Point #1 was raising alarm bells on my mind, too naive, so thank you for asking the question.
Cervera
Karma: 68
What’s up with the weird units for money?
I dont think any of that invalidates that Gwern is a usual, usually right.
I dont think this is a good take.
The Cybertruck does not break on that pull. It breaks on this one: 0:27
“We’ll be building a cluster of around 22,000 H100s. This is approximately three times more compute than what was used to train all of GPT4.
This bothers me. It’s a naive way of seeing compute. It’s like confusing Watts and Watt-hours
22,000 H100s is three times the amount of FLOP/s than what was used to GPT-4, so you could train it in 3x less time, of with 1⁄3 of your cluster and the same time.
I think this view of looking at compute helps making naive asumptions about what this compute can be used to. And FLOP/s are not a perfect unit for normal discourse when we’re at x10¹⁵ scales.
If ancestor is parent/mother/grandparent etc but nothing else. Obviously non hunters.
If we count how many people dead or alive are you related to. Farmers.
Better than 90% of amateur fiction posted.
This gap will only widen over time; China is failing to develop a domestic semiconductor industry, despite massive efforts to do so, and is increasingly cut off from international semiconductor supply chains.
I would say this is a falsehood
The US export ban on Controlled GPUs has really made china push for local semiconductor manufacturing way, and accelerate their projects, they dont have 5nm TSMC quality wafers, fine, but they’re developing the full stack.I mean if this was a “The AGi Race Between the US and Russia doesnt exist” okay fine, but Seeing how more than half the papers that land in ArXiv have chinese authors in them, plus the whole China does 90% of electronic manufacturing in the world. I dont understand how you come to the conclusion that china is hopelessly dead in the water.
The day the US export ban on GPU happened, okay, most of us really wondered, but seeing how they’re operating 6 months to a year afterwards, it just obvious that they will be able to make it happen.
I dont think the difficulty of the task has much with the outcome.
I mean, I take your comment at face value and update to “it’s going to get powerful faster” and not the other way around.
A “moonshot idea” I saw brought up is getting Yudkowsky’s Harry Potter fanfiction translated into Chinese (please never ever do this).
Can you expand on this? Why would it be a bad idea? I have interacted with mainland chinese people (outside of china) and I’m not really making the connection.
How do you think discharge rates would affect the battery? Would it behave like a LFP that basically outputs mostly the same rate of Amps in the full spectrum of charge (Voltage varies little with discharge %)
Do you think an approach like this generates bateries with long lifetimes?
Did you expect balancing of these particular batteries to be particularly complicated?
Basically, tell us more!
Well, Evals and that stuff OpenAI did with predicting loss could be a starting point to work in the tables.
But we dont really know, I guess that’s the point EY is trying to make.
Use tables for concrete loads and compare experimentally with the to be poured concrete, if a load its off, reject it.
We dont even have the tables about ML. Start making tables, dont build big bridges until you got the fucking tables right.
Enforce bridge making no larger than the Yudkowski Airstrike Threshold.
I appreciate the concept.
I wonder about how the hardware overhangs would look like after the moratorium ends or somebody bypasses it.
On first thought it doesnt look like a robuts solution, I assume a 15-20% improvement on Compute access per year. Would need to plot it against the moratorium treshold and see if over time one gets closer to the Yudkowski Airstrike treshold but I assume no, 20% vs 54%,,
I dont know! maybe this is a good idea.
I dont think I never learned something because it would make me a better worker / provide me with more economical resources when I was a child and was in need of tutoring, I got lucky to have a somewhat curious mind and I tried to saciate it.
Of course as an adult I choose to do things that are useful, overall, and that normally repercutes on being a human with skills that other people pay for. But the explicit bayesian calculation about knowledge and money is not one I tend to do, what interest me interest me. Of course when trying to learn something as an adult the friction of the subject is a marker that determines how likely I’m going to try to acquire the information, for example, I have read about Laser gyros, but the mountain of knowledge was too unsumournable to actually learn how laser gyros work, really.
If the LLMs can lower the friction I guess everybody will be more likely to learn things. Also, there is no “big mistery” in most fields, you just need a structured idea of a certain amount of concepts. Some of them more palatable than others. (I know what a sigmoid activation is, but I dont have the high school knowledge about the funcion very fresh on my mind) These tools could help with it.
(I tried to parse this comment on Bing writing tool to get a better output and it just came as more corporate) English is not my first language but this just feels worse.
I have always been curious about learning new things, regardless of their economic value or usefulness for my career. When I was a child and needed tutoring, I did not choose subjects based on how they would make me a better worker or provide me with more resources. Instead, I followed my interests and tried to satisfy them. Of course, as an adult, I also consider the practical aspects of learning something new, such as how it can benefit me professionally or personally.
But I do not usually make explicit Bayesian calculations about knowledge and money; rather, I learn what interests me. However, sometimes the difficulty of learning something can discourage me from pursuing it further. For example, I have read about laser gyros, but the amount of knowledge required to understand how they work was too overwhelming for me.
If there were tools that could lower the friction of learning new things, such as language models that could explain concepts in simple terms or provide structured overviews of different fields, I think everyone would be more likely to learn more things.
After all, most fields do not have “big mysteries” that are impossible to grasp; they just require familiarity with certain concepts and their relationships. Some of these concepts may be more intuitive than others (for instance, I know what a sigmoid activation is in neural networks, but I do not remember much about the function itself from high school math). These tools could help bridge these gaps and make learning easier and more enjoyable.
So first thoughts while reading the research/Gpt-4 page
- ChatGPT System Prompt open soon to users, not API holders, that’s going to be interesting.
-Only trained adversarialy with 50 experts? One would think you would spend a bit more and do it with 100x more people? Or if hard to coordinate, at least 500.Closing in on human performance, but I would love to see numbers on the compute needed to train. They can predict loss accurately, but what that loss means seems to be mostly an open problem.
Will need a couple days and rereads to digest this.
I did finish it, and was going to edit the original comment, I get the purpose of what you did here, thanks for the post, it’s an interesting read.
I plan on reading the whole post, but a thing struck with me on the first paragraph.
it’s not a “her”, it’s a “it”
Happened yesterday too in Spanish in another forum, using feminine pronouns to refer to ChatGPT (Terrible name) I guess it’s more “normal” in my native language because Intelligence has a feminine gender on my language. But seeing it in English really makes me notice.
I was trying to imply that the flow of water goes in a single direction. And that it’s a quite irrational fear to be sprayed by others fecal matter in this case.
I dont like the japanese style ones, at my home we have a hand mini showerhead that we direct ourselves. We clean it periodically and it rarely touch anybody’s skin. All the muslim world has that.
Could we not devise AlphaFold but for LLM alignment?
Your P/NP remark reminded me of the scepticism around Protein folder before the Alpha fold days.