Sean Grullon

@hulu_support The live stream just cut off nearing the end of the broadcast

Our team has identified the issue and users affected should be able to log back in again soon. We apologize for the inconvenience.

@hulu_support And now - the stream cut off right when the Oscars are nearing the end

Thanks so much for hanging in there! Our team took the necessary steps to resolve this, so you should be all set after rebooting your device. We appreciate your patience!

@elonmusk “Free” as in trying to bribe voters in PA. Pathetic hypocrite.

My involvement in politics is not because I wish to be, but because I believe the future of the free world is at stake

People have been rewriting history and saying that "everyone has always believed that LLMs alone wouldn't be AGI and that extensive scaffolding around them would be necessary". No, throughout most of 2023 (the "sparks of AGI" era) the mainstream bay area belief was that LLMs were *already* AGI, and that merely scaling their parameter count and training data size by ~2 OOM without changing anything else would lead to super-intelligence.

@ylecun @jmeierX Lean left by American standards... I don't have good data on this, but from my personal experience scientists lean center by European standards.

1. I also know Right-leaning scientists. But the data is clear: people with a postgraduate education (not just scientists) generally lean Left (see chart below). 2. Scientists are accustomed to ignoring their biases and prejudices when collecting evidence and testing a hypothesis. They are immune to misinformation, but they are trained to minimize the influence of their biases, and rely on evidence, to arrive at a good model of the world.

People studying misinformation lean left for two reasons: 1. scientists lean left, regardless of specialty, because they care about facts. 2. misinformation today primarily comes from the Right ("they're eating the dawwwgs!") which makes it worth studying and fighting against for people leaning left.

Yep. Strawberry is good work just not a breakthrough. Can we stop talking nonsense and actually address difficult problems in the field?

This is also one reason why serious ML researchers looked down their nose at deep learning for so long. No fancy theorems, no complicated, esoteric math. Just clip your gradients and set your weights to 0.8 and you can't fail to dominate.

BREAKING: Inflation now back to normal historical level of 2% According to latest calculation of underlying inflation by New York branch of the Federal Reserve Only the second time in American history that the government has alleviated inflation without causing a recession

If you’re registered for #ICML virtually, here’s the livestream link for our mechanistic interpretability workshop that’s about to start: icml.cc/virtual/2024/w…

⚡️ Excited to share that I am starting an AI+Education company called Eureka Labs. The announcement: --- We are Eureka Labs and we are building a new kind of school that is AI native. How can we approach an ideal experience for learning something new? For example, in the case of physics one could imagine working through very high quality course materials together with Feynman, who is there to guide you every step of the way. Unfortunately, subject matter experts who are deeply passionate, great at teaching, infinitely patient and fluent in all of the world's languages are also very scarce and cannot personally tutor all 8 billion of us on demand. However, with recent progress in generative AI, this learning experience feels tractable. The teacher still designs the course materials, but they are supported, leveraged and scaled with an AI Teaching Assistant who is optimized to help guide the students through them. This Teacher + AI symbiosis could run an entire curriculum of courses on a common platform. If we are successful, it will be easy for anyone to learn anything, expanding education in both reach (a large number of people learning something) and extent (any one person learning a large amount of subjects, beyond what may be possible today unassisted). Our first product will be the world's obviously best AI course, LLM101n. This is an undergraduate-level class that guides the student through training their own AI, very similar to a smaller version of the AI Teaching Assistant itself. The course materials will be available online, but we also plan to run both digital and physical cohorts of people going through it together. Today, we are heads down building LLM101n, but we look forward to a future where AI is a key technology for increasing human potential. What would you like to learn? --- @EurekaLabsAI is the culmination of my passion in both AI and education over ~2 decades. My interest in education took me from YouTube tutorials on Rubik's cubes to starting CS231n at Stanford, to my more recent Zero-to-Hero AI series. While my work in AI took me from academic research at Stanford to real-world products at Tesla and AGI research at OpenAI. All of my work combining the two so far has only been part-time, as side quests to my "real job", so I am quite excited to dive in and build something great, professionally and full-time. It's still early days but I wanted to announce the company so that I can build publicly instead of keeping a secret that isn't. Outbound links with a bit more info in the reply!

Excited to go to ICML! Come say hello in Vienna...

Yesterday, a team of PathAI scientists published “Interpretability analysis on a pathology #foundationmodel reveals biologically relevant embeddings across modalities.” 📰Read the preprint here: arxiv.org/abs/2407.10785 #pathology #PathTwitter #AI #deeplearning

@andrewgwils Completely agree - but as a recovering physicist it was significantly easier to get a job in AI and make a living in it than it was physics.

Physics is cooler than AI. Gravitational time dilation? The twin paradox? Quantum entanglement? A chatbot just doesn't compare...

To capture both high and low-level details of biological objects across resolutions, we modify the #DINOv2 training by adding MAE decoder and a Fourier reconstruction loss. This ‘flexified’ training operates on masks of different sizes

PLUTO - our Pathology #foundationmodel specifically designed to be performant, compact, and generalizable. It's been an honor to help the guide the PLUTO direction. Our manuscript can be found here: lnkd.in/eHqtBBqW

We integrate the #flexivit approach for backbone training. FlexiViT is esp useful here - it enables us to customize the patch size based on the scale of biological entities we want to capture (small for nuclei, large for glands) without a drop in performance

Today, we shared our pre-print describing our new #FoundationModel PLUTO (PathoLogy Universal TransfOrmer) to fuel #drugdevelopment & #CancerResearch for fast, efficient and powerful representation of #pathology images Read the Paper: arxiv.org/abs/2405.07905

@akshay_pachaar That’s a misleading and silly analogy.

The new e=mc**2

I did a double take when I saw this. There are caveats, but this is still a remarkable thing to consider.

# on shortification of "learning" There are a lot of videos on YouTube/TikTok etc. that give the appearance of education, but if you look closely they are really just entertainment. This is very convenient for everyone involved : the people watching enjoy thinking they are learning (but actually they are just having fun). The people creating this content also enjoy it because fun has a much larger audience, fame and revenue. But as far as learning goes, this is a trap. This content is an epsilon away from watching the Bachelorette. It's like snacking on those "Garden Veggie Straws", which feel like you're eating healthy vegetables until you look at the ingredients. Learning is not supposed to be fun. It doesn't have to be actively not fun either, but the primary feeling should be that of effort. It should look a lot less like that "10 minute full body" workout from your local digital media creator and a lot more like a serious session at the gym. You want the mental equivalent of sweating. It's not that the quickie doesn't do anything, it's just that it is wildly suboptimal if you actually care to learn. I find it helpful to explicitly declare your intent up front as a sharp, binary variable in your mind. If you are consuming content: are you trying to be entertained or are you trying to learn? And if you are creating content: are you trying to entertain or are you trying to teach? You'll go down a different path in each case. Attempts to seek the stuff in between actually clamp to zero. So for those who actually want to learn. Unless you are trying to learn something narrow and specific, close those tabs with quick blog posts. Close those tabs of "Learn XYZ in 10 minutes". Consider the opportunity cost of snacking and seek the meal - the textbooks, docs, papers, manuals, longform. Allocate a 4 hour window. Don't just read, take notes, re-read, re-phrase, process, manipulate, learn. And for those actually trying to educate, please consider writing/recording longform, designed for someone to get "sweaty", especially in today's era of quantity over quality. Give someone a real workout. This is what I aspire to in my own educational work too. My audience will decrease. The ones that remain might not even like it. But at least we'll learn something.