This guy is a genius.

super chad

Clearly and unquestionably uncreative, by definition. Ilya lies

Elicit did mot picj uo my paper. Google schilar did. You know what Max Tegmark is proposing for AI safety. Extract the program from the AI model and then prove the program will not do certain things in a provabke manner. And then release the AI. I did do a gradcam project on explainable ai.

The World's biggest AI Brains are in a single frame

26:18 -- Those results are highly suspicious. I suspect there's a confirmation bias in the human scores. The workers don't want to lose their qualifications, and so won't perform tasks that are too difficult. So there may be a drop off in participation or submission if a human worker feels like they may be in error for the more complex tasks. Furthermore, the human workers were given "room to think", where the prompting of the LLMs suggest they were not. I suspect allowing GPT-4 to use step-by-step reasoning would improve its score across the board. And dramatically more so if its allowed to create a python script to solve the problem.

Just binge watch. Public. No one shall kick us out of the Tegmark paradise😊I will watch in VR. 🎉

Math, cannot afford to have such jokes..irrelevant, . Because, correlation and causations are only sort, especially if you think...in epidemiology and many diseases people read this only to get that 'core' really minded right

The AI understanding of the world is different from Human being.

27:44

It would be interesting to see the difference between LLMs trained on non-fiction, realistic fiction, and and fantasy.

it could break my neurons. 😅

The tasks where llm fail are either useless (not trained on many data), or based on vision capabilities (where are a lot worst yet these systems)

Really good talk :)

So why did her methodology fail to work here? So we have to go to the basics, because you simply can't skip them: 1) Humans are repetition machines, they repeat and recombine their experiences, you can see it ESPECIALLY in the arts, but there we call it inspirations, humans take "inspirations" from their life experiences, and recombine them. 2) AI is the same, they too are repetition machines that recombine experiences, but their experiences are different from humans'. 3) Hence, for comparison, you may not test humans on subjects that they have not experienced, and for AI you may not do the same. However her entire testing methodology was based on testing on experiences that only humans had. Basically, she almost got it when she said that a child learns to wear socks under the shoes by his experiences, but then did not narrow her tests to be based on common experiences of AI and humans, rendering them a curiosity of translation, but not of understanding.

Thanks.

Lecun's too overrated in AI community. 🤭

22:35 - how many of those tasks require knowledge of the exact spelling of the word? - LLM's are only passed the encoded tokens, and may not be aware of word spelling in a way that allows eg: acronyms?

In my opinion understanding is actually pretty clear. In humans, a very useful skill in language acquisition is circumlocution, or referring to a concept without using the name of it. Now, for a true LLM it's possible that could be done with regurgitation of training data directly (what are the odds that some common turns of phrase show up on Wikipeda? Or that a dictionary would find its way into a dataset?), but in a multi modal LLM I think the ability to verify similar patterns of neuron activity for analogous inputs across modalities is pretty indicative of strong understanding and generalization. In other words, I think the strength of understanding can be measured as the number of unique inputs that can implement a given pattern of behavior in the FFN, or lead to the same or similar output in the language head, roughly.

Mitchell is great! Love her work. But Lecun's reckless, self-serving comments should not be elevated so high. It's like a TV news program hosting a flat-Earther to give both sides of the story.

It seems like tokenization would really fuck with one's ability to do some of these tests. Like, I don't know how much the format of something like that would even be preserved? Also, makes me wonder how much the lack of physical 3d movement data/training would impact some of these reasoning tasks. Like, you can even notice in her language use about concepts and stuff how much "spatial" reasoning is involved. It seems like to do one of these tests fairly, you would have to completely sort of homogenize the way the test taker would be taking it? It brings to mind like, disabilities, you know? You wouldn't expect someone that was born not only blind, but completely unable to process visual data in any way we would recognize to be able to solve visual tasks necessarily, unless they generalize well to whatever mediums they do know.

lol! the difference in performance between the humans doing the test for free and the people being paid is about the same jump in performance you get from chatgpt when you just give it a prompt vs if you tell it "i'll give you $20 if you do a good job" xD

can you post discussion?

Show meeee

سلام بر اندیشمندان جهان در دین من که رشته ای از تمام ادیان الهی است خداوند در قرآن می فرماید یک لحظه فکر کردن و تفکر سالم و بدون تعص و با فکر باز بهتر است از هزار سال عبادت لطفا از شما خواهش میکنم نامه ای که آقای خامنه ای برای جوانان غرب نوشته را با عنوان (نامه ای برای تو) را بخوانید. بدون تعصب فکر گنید. اگر برای شما جالب بود در مورد مذهب ایشان کتاب های ایشان اقای خامنه ای تحقیق کنید عزیزان من، ما با زیست حیوانی به دنیا مي آییم دنیا را در موقعیت جغرافیایی خود بگونه ای میبینیم که همه برای بقای خود تلاش می کنند . لی کسانی بعد از مرگ هنوز نام آنها زنده است که برای بقای ناتوان ها و کم توٱن ها تلاش میکنند جاودان است. ای دوست من تسلیم این دنیای خیالی نشو که پر ان همه به دنبال سکس پول و مقام هستند نشو. برای هدف از خلقت خود ثقت بگذار که چرا یک موش نمیتواند بمب هسته ای بسازد ولی انسان میتواند، پس انسان متفاوت است. و برای چیزی مقدس تر از آنچه دیده می شود در این دنیا پا نهاده انسان به گفته خداوند بهترین مخلوق خداوند است پس نباید در حد نیاز های حیوانی سقوط کند انسان قطره ای از دریای خداوند است که ا تنهایی هیچ نیست ولی وقتی به دریا می رسد تبدیل به خود دریا می شود از شما عاجزانه خواهش می کنم کمی فکر کنید،،، خداوند با آغوش باز منتظر توست اگر باور نداری آغوش ات را به سوی خداوند باز کن تا ببینی چگونه تورا به جایی می برد که تمام طلا های دنیا را تکه سنگی بی ارزش ببینی،،، 🙏🙏🙏🙏 کمی تفکر کنید الله# شیعه# خامنه ای# ❤❤❤❤❤❤❤❤❤❤❤❤❤

An awesome discussion. These kind of panels where expert condensus and debates are exchanged would be great at staying grounded on AI.

25:06 I consider myself to be reasonably intelligent, but I am absolutely stumped by Problem No. 1. How are you supposed to evaluate the three blocks of letters below the alphabet? Are the two blocks on the first line supposed to serve as an example? Are you supposed to consider all three blocks together? Does the order of the blocks matter? I suspect that there is something implied here that I am missing.

Melanie Mitchell, ILY <3

EXCELLENT insights! Thank you so much for sharing!

Such a great explanation

Nó không hiểu được tiếng nào

literally the best explaination!!!

I wonder if Patrick is referring to Toba at 1:45:44

AI playing practical joke

Is nocomment coded the same as null? Zerotolance?

the fly has a different system of digestive when she transfrome all the things she touches. I think that her mouth they invented a stethoscope it just look like her mouth she has role, like others, the insects, and the organisms in the environment in the circle of nature.,

Flyologist

Little did Turing know, the infant brain starts with a bizarrely over-connected brain and learns by cutting the unnecessary connections, so adults are actually simpler.

hOLLY FUCKING JESUS THE LEVELOF DETAIL IS INSANE!!!

One day we will invent an AI powerful enough to record a lecture that is watchable to the human eye!

Why are the interesting parts cut out?!

Thanks. I'll be able to use this info someday

Yall should let an avg curious citizen like me ask the panel questions. So more people can learn from this. I volunteer

ilya geoffrey and demis all on a call is actually such an absurdly must watch

How to get the information out of this video: Just below the vid, there is "Abstract: ...more". Click that and you can read the story: just the abstract, or the whole script. 😁

Counted the manually and with no computer program? Damnn that really reliable doc 😂😅 kidding Sweet work I love dev bio

I don't know how I got here. But I like the information I acquired

Explore exploit the control response; that’s why we began our project with the history of the Generation X project and our hypothesis is one that is focused on activating the programed response

That is awesome!

brilliant lectures, the whole series is