AI
Matt Ross and
AI
Matt Ross and
Through extensive experimentation across diverse puzzles, we show that frontier LRMs face a complete accuracy collapse beyond certain complexities. Moreover, they exhibit a counter-intuitive scaling limit: their reasoning effort increases with problem complexity up to a point, then declines despite having an adequate token budget. By comparing LRMs
... See more“The Illusion of Thinking: Understanding the Strengths and Limitations of Reasoning Models via the Lens of Problem Complexity.” They can’t think. Water is wet.
In 2013, the Chicago Sun-Times fired all 28 of its staff photographers. Every single one. Including John H. White, who had won a Pulitzer Prize.
The plan was simple. Reporters would shoot their own photos using iPhones. The technology had gotten good enough. The cameras were in everyone’s pockets. Why pay 28 salaries for something the existing staff
... See moreI feel this intensely and think about it all the time. Many people for whom I have tremendous respect, people I have been reading for many years, people who I look to to make sense of the world really do seem to dislike AI. All I can tell you is that for the last few years, I have been obsessively playing with it, building things with it and spending a lot of time trying to make sense of it. I can’t stop. It’s the exact same feeling and drive I had when I first installed Mosaic in the 90s. But this feels quite a bit different.