p1esk

32 ms·

571.
by p1esk 2y ago
I don’t see gpt4 scores there. In fact I’m particularly interested in the performance of a natively multimodal model, like gpt4o or gemini. It does not really make sense to test a model trained on text on those visual/spatial puzzles.
572.
by p1esk 2y ago
What many people don’t seem to realize is there are a lot of early stage but already well funded (10M+) startups who are desperately looking for top quality people. Once I was approached by a founder who offered 500k base salary (wasn’t a
573.
by p1esk 2y ago
Thank you. I did not realize old wars were so devastating in terms of lives lost, especially considering much smaller total population. It’s incredible that modern governments, being so incompetent, corrupt, and dysfunctional, are still a l
574.
by p1esk 2y ago
Why does everyone thinks startups don’t pay well? I have worked for various startups all my life, most of them well funded, and competing for talent with faangs. Yes, I could probably make more at Google but I don’t feel like I’m underpaid.
575.
by p1esk 2y ago
With cofounders it’s fairly typical for one to be technical and the other business savvy.
576.
by p1esk 2y ago
So a billion years of evolutionary search plus 20 years of finetuning is a better method?
577.
by p1esk 2y ago
True. Altman said he * hopes* the quality improvement will be as much as 3.5 —-> 4.
578.
by p1esk 2y ago
That’s what I expect from gpt5.
579.
by p1esk 2y ago
Is there a leaderboard for the no-restriction version of the competition? I want to see how gpt4 does on it.
580.
by p1esk 2y ago
illusion that these models are learning like children do, when in reality they are not even entirely capable of learning novel concepts Now imagine how much would your kid learn if the only input he ever received was a sequence of words?
581.
by p1esk 2y ago
If we look at 1800s, and take any single year in terms of causalities, how would it compare to 2024? Would 2024 be closer to the top or bottom ten deadliest years of 1800s?
582.
by p1esk 2y ago
2024 is not * relatively* peaceful.
583.
by p1esk 2y ago
They might not need to raise any more money. If they manage to release a significantly smarter model (gpt-5), I, and millions of others, will gladly pay hundreds of dollars per month to use it.
584.
by p1esk 2y ago
2024 is not peaceful.
585.
by p1esk 2y ago
I thought the hormonal issues are caused by steroids, no?
586.
by p1esk 2y ago
Not sure who are you arguing with. I compared learning on your own with learning in grad school. To prepare for real world jobs. Grad school is better imo.
587.
by p1esk 2y ago
In a typical phd program you have a lot more freedom (as long as you publish), a lot more flexibility than in a typical 9-5 job, and a lot less politics on average. Pressure is about the same.
588.
by p1esk 2y ago
What you described is similar to a typical work environment you should be ready for, unless you were born rich.
589.
by p1esk 2y ago
Sure, but you’re getting something that will help you get a good job in the future, so it seems like a good investment.
590.
by p1esk 2y ago
If you do all that why not just enroll and get a degree while you’re at it?
591.
by p1esk 2y ago
I found that video pretty interesting.
592.
by p1esk 2y ago
FUB: Functional Unit Block. It seems that they are moving from higher level (larger) building blocks to lower level (smaller) building blocks to increase efficiency.
593.
by p1esk 2y ago
I love leetcode type puzzles. I still do an occasional puzzle even though I’m not looking for work. I like the feeling of intellectual satisfaction when I come up with a solution to an interesting problem.
594.
by p1esk 2y ago
the same type of optimism will make itself known during periods of strong mathematical culture and high rates of progress Would you characterize present time as one of such periods? I view current AI research as a branch of math, and the
595.
by p1esk 2y ago
Why do you care how others feel? How do you feel about working for Meta?
596.
by p1esk 2y ago
Is it comfortable to ride with no hands, sitting in a fully upright position? How does it compare to a wide plush seat like ones on cruisers?
597.
by p1esk 2y ago
Has anyone tried to measure gpt4 IQ?
598.
by p1esk 2y ago
It’s aimed at maybe a few hundred people in the world who can understand it. Maybe even few dozen.
599.
by p1esk 2y ago
What is “theoretical max” for intelligence?
600.
by p1esk 2y ago
In my system prompt I ask it to always start with repeating my question in a rephrased form. Though it’s needed more for lesser models, gpt4 seems to always understand my questions perfectly.

More ›