Matt Figdore (@mattfigdore) profile picture

building @ivorymindai | 30 under 30 | Z Fellow | prev @speechifyai

NYC
Joined February 2021
Photo by Matt Figdore (@mattfigdore)Photo by Matt Figdore (@mattfigdore)Photo by Matt Figdore (@mattfigdore)Photo by Matt Figdore (@mattfigdore)Photo by Matt Figdore (@mattfigdore)Photo by Matt Figdore (@mattfigdore)
Matt Figdore retweeted
Life is beautiful. Do not miss it. Do not take it for granted. Do not tell yourself something is missing, as you will always find something to look for. The thing you are looking for is you.
74
469
53
2,260
609,168
للمهتمين في مجال البحث العلمي… أفضل أداه استخدمتها للقراءة الدقيقة 👇🏻 Coral AI
3
1
29
10,356
Matt Figdore retweeted
Worrying about wasting your life is one of the top 3 reasons I see “smart” people do nothing important. Up there with rationalizing something is beneath them when they really want that thing and believing they can rationalize a perfect move at all. Like life is some puzzle where they can comprehend how it all fits together.
Einstein wasted the second half of his life on a fruitless quest. In the second half of his life, von Neumann invented game theory, computer architecture, implosion nuclear weapons, cellular automata and weather prediction, among other things.
13
45
8
844
75,764
Will scaling reasoning models like o1, o3 and R1 unlock superhuman reasoning? I asked Gwern + former OpenAI/DeepMind researchers. Warning: long post. As we scale up training and inference compute for reasoning models, will they show: A) Strong general reasoning skills that work across most logically-bound tasks B) Some generalization to other logic tasks, but perhaps requiring domain-specific retraining or C) Minimal generalization? Finbarr Timbers (Artfintel, former DeepMind) "We'll achieve superhuman performance on specific tasks with verifiable rewards. I see no evidence for general broad transfer, but it seems extremely plausible." Gwern Branwen ""Everyone neglects to ask, what are we scaling?" Depends on what data they scale up on. The more you scale up on a few domains like coding, the less I expect transfer, as they become ultra-specialized. Far transfer is rare, and after a certain point, meta-learning collapses into learning for additional optimality. AlphaZero doesn't develop general gameplaying ability that transfers strongly across all domains either." Jacob Buckman (Co-founder at Manifest AI, former Google Brain) "Generalization can't really be predicted like that except empirically. All I know is as you add more compute and data you go from minimal transfer to some transfer to broad transfer. I have no clue where on that spectrum we stand when we run out of compute or data" Karthik Narasimhan (Sierra AI, former OpenAI, co-author of GPT paper) "I expect some generalization with domain specific retraining" Near (Independent) "I think the "spikiness" of intelligence will continue to be notable (models which are extremely good at some things yet quite 'dumb' at others), but that it is easy to improve generalization in the areas we care about, since it just requires some data/RL fun." Nathan Lambert (Post-Training Lead at Allen AI) "With chain of thought generation as a general approach for autoregressive language models to break down and process information, the new models trained to reason heavily about every subject will come to have better average performance than standard autoregression. In domains with explicit verifiers, this performance will be superhuman, in domains without, reasoning will still enable better performance, but maybe not more economical performance." Pengfei Liu (Shanghai Jiao Tong University) "Increased compute and inference time will drive reasoning capabilities to expert-level performance where rich feedback loops exist. However, the development of general reasoning will be gated by two factors: the availability of problems requiring genuine deep thinking, and access to high-quality expert cognitive process data or well-defined reward signals." Ross Taylor (Previously led Meta AI reasoning team) "I think general reasoning will come fairly quickly. Right now it's easier to scale in domains where problems are easy to verify with an external signal. The generalisation will come if models themselves become good verifiers across domains." Shannon Sands (Nous Research) "There's at least some generalisation to other tasks like logic puzzles - but it might require more domain specific training to improve on many more out of domain tasks." Steve Newman (Co-founder of AI Soup, Co-founder of Writely aka Google Docs) "This is a trillion dollar question. If I had to guess: we'll see some transfer of reasoning skills across domains, but (on anything resembling current architectures) some specialized training will be needed in each domain. We'll learn a lot one way or another this year." Tamay Besiroglu (Co-founder of Epoch AI) "I think minimal transfer is wrong because reasoning is a very general skill that you can apply to perform a wide range of actions. Planning, for instance, is something that requires good reasoning, and is useful to achieve high levels of performance across a wide range of tasks. Differentiating between broad transfer and some transfer is hard without being more precise or quantitative" Teortaxes (Independent) "I think there will be a period of strong 'natively verifiable reasoning overhang' which translates to more general verifiers using models' strong coding ability and general knowledge+tools, then they grok more general regularities of sound reasoning, and the next generation can natively generate good reasoning data for all domains." Xeophon (Independent) "We will see some generalizations into other domains where the model was not explicitly trained on. For example, R1 writes better and more creative stories than V3, the model it is based on. To push this further, models need to be trained on more data in other domains." Chris Barber (Independent, Creator of this roundup) Synthesis: "The expert takes point to generalization for all logically bound domains where we can construct verifiers for now, and trending in the direction of broad transfer in the future." Implications and follow-up questions Gwern "The important question is to what extent the verifiers or judge models can be cheaply set up for each new domain, I'd think. If you can quickly 'o3-ize' every important domain, particularly with general-purpose coders & mathematician expert models, then the scaling can continue cheaply." Gwern's recent comment on LW and McLaughlin's essays are also great reads. Ross Taylor "Verification will become a test-time compute problem -> the longer you spend checking a solution, the more accurate the verification signal should be. Then the question will be: how do you verify the verifiers? I suspect we'll end up in a world where multiple agents are checking each others' work (a bit like how the research community works). Not clear if we're at the level of model capability for this to work yet, but I wouldn't bet against it!" A Separate Question - Is There a Ceiling? Finbarr Timbers "I see no reason for reasoning models to plateau around world-expert level tbh. The existing paradigm for reasoning models looks like RLVR (reinforcement learning with verifiable rewards), where the model learns to solve tasks with deterministic/verifiable rewards. No reason that has any limit around human levels. Look at deepseek r1, for instance." Jacob Buckman "In general search can surpass humans. This was clear even pre-deep-learning, e.g. with "classical" chess AI. But most real-world domains don't have the nice properties (cheap to simulate, clear success condition) that make them amenable to search. In a setting without this niceness, reasoning/search amounts to learning a human expert's "search function", and then unrolling it. It's possible to surpass the original expert using this technique, by unrolling for more steps than they do. The bottleneck in this case becomes rate-of-search, that is, you won't be able to improve performance faster per step of search than the expert." Tamay suggested a good follow up would be to "elicit ideas for experiments that people would expect to turn out one way conditional on “Weak transfer” and another if “Strong transfer” is correct" – let me know if you have ideas. Thank you to Amir Haghighat, Arun Rao, Ash Bhat, Avery Lamp, Charlie Songhurst, Connor Mann, Daniel Kang, Dhruv Singh, Eric Jang, Ethan Beal-Brown, Finbarr Timbers, Flo Crivello, Griffin Choe, Gwern, Herrick Fang, Jacob Buckman, James Betker, Jay Hack, Josh Singer, Julian Michael, Katja Grace, Karthik Narasimhan, Logan Graham, Matt Figdore, Mike Choi, Nathan Lambert, Nicholas Carlini, Nitish Kulkarni, Pengfei Liu, Rick Barber, Robert Nishihara, Robert Wachen, Rohit Krishnan, Ron Bhattacharyay, Ross Taylor, Shannon Sands, Spencer Greenberg, Steve Newman, Tamay Besiroglu, Teknium, Teortaxes, Tim Shi, Tim Wee, Tom, Tyler Cowen, and Xeophon. Upcoming posts: 1) Computer-use benchmark timeline predictions 2) Clips from researcher interviews (e.g. Finbarr Timbers and Jacob Buckman from the list above) 3) More takes on scaling reasoning models Questions: 1. What's your take? 2. Who else should I ask? 3. What follow up questions should I ask? 4. What's the even more important question that I should be asking instead?
40
98
26
683
152,765
Matt Figdore retweeted
it doesnt matter how smart you are it doesnt matter how many connections/how much wealth your family has it doesnt matter where you went to school if you arent willing to grind out as many hours as your body is capable of you will never accomplish anything truly extraordinary – at Florida, USA
72
125
14
1,412
201,447
everyone end of worlding about announcements today meanwhile cofounder and I frothing because this is exactly what we hoped for literally the best day of my life I can't wait to shill next week what we shipping right now
20
5
2
127
72,986
Matt Figdore retweeted
Just shipped! If you want to make TikToks for your brand, you can now generate influencer-quality video storyboards in seconds. Each storyboard has: - A full video script - Cut direction - Sample frame thumbnails - On-screen text suggestions - Caption Demo: (🔊 on)
34
18
3
259
101,262
Matt Figdore retweeted
We built an AI assistant that does your busywork in Google Sheets and Excel. It's being used by thousands to: - Write and rewrite text in batch - Extract and clean data - Categorize cells and SO much more 👇 Install at Numerous.ai
17
91
13
627
157,644
One of the best essays I've read on deciding what to work on
"The market you pick and the waves you ride are more important than your execution." Wrote some thoughts/reflections on picking markets, seeking waves, and why your market matters more than your hard work and execution. andrewrea.substack.com/p/rid…
2
1
24
8,904
This really works, highly recommend
Try starting the day (post sunlight viewing) w/1-3min of uncomfortably cold (yet safe) water exposure (shower or immersion to the neck). The research-supported benefits are significant & long lasting elevations in dopamine & adrenaline, & thereby elevated mood, alertness & focus.
2
17
7,142
I spent the last 2 weeks building the ultimate AI assistant for spreadsheets. It can: - Autofill entire sheets just based off context - Generate formulas and filter rows using text commands - Directly integrate GPT-3 into your sheets And more… Here are a few examples: (1/8)
145
707
160
4,612
1,143,496
Matt Figdore retweeted
Been thinking about this for 24 hours straight
12
35
1
328
Athena: a video-native search engine Visual Content Search: ✅ Transcription Semantic Search: ✅ and now, finally, a solid UI to show it off 🎉 not too shabby for a project that's only a week old!
9
10
4
87
Excited to announce Contrary NYC 🗽, a 4000+ sq ft community space in the heart of Flatiron, NYC. We’re building the definitive beacon for the most brilliant NYC technologists, thinkers and tinkerers.
39
13
7
538
Matt Figdore retweeted
the entirety of progression I think is traced to just thinking and caring less. JUST LET SHOTS FLY, HYPERINSTICTUAL
1
10
90
The real differentiator.
All businesses eventually realize they’re not in the selling stuff business they’re in the recruiting people business.
1
4
Ninety percent of success can be boiled down to consistently doing the obvious thing for an uncommonly long period of time without convincing yourself that you're smarter than you are.
48
797
47
4,901
Self image is all inner dialogue. Stop listening to urself and start talking to urself instead Every story you tell yourself in ur head has a drastic impact on who you become
Replying to @brutedeforce
What would you tell a young cat to improve his self image
5
118
6
728
The world belongs to those who can keep doing without seeing the result of their doing.
75
329
27
1,985