PostOnce

15 ms·

1. 271.▲
   by PostOnce 3y ago

   That's interesting that people felt it worth the time to email. I think what's going on here is that some people (myself included) find it extraordinarily offensive to question someone's right to procreate, whether they&#x2
2. 272.▲
   by PostOnce 3y ago

   I'm not trying to attack you, just the idea. (I saw the same sentiment from several commenters) I'm just saying I strongly disagree with publicly questioning a dead man's decision to have kids when the kid still has at least
3. 273.▲
   by PostOnce 3y ago

   You're right; I could've said that without calling it "a BS sentiment".
4. 274.▲
   by PostOnce 3y ago

   We don't live in an ideal world. That doesn't mean it's not worth living in. The man didn't hang himself, he got a case of severe bad luck. I'm sure he'd be here doing the father figure stuff if he could, but i
5. 275.▲
   by PostOnce 3y ago

   None of the commenters expressing this bullshit sentiment will provide their children with 1% of the education, health, freedom, security, etc that Mitnick will have left for his child. We struggle our whole lives to provide, it looks like
6. 276.▲
   by PostOnce 3y ago

   The company running this website was cofounded by the man who wrote the first internet worm, there's a chapter about rtm and pg in some of the same books that talk about Mitnick. Nobody's perfect.
7. 277.▲
   by PostOnce 3y ago

   Yes, they're not "citizen scientists", they're just scientists. How many people's job title is "scientist"? Debt from a PhD doesn't mean your science is necessarily more rigorous or valid; it may impl
8. 278.▲
   by PostOnce 3y ago

   I was going to write a post about how System Shock 3 is in development, but I'm dismayed by some details I found on wikipedia. They had Warren Spector and Paul Neurath on board, but maybe funding isn't so easy. I thought System Sh
9. 279.▲
   by PostOnce 3y ago

   I have a normal qwerty layout and opened it on a laptop and my first thought was "why the hell cant i scroll?" It's a stupid website. It took more work to make it shittier to use. It's a loss for the author AND the viewe
10. 280.▲
    by PostOnce 3y ago

    OpenAI was so destitute, having only $100,000,000, that they had no choice but to fuck us all and lobby congress to make it illegal for us to use this new technology? That doesn't make sense. If your choice is between $100M + doing ear
11. 281.▲
    by PostOnce 3y ago

    is it compression though? It's a model of the way our language works and contains a set of knowledge. If I read 200 physics textbooks, I couldn't recall any of them exactly , but I could write another book... that would contain
12. 282.▲
    by PostOnce 3y ago

    I tried to plug in a wii the other day only to realize the TV had no RCA inputs, so I hooked the wiimotes up to the laptop for the kids and fired up the emulator...
13. 283.▲
    by PostOnce 3y ago

    So far as anyone knows, this is not a derivative work, its transformative, and therefore not subject to any licensing requirement. You're right though, that's arguably still up for debate, but I think the precedent of transformati
14. 284.▲
    by PostOnce 3y ago

    > "outputs al qaeda propaganda" So does Microsoft Word. Both require a human to tell the software what to output. > as if none of the other FAANGs could trivially cross any regulatory barriers that emerge. That's the po
15. 285.▲
    by PostOnce 3y ago

    So fold. Running out of money to run a charity means they now suddenly need to fuck over the people the charity was supposed to help (that is, everyone on earth ) ? Imagine if a feed the homeless charity accepted private investment and w
16. 286.▲
    by PostOnce 3y ago

    It's several reasons, first it's the lies and the abuse of a charity, that wouldn't be an issue if they had began as a private company instead of robbing a charity. But secondly, even if they were a private company, it's
17. 287.▲
    by PostOnce 3y ago

    "Open" AI, a charity to benefit us all by pushing and publishing the frontier of scientific knowledge. Nevermind, fuckers, actually it's just to take your jobs and make a few VCs richer. We'll keep the science a secret a
18. 288.▲
    by PostOnce 3y ago

    We need more teeny-tiny projects that illustrate somewhat novel or lesser-known "things you can do". I love it.
19. 289.▲
    by PostOnce 3y ago

    Hobby? PrintNC? I'll read whatever you feel like typing about it.
20. 290.▲
    by PostOnce 3y ago

    "Will OpenAI actually be open" That ship sailed, friend. OpenAI is no longer a charity in any meaningful sense of the word anymore, it's now an adversarial organization working against the public good with the sole aim of m
21. 291.▲
    by PostOnce 3y ago

    CPU RAM
22. 292.▲
    by PostOnce 3y ago

    Here is a short test of a 7B 4bit model on an intel 8350U laptop with no AMD/Nvidia GPU. On that laptop CPU from 2017, using a copy of llama.cpp I compiled 2 days ago (just "make", no special options, no BLAS, etc): .&#x2
23. 293.▲
    by PostOnce 3y ago

    https://mpost.io/phi-1-a-compact-language-model-outpaces-gpt... a 1billion parameter model beats 175billion parameter GPT3.5 OpenAI wants us all to drink the kool-aid.
24. 294.▲
    by PostOnce 3y ago

    Different quantizations can give you a big speedup if you've had "depressingly slow" issues. Even the slowest ones (that fit in RAM) will run at basically interactive speed, not instant, but also not "email speed".
25. 295.▲
    by PostOnce 3y ago

    I'm using a variety of 7 and 13B models (and a 3B one for fast feedback loop debugging) at between 8bit and 4_K_M quantizations. Depending on your pre-prompt, your fine-tune (i.e. which model you downloaded), and your specific task, th
26. 296.▲
    by PostOnce 3y ago

    I have a Dell 7490 (intel 8350u cpu) I paid $250 for and I have no trouble running 13B models through a custom interactive interface I wrote as a hobby project in an afternoon. It can still get a lot better. I made it async the following da
27. 297.▲
    by PostOnce 3y ago

    The benchmark you linked was to "programming performance", not generic LLM "intelligence". The situation for the little guy is wildly better than most people imagine.
28. 298.▲
    by PostOnce 3y ago

    Your benchmark lacks the current #2 https://github.com/nlpxucan/WizardLM/tree/main/WizardCoder It beats Claude and Bard. You could probably get a 4bit 15B model going in 16GB of RAM and be approaching GP
29. 299.▲
    by PostOnce 3y ago

    OpenLLaMA is though. https://github.com/openlm-research/open_llama All of these are surmountable problems. We can beat OpenAI. We can drain their moat.
30. 300.▲
    by PostOnce 3y ago

    https://github.com/ggerganov/llama.cpp https://huggingface.co/TheBloke There's a LocalLLaMA subreddit, irc channels, and a whole big community around the web working on it on GitHub nd elsewhere.