Senko @senkobento - Twitter Profile

about 1 month ago

Some people refer to AI (LLMs) as simply "stochastic parrots." I think that misses a really profound bigger picture. In 1960 the Physicist Eugene Wigner wrote an essay called "The Unreasonable Effectiveness of Mathematics in the Natural Sciences." His puzzle was that mathematics invented for its own sake-- for elegance, with no application in mind-- kept turning out to describe the physical world (physics). Since then, the same unreasonable effectiveness has reached the human stuff. Language. Conversation. Reasoning. The human mind. Nobody intentionally designed next-token prediction to be a theory of thought. What I think is absolutely incredible is just how much LLMs can seemingly comprehend real human interactions and can also reason just as good if not better than most humans purely through the poetry of mathematics. The obvious objection is that this isn't surprising at all: language encodes human reasoning because humans wrote it down, so of course modeling language recovers it. Fair enough. But I think that just moves the wonder goal post. It means language itself carries far more structure than we assumed. What's more, the record we left of our own thinking turns out to be complete enough to reconstruct that thinking. Either way, I'm truly in awe of this. Something is being described that we didn't think was describable. I'm very grateful for the AI safety researchers working hard on worst-case scenarios. AND, I also think there's room to find this beautiful and have some hope for the future. Mathematics doesn't only structure the universe. It seems to structure how we think, reason, and understand each other-- and we're only finding that out now.

sbkaufman's tweet photo. Some people refer to AI (LLMs) as simply "stochastic parrots." I think that misses a really profound bigger picture.

In 1960 the Physicist Eugene Wigner wrote an essay called "The Unreasonable Effectiveness of Mathematics in the Natural Sciences." His puzzle was that mathematics invented for its own sake-- for elegance, with no application in mind-- kept turning out to describe the physical world (physics).

Since then, the same unreasonable effectiveness has reached the human stuff. Language. Conversation. Reasoning. The human mind. Nobody intentionally designed next-token prediction to be a theory of thought. What I think is absolutely incredible is just how much LLMs can seemingly comprehend real human interactions and can also reason just as good if not better than most humans purely through the poetry of mathematics.

The obvious objection is that this isn't surprising at all: language encodes human reasoning because humans wrote it down, so of course modeling language recovers it. Fair enough. But I think that just moves the wonder goal post. It means language itself carries far more structure than we assumed. What's more, the record we left of our own thinking turns out to be complete enough to reconstruct that thinking.

Either way, I'm truly in awe of this. Something is being described that we didn't think was describable. I'm very grateful for the AI safety researchers working hard on worst-case scenarios. AND, I also think there's room to find this beautiful and have some hope for the future. Mathematics doesn't only structure the universe. It seems to structure how we think, reason, and understand each other-- and we're only finding that out now.

39

141

30

94

12K

senkobento retweeted

Alex Veremeyenko

@alex_verem

about 1 month ago

MIT published a brutally honest report on what AI is doing to students. A committee of professors and students spent five months studying how AI changed learning on campus, and the findings read like a warning to every university on the planet. Study groups are disappearing. Office hours are emptying out. Problem sets and take-home exams no longer prove anything, because AI can produce credible solutions to almost any written assignment in the undergraduate curriculum. Students who lean on chatbots lose mastery and confidence, and some slip into what the report calls cognitive surrender, reaching for AI at the first hint of struggle. The numbers are rough. 46 percent of surveyed MIT undergrads use LLMs daily. 90 percent worry about their own overreliance. Undergrads who feel AI makes them replaceable now outnumber those who feel it makes them capable. The committee's answer surprised me. They refused to fight AI with surveillance. The report calls AI detectors unreliable, says lockdown browsers feel like spying, and warns that policing students builds a classroom atmosphere of mutual distrust. Instead, MIT wants to rebuild education around the things AI can't replace. That means oral exams, semester portfolios, in-person project work, and a required social component in every subject. The report even floats the idea of rethinking grades entirely, since without a GPA to optimize, much of the incentive to cheat with AI evaporates. The committee warns professors against replacing undergrad research assistants with AI agents just because they're cheaper, because a university exists to grow people, not output. The most famous tech school on earth admitted the machines broke its way of teaching. Its answer is more humans, not more software.

alex_verem's tweet photo. MIT published a brutally honest report on what AI is doing to students.

A committee of professors and students spent five months studying how AI changed learning on campus, and the findings read like a warning to every university on the planet.

Study groups are disappearing. Office hours are emptying out. Problem sets and take-home exams no longer prove anything, because AI can produce credible solutions to almost any written assignment in the undergraduate curriculum. Students who lean on chatbots lose mastery and confidence, and some slip into what the report calls cognitive surrender, reaching for AI at the first hint of struggle.

The numbers are rough. 46 percent of surveyed MIT undergrads use LLMs daily. 90 percent worry about their own overreliance. Undergrads who feel AI makes them replaceable now outnumber those who feel it makes them capable.

The committee's answer surprised me. They refused to fight AI with surveillance. The report calls AI detectors unreliable, says lockdown browsers feel like spying, and warns that policing students builds a classroom atmosphere of mutual distrust.

Instead, MIT wants to rebuild education around the things AI can't replace. That means oral exams, semester portfolios, in-person project work, and a required social component in every subject. The report even floats the idea of rethinking grades entirely, since without a GPA to optimize, much of the incentive to cheat with AI evaporates.

The committee warns professors against replacing undergrad research assistants with AI agents just because they're cheaper, because a university exists to grow people, not output.

The most famous tech school on earth admitted the machines broke its way of teaching. Its answer is more humans, not more software.

484

25K

8K

14K

3M

Senko

@senkobento

Who to follow

Last Seen Users on Sotwe

Trends for you

Most Popular Users