p1esk

16 ms·

301.
by p1esk 1y ago
Exactly. There’s one guy - Ray Kurzweil - who predicted in late 90s that AGI will happen in 2029 (yes, the exact year, based on his extrapolations of Moore’s law). Everybody laughed at him, but it’s increasingly likely he’ll be right on the
302.
by p1esk 1y ago
All you need is to publish a couple of right papers and/or significantly contribute to a couple of right projects. If you have brains for that you’ll be noticed.
303.
by p1esk 1y ago
You can say the same about any set of 100 people.
304.
by p1esk 1y ago
If comparing Deepseek to OpenAI, sure. But OP is comparing DeepSeek to Meta: both are followers.
305.
by p1esk 1y ago
Do you think Anaconda founders will ever become billionaires? If not, why would you expect it for Astral founders?
306.
by p1esk 1y ago
What is their revenue?
307.
by p1esk 1y ago
Thank you! That’s quite a lot of information to digest. What do you mean by “Prepare for every workout with some carbohydrates and electrolytes”? What should I prepare?
308.
by p1esk 1y ago
I don’t understand this: if I’m fat and I start exercising without changing my diet, I should be losing fat and gaining muscle mass. As a result, my weight can be going down.
309.
by p1esk 1y ago
I think by default ChatGPT will choose 4o for you. So unless you actually chose o3 you haven’t used the best model.
310.
by p1esk 1y ago
When you say the latest chatGPT, do you mean o3?
311.
by p1esk 2y ago
You like being on small collaborative teams building cool impactful things, but you don’t like how the companies you worked at do things (depend on VC, or whatever). So just start your own - build cool things and don’t take VC money. You on
312.
by p1esk 2y ago
It sounds like you should start your own company.
313.
by p1esk 2y ago
I don’t see any testing performed in that video. Did I miss it?
314.
by p1esk 2y ago
Source for your 10% number?
315.
by p1esk 2y ago
they are designed to elucidate the author's thought process - not the reader's learning process No, it’s exactly the opposite: when I write papers I follow a rigid template of what a reader (reviewer) expects to see. Abstract, i
316.
by p1esk 2y ago
I don’t see o1 in their model list. That’s the only model that produces decent answers to my questions.
317.
by p1esk 2y ago
I don’t want to follow anyone, but I do give stars to repos I like.
318.
by p1esk 2y ago
This answer is probably better than 99.99% of human answers. Feel the AGI!
319.
by p1esk 2y ago
Pretty sure a waymo car drives better than an average SF driver.
320.
by p1esk 2y ago
Yes, we pay well. No, we don't pay as much as Meta. How well do you pay? If I were at Meta, my total comp would be 500-600k. I make half that at a small startup. Can you afford me?
321.
by p1esk 2y ago
I am also disappointed, and I have not missed the context. The talk is empty for anyone who follows the field for more than two years, and especially for those who are familiar with his 2014 paper. Yes, he had amazing insight and intuition
322.
by p1esk 2y ago
multiple GPT-4 models working together What do you mean? Is o1 not a single model?
323.
by p1esk 2y ago
Wow! That’s all I need to know about Google’s model.
324.
by p1esk 2y ago
Are these benchmarks still meaningful?
325.
by p1esk 2y ago
Can you please give an example of a “completely illogical statement” produced by o1 model? I suspect it would be easier to get an average human to produce an illogical statement.
326.
by p1esk 2y ago
How is TTT anything other than a deep learning algorithm? We have a deep learning model, we generate training data based on an example and use a stochastic gradient descent to update the model weights to improve its predictions according to
327.
by p1esk 2y ago
World understanding.
328.
by p1esk 2y ago
Bigger LLMs, mainly for video frame prediction - that’s where we have unlimited training data.
329.
by p1esk 2y ago
How big is your model?
330.
by p1esk 2y ago
OK, so the question becomes: TSMC could do it. Why can’t Intel?

More ›