p1esk

37 ms·

601.
by p1esk 2y ago
Sure, but in the event of liquidity, can they refuse to pay me the value of my shares? For any reason?
602.
by p1esk 2y ago
Can you give us a couple of examples? I mean your ideas.
603.
by p1esk 2y ago
Ideas are cheap. It’s all about “thinking about problems and trying to solve them”.
604.
by p1esk 2y ago
Yes, you can learn how to sue. You can learn how to be a doctor too. You can also learn rocket science. The third one is the easiest to me, personally.
605.
by p1esk 2y ago
My understanding is as soon as you exercise your options you own them, and the company can’t take them from you. Can anyone confirm this?
606.
by p1esk 2y ago
But you can opt out, right? So what’s the problem? Also, is Teams (and other messengers) any different?
607.
by p1esk 2y ago
It’s like saying SSD is a YOLO. Both are single shot object detectors, but only YOLO is “a YOLO”.
608.
by p1esk 2y ago
I’d be surprised if more 1% opt out.
609.
by p1esk 2y ago
GPT refers to the specific family of models developed at OpenAI.
610.
by p1esk 2y ago
GPT4 is great for brainstorming. It helped me come up with an idea for my last paper.
611.
by p1esk 2y ago
Note all mentions of FLOPS in this thread refer to FP64 (double precision), unlike more popular “AI OPS”, which are typically INT8, specified for modern GPUs.
612.
by p1esk 2y ago
Nvidia has about a hundred direct competitors in AI accelerator space, both in datacenters and in edge devices. Apple is typically only interested in edge devices. Mac servers, despite the name, are edge devices. It is possible that Apple d
613.
by p1esk 2y ago
I’ve used it for a couple of hours to help with coding and it feels very similar to gpt4: still makes erroneous and inconsistent suggestions. Not calling it 4.5 was the right call. It is much faster though. The expectations for gpt5 are sky
614.
by p1esk 2y ago
This is still gpt4. I don’t expect much more from this version than what previous version could do, in terms of reasoning abilities. But everyone is expecting them to release gpt5 later this year, and it is a bit scary to think what it will
615.
by p1esk 2y ago
GPT4 has been a lot more useful to me than most normal people I interact with.
616.
by p1esk 2y ago
The key is to use GPT4, not the free version.
617.
by p1esk 2y ago
It’s been increasingly difficult to make money writing, even before AI. I believe AI will kill >90% of writing jobs in the near future. With a GPT-5 level model, a human manager/editor will be able to lay off most of their writers.
618.
by p1esk 2y ago
Rumors are it will be a really good voice integration. Maybe something like in the movie Her?
619.
by p1esk 2y ago
Hard to try it on my phone.
620.
by p1esk 2y ago
That’s what people said about CV models in 2011.
621.
Managers have no human rights(yosefk.com)
13 points by p1esk 2y ago|7 comments
622.
by p1esk 2y ago
I want to play with the best model I can get my hands on. In the nearest future, this will probably be llama 400B. Even that model will probably be dumber than GPT4, and GPT4 feels pretty dumb sometimes.
623.
by p1esk 2y ago
what is considered affordable hardware for running large language models locally today? I’d say “under $20k” is considered affordable. In comparison, a single H100 server is $250k. At least if you want to run decent models (>70B) at be
624.
by p1esk 2y ago
Why don’t you test it first with a small model on mnist or cifar?
625.
by p1esk 2y ago
How many young people would be driven to programming in this scenario?
626.
by p1esk 2y ago
So why can’t you just smoke weed after work? Or do you need it all day long?
627.
by p1esk 2y ago
Common plebs also include different people.
628.
by p1esk 2y ago
BDSM people (hopefully) willingly choose to have that experience.
629.
by p1esk 2y ago
Unfortunately different people often want different things.
630.
by p1esk 2y ago
That substitution is the main difference between SGD and RWP. It’s like describing bubble sort when you meant to describe quick sort. Would not fly on an ML 101 exam, or in an ML job interview.

More ›