in the last decade.
March 21, 2025 at 2:10 AM
In some ways, the internet will be fine post LLM steady state. Some of its best content has been raw data and research put up for free from financially protected academic resources, which have pushed hard to open up in the last decade.
The ideal human data collection UI is screen recording everything on both sides of a remote work session with a thinkaloud, with AI usage is allowed. Scaled oversight.
March 20, 2025 at 7:03 PM
The ideal human data collection UI is screen recording everything on both sides of a remote work session with a thinkaloud, with AI usage is allowed. Scaled oversight.
Tiktok remains the most wondrous tool ever made to vectorize a human's limbic preferences
March 20, 2025 at 6:35 PM
Tiktok remains the most wondrous tool ever made to vectorize a human's limbic preferences
Pretty sure that Claude 3.7’s effusiveness is a play to collect more nuanced feedback.
March 20, 2025 at 5:23 PM
Pretty sure that Claude 3.7’s effusiveness is a play to collect more nuanced feedback.
A useful heuristic for UI post AI — interfaces are just templates for good thinking.
March 20, 2025 at 3:05 PM
A useful heuristic for UI post AI — interfaces are just templates for good thinking.
Flatter Gmail today. Excited for the UI trend of *get out of my way and just make it easier for my AI to understand you*
March 18, 2025 at 10:32 PM
Flatter Gmail today. Excited for the UI trend of *get out of my way and just make it easier for my AI to understand you*
I want a biomechanics-vision-language model. I half joke I should do yoga teacher training so I still have a job. But but the physiotherapy part would be cool to encode.
Anyone know of one?
Anyone know of one?
March 18, 2025 at 4:57 PM
I want a biomechanics-vision-language model. I half joke I should do yoga teacher training so I still have a job. But but the physiotherapy part would be cool to encode.
Anyone know of one?
Anyone know of one?
The new software engineering interview should be a reverse turing test where you review code on a spectrum of amateur vibe coder to Jeff Dean
March 18, 2025 at 2:01 PM
The new software engineering interview should be a reverse turing test where you review code on a spectrum of amateur vibe coder to Jeff Dean
Programming languages (their most common use) has a thinking to writing ratio. JSX is probably the lowest. But if you spend a lot of time in it you might think language models can do anything.
March 18, 2025 at 2:59 AM
Programming languages (their most common use) has a thinking to writing ratio. JSX is probably the lowest. But if you spend a lot of time in it you might think language models can do anything.
The more lines of code you put down that feel like a harsh reality tradeoff, the less likely it’ll be replaced by an LLM
March 18, 2025 at 2:35 AM
The more lines of code you put down that feel like a harsh reality tradeoff, the less likely it’ll be replaced by an LLM
Software, like clocks, is largely a coordination tool
March 14, 2025 at 4:30 PM
Software, like clocks, is largely a coordination tool