Xin Chen @outformatics

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 x.com/karpathy/statu… 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 x.com/karpathy/statu… . 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. x.com/karpathy/statu… . 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" x.com/karpathy/statu…. 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" x.com/karpathy/statu… , 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": x.com/karpathy/statu… , 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" x.com/karpathy/statu… Time travel to Yann LeCun 1989. This is the post that I did a very hasty/bad job of describing on the pod: x.com/karpathy/statu… . 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) x.com/karpathy/statu… 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. x.com/karpathy/statu… Job automation. How the radiologists are doing great x.com/karpathy/statu… 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 x.com/karpathy/statu… 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!

Dwarkesh Patel @dwarkesh_sp

a year ago

The @karpathy interview 0:00:00 – AGI is still a decade away 0:30:33 – LLM cognitive deficits 0:40:53 – RL is terrible 0:50:26 – How do humans learn? 1:07:13 – AGI will blend into 2% GDP growth 1:18:24 – ASI 1:33:38 – Evolution of intelligence & culture 1:43:43 - Why self

535 3K 18K 10.8M 25K

Great touch on the forefront of AI robotics open.spotify.com/episode/6j0oHF…

@KevinNaughtonJr -

@aaditsh Free

How to communicate clearly? Majority of our time goes into emails, slack messages, and loads of meetings. Yet, as a student, no one taught me how to communicate well. (I must have annoyed a lot of people along the way ...😬) Here is a template for organizing your message.

We're introducing embeddings, a new feature of our API that distills relationships between concepts, sentences, and even code in a simple numerical representation — for more powerful search, classification, and recommendations. openai.com/blog/introduci…

Yuval Noah Harari: Lessons from a year of Covid | Free to read ft.com/content/f1b30f… via @financialtimes

Chinese American community works very hard on donating urgent PPEs to our hospital. They share what they have in labs/clinical offices/home, and seek supplies all around the world. Two groups delivered the supplies in the weekend. Now used at ER! #FightCOVID19 @StonyBrookMed

keyboard_arrow_left Previous keyboard_arrow_right Next

28 years ago today Linus Torvalds announced in a newsgroup that he was working on a free operating system that later came to be called Linux. bit.ly/2ZjgGtO #Linux #opensource #tdih #otd

"Because You Can Too"

keyboard_arrow_left Previous keyboard_arrow_right Next

@Waymo 👍👍

Today, we're launching our Waymo Open Dataset. This high resolution lidar and camera data has been collected by our self-driving cars across a diverse range of situations. We're excited to share it directly with the research community. Download now: waymo.com/open

This was a mind-blowing conversation with @harari_yuval -- latest episode of #TheTEDInterview. At the heart of it is a chilling question. Does the future need us? bit.ly/2JRTk5w Please share, comment, review, rate. Thanks all.

The Most Comprehensive Guide to K-Means Clustering you may ever need: analyticsvidhya.com/blog/2019/08/c… by @AnalyticsVidhya ———— #BigData #DataScience #Ststistics #MachineLearning #Algorithms #DataLiteracy

New blog post: The State of Transfer Learning in NLP A review of key insights and takeaways from our NAACL 2019 tutorial with updates based on recent work. ruder.io/state-of-trans…

keyboard_arrow_left Previous keyboard_arrow_right Next

Your salary is just your company’s monthly subscription of you

A nice blog post explaining recent progress in NLP and translation and some Facebook AI's contributions. BERT, RoBERTa, backtranslation, GLUE, SuperGLUE, WMT explained.... ai.facebook.com/blog/new-advan…

Three scaling breakthroughs for NLP: fastest BERT-Large training (under one hour), fastest BERT inference (2.2ms on T4), and largest Transformer (GPT-2 8.3B). Code is open source. devblogs.nvidia.com/training-bert-…

I get many good questions that I answer on a daily basis on social media so I have been sharing the questions and answers weekly in this format so you and others can see them all: linkedin.com/posts/raydalio…