Matt Beane
@mattbeane
Followers
4K
Following
11K
Media
249
Statuses
7K
Study work with intelligent machines, esp. robots. @MITSloan PhD, @Ucsb Asst Prof, @Stanford fellow, @tedtalks Book, Substack: https://t.co/OJQa0by9MT
Santa Barbara, CA
Joined November 2008
I'm so grateful and exited to share that my book will be published on June 11th!. It's called "The Skill Code: How to Save Human Ability in an Age of Intelligent Machines". Available for preorder now, here's a bit of the story:.
10
33
198
In a post-truth landscape, a difference in quantity amounts to a difference in quality. Careful thinking about this will produce a better model to predict diffusion and its consequences, imo.
0
0
1
A smeared and intensified interpretation landscape is a unique feature of genAI diffusion relative to other general purpose technologies. It is both caused by build/marketing/deployment/use of the tech and shapes it. Yes, there have been naysayers and enthusiasts before. But:.
1
1
2
RT @lugaricano: 1/@EpochAIResearch doubles down on preiction AI will drive 20%+ annual GDP growth. Economists remain skeptical. This is the….
0
185
0
RT @The_Maintainers: 🎉 We're excited to announce the release of The Maintainers Study Guide, a resource to help ground yourself in the prin….
0
9
0
A *great* potential move that could help with skilling/deskilling related to genAI. Eager to test it out.
ChatGPT is rolling out a new mode called “Study Together” 👀. Instead of giving you answers, it acts like a tutor - asking guiding Qs and walking through problems step by step. Feels like a big step towards personalized learning (if it works!)
0
3
7
RT @msbernst: Thank you to everyone for your energy and enthusiasm in joining this adventure with me so far!
0
13
0
I don't much tweet about robotics these days - all my research is focused on LLMs - but these robots (made by @Dyson) are promising and innovative. Kudos!.
0
1
1
RT @goodside: ChatGPT o3-pro identifies a 1965 quote by I. J. Good hand-written in a mix of print and cursive on a note ripped into four st….
0
161
0
This article by @GeoffreyHuntley is *brilliant*. Highly engaging, too. At the same time, I think it underestimates the risks of suboptimal genAI use in SWE. Working papers coming soon, but we show small self-owns are far more common than the (very real) 10x-ing Geoff highlights.
@mattbeane @GeoffreyHuntley gave a very interesting talk about that recently
0
2
11
First consumer-grade table saw: 1928. First SawStop (auto-stop of blade on human flesh contact): 2004. Deep research o3 says that from 1928-2004:.-ED-treated table-saw injuries ≈ 1.6 million.-Finger/hand amputations ≈ 170 000. How many SWEs are injuring themselves while vibing?.
Quitting programming as a career right now because of LLMs would be like quitting carpentry as a career thanks to the invention of the table saw.
1
0
5
First wind-powered table saw: 1777. First consumer-grade table saw: 1928. TONS of time for us to adapt. First LLM paper: 2017. First consumer-grade genAI release: 2022. Today: 2024. No time. So Simon's right conceptually, likely *very* wrong wrt humanity's adaptive capacity.
Quitting programming as a career right now because of LLMs would be like quitting carpentry as a career thanks to the invention of the table saw.
0
0
1
RT @simonw: Quitting programming as a career right now because of LLMs would be like quitting carpentry as a career thanks to the invention….
0
1K
0
Amen!. Default AI design, implementation, and use are leaving too many of us with less valuable skill. But systematic positive exceptions are here, now. Have been since I started studying this in 2011. Now, more than ever, we must make AI part of the skilling solution.
7/ We can use these tools to help people climb the of work, by embedding intelligence into the tools themselves. We can lower the cost of learning, speed up the acquisition of expertise, and build systems where more people do high-leverage work.
0
0
1
RT @MITSloan: These research-backed titles cover retirement transitions, artificial intelligence and skills development, how to build an in….
0
3
0
RT @timoreilly: Henry Farrell's rant about how science fiction has led Silicon Valley astray is well worth your time. Who knew that Hugo Ge….
0
7
0
Honored that is on MIT Sloan's summer reading list!.
From life after work to working with artificial intelligence, MIT Sloan researchers tackle our moment’s big questions in these eight books.
0
1
11
WHOA, @AmazonMGMStudio you better not have messed this up because I am now officially excited:
0
0
1
Okay this is marvelous, crafty, and subtly powerful! The hallucinations section was by far the most entrancing. I won't spoil it, but everyone should at least read that. We REALLY don't understand these models on a deep level, especially in complex, long-chain interaction. Wow.
New Anthropic Research: Project Vend. We had Claude run a small shop in our office lunchroom. Here’s how it went.
1
2
12
Strongly recommend this paper. Innovative methods and useful conclusions. Of course much to do, but this helps us think more precisely about the impact of LLMs on the labor market.
(1/4) When some job tasks are automated, do the tasks that remain become more or less valuable? 🧵👇. In a new working paper, @davidautor and @ProfNeilT argue the answer depends on how much expertise is required for the tasks still done by humans.
0
0
4
Quite something for a startup to burn more dollars on tokens than wages.
2
0
8