Gabe Lozano

Learning by building.

[Cryptic Swarm @CrypticSwarm]

Software engineering is phoenixing.

Companies are downsizing.

Agent swarms are promising to replace entire engineering teams.

And yet, the best engineers are busier than ever.

[Andrej Karpathy @karpathy]

My pleasure to come on Dwarkesh last week, I thought the questions and conversation were really good.

I re-watched the pod just now too. First of all, yes I know, and I'm sorry that I speak so fast :). It's to my detriment because sometimes my speaking thread out-executes my thinking thread, so I think I botched a few explanations due to that, and sometimes I was also nervous that I'm going too much on a tangent or too deep into something relatively spurious. Anyway, a few notes/pointers:

AGI timelines. My comments on AGI timelines looks to be the most trending part of the early response. This is the "decade of agents" is a reference to this earlier tweet https://t.co/NiSn6jftqq Basically my AI timelines are about 5-10X pessimistic w.r.t. what you'll find in your neighborhood SF AI house party or on your twitter timeline, but still quite optimistic w.r.t. a rising tide of AI deniers and skeptics. The apparent conflict is not: imo we simultaneously 1) saw a huge amount of progress in recent years with LLMs while 2) there is still a lot of work remaining (grunt work, integration work, sensors and actuators to the physical world, societal work, safety and security work (jailbreaks, poisoning, etc.)) and also research to get done before we have an entity that you'd prefer to hire over a person for an arbitrary job in the world. I think that overall, 10 years should otherwise be a very bullish timeline for AGI, it's only in contrast to present hype that it doesn't feel that way.

Animals vs Ghosts. My earlier writeup on Sutton's podcast https://t.co/rSp1noyGBr . I am suspicious that there is a single simple algorithm you can let loose on the world and it learns everything from scratch. If someone builds such a thing, I will be wrong and it will be the most incredible breakthrough in AI. In my mind, animals are not an example of this at all - they are prepackaged with a ton of intelligence by evolution and the learning they do is quite minimal overall (example: Zebra at birth). Putting our engineering hats on, we're not going to redo evolution. But with LLMs we have stumbled by an alternative approach to "prepackage" a ton of intelligence in a neural network - not by evolution, but by predicting the next token over the internet. This approach leads to a different kind of entity in the intelligence space. Distinct from animals, more like ghosts or spirits. But we can (and should) make them more animal like over time and in some ways that's what a lot of frontier work is about.

On RL. I've critiqued RL a few times already, e.g. https://t.co/mYrMFVdVDW . First, you're "sucking supervision through a straw", so I think the signal/flop is very bad. RL is also very noisy because a completion might have lots of errors that might get encourages (if you happen to stumble to the right answer), and conversely brilliant insight tokens that might get discouraged (if you happen to screw up later). Process supervision and LLM judges have issues too. I think we'll see alternative learning paradigms. I am long "agentic interaction" but short "reinforcement learning" https://t.co/2L7FiaoKsw. I've seen a number of papers pop up recently that are imo barking up the right tree along the lines of what I called "system prompt learning" https://t.co/df5mJDdN3C , but I think there is also a gap between ideas on arxiv and actual, at scale implementation at an LLM frontier lab that works in a general way. I am overall quite optimistic that we'll see good progress on this dimension of remaining work quite soon, and e.g. I'd even say ChatGPT memory and so on are primordial deployed examples of new learning paradigms.

Cognitive core. My earlier post on "cognitive core": https://t.co/q2s1ihGy0T , the idea of stripping down LLMs, of making it harder for them to memorize, or actively stripping away their memory, to make them better at generalization. Otherwise they lean too hard on what they've memorized. Humans can't memorize so easily, which now looks more like a feature than a bug by contrast. Maybe the inability to memorize is a kind of regularization. Also my post from a while back on how the trend in model size is "backwards" and why "the models have to first get larger before they can get smaller" https://t.co/6k0FZRGXsb

Time travel to Yann LeCun 1989. This is the post that I did a very hasty/bad job of describing on the pod: https://t.co/fQgqaXPyp6 . Basically - how much could you improve Yann LeCun's results with the knowledge of 33 years of algorithmic progress? How constrained were the results by each of algorithms, data, and compute? Case study there of.

nanochat. My end-to-end implementation of the ChatGPT training/inference pipeline (the bare essentials) https://t.co/SIetgyoKWN

On LLM agents. My critique of the industry is more in overshooting the tooling w.r.t. present capability. I live in what I view as an intermediate world where I want to collaborate with LLMs and where our pros/cons are matched up. The industry lives in a future where fully autonomous entities collaborate in parallel to write all the code and humans are useless. For example, I don't want an Agent that goes off for 20 minutes and comes back with 1,000 lines of code. I certainly don't feel ready to supervise a team of 10 of them. I'd like to go in chunks that I can keep in my head, where an LLM explains the code that it is writing. I'd like it to prove to me that what it did is correct, I want it to pull the API docs and show me that it used things correctly. I want it to make fewer assumptions and ask/collaborate with me when not sure about something. I want to learn along the way and become better as a programmer, not just get served mountains of code that I'm told works. I just think the tools should be more realistic w.r.t. their capability and how they fit into the industry today, and I fear that if this isn't done well we might end up with mountains of slop accumulating across software, and an increase in vulnerabilities, security breaches and etc. https://t.co/8556ESSpyY

Job automation. How the radiologists are doing great https://t.co/FVUI872dkD and what jobs are more susceptible to automation and why.

Physics. Children should learn physics in early education not because they go on to do physics, but because it is the subject that best boots up a brain. Physicists are the intellectual embryonic stem cell https://t.co/p72Elk8lPV I have a longer post that has been half-written in my drafts for ~year, which I hope to finish soon.

Thanks again Dwarkesh for having me over!

[Gabe Lozano @gabelozano]

@BrentBeshore Thanks, bud. Grateful for you as well. ❤️

[Greg Brockman @gdb]

Perhaps the single most important virtue in ML engineering is persistence. The ML engineering process is one of repeatedly checking & understanding every detail of the system, until it finally goes through a phase transition from "not working at all" to "working shockingly well".

[Gabe Lozano @gabelozano]

@fchollet Build a company, then report back.

[Gabe Lozano @gabelozano]

"Trusting yourself to be patient enough to wait for the optimal opportunity requires a deep structural understanding of constraints, resources, capabilities, and potential outcomes."

Exceptional post on patience by @BrentBeshore and his team: https://t.co/CEuxhVYK16

[François Chollet @fchollet]

A clear sign that you're wrong about something is if you display incuriosity. If you come across information that seems to contradict what you believe, the logical move is to try to learn more about it and get to the bottom of it, not outright reject it and plug your ears.

[MIT CSAIL @MIT_CSAIL]

"Programs must be written for people to read, and only incidentally for machines to execute." ― MIT professor Harold Abelson

[Paul Graham @paulg]

To get startup ideas, it's necessary to have fairly deep expertise in some domain. The idea is usually in the same domain, but it doesn't have to be. A startup idea can also consist of importing insights from one domain to another.

[Steven Strogatz @stevenstrogatz]

⁦A long yet gracefully written blog post by @stephen_wolfram⁩ about his latest thinking. Very ambitious, very provocative, and to me at least, very interesting. See what you think. https://t.co/XRytKUBIZC

[Reid Hoffman @reidhoffman]

I believe starting a company is like jumping off a cliff and assembling a plane on the way down -- your willingness to jump is your most valuable asset as an entrepreneur.

[Gabe Lozano @gabelozano]

🤣

[Peter Wang 🦋 @pwang]

My tensors bring VCs to the yard
And they’re like, “The math is too hard!”
Damn right, the math is too hard
I can teach them, but I’d have to charge

[hattie @oh_that_hat]

Constraint is fundamental to intelligence.

With limited capacity, we are forced to develop "understanding" in order to simplify the amount of signals we need to analyze.

You can have a model with infinite capacity and perfect predictions but no understanding.

[Matt Crisp @EnergyCrisp]

Great news this morning on stockholder vote. NYSE here we come! $STPC -> $BHIL
https://t.co/4XHk9dHzQ0

[Nerdy @nrdyinc]

A proud moment for Nerdy, live #learning and the #education industry. We rang the Opening Bell at the @NYSE this morning to celebrate $NRDY going public. (📷: NYSE)

[Amjad Masad @amasad]

Success doesn't seem to be attainable without massive stress tolerance. Perhaps it's the number one required attribute.

[Brian Greene @bgreene]

"A man really dies when he loses the power to take in a new idea."--George Orwell, born OTD, 1903.

[Gabe Lozano @gabelozano]

@AbePalmer With most noise filtered out, the daily volume of unread emails hitting my inbox tends to be rather low and always manageable, even when the sender doesn’t know or care to use my preferred consumption format.

[Gabe Lozano @gabelozano]

@AbePalmer e.g., I use filters to automate (1) trashing emails I don’t want to see; (2) archiving emails that don’t need to be read upfront, but that I want indexed for future searches; and (3) marking as read, but leaving in the inbox, emails that I want to read in batches, as time allows.