Technology
This is the official technology community of Lemmy.ml for all news related to creation and use of technology, and to facilitate civil, meaningful discussion around it.
Ask in DM before posting product reviews or ads. All such posts otherwise are subject to removal.
Rules:
1: All Lemmy rules apply
2: Do not post low effort posts
3: NEVER post naziped*gore stuff
4: Always post article URLs or their archived version URLs as sources, NOT screenshots. Help the blind users.
5: personal rants of Big Tech CEOs like Elon Musk are unwelcome (does not include posts about their companies affecting wide range of people)
6: no advertisement posts unless verified as legitimate and non-exploitative/non-consumerist
7: crypto related posts, unless essential, are disallowed
view the rest of the comments
I'm just waiting for the closed-source AI industry to have their Black Swan moment.
Something's going to come out of left field from the open AI community and all that investment in proprietary models and mega-compute is going to be rendered useless.
You still need the mega-compute. Even for inference, 1.5 terabytes of RAM in modern servers isn't cheap.
You can run Claude Sonnet equivalent quantised models locally on much less RAM.
Local LLMs are close to being viable. We're almost at the point where they fit a Macbook.
You're not thinking black-swan enough.
You're typing your comments using a blob of goo with about a hundred million neurons in it that cycles under a hundred hertz and draws less than 20 watts.
I don't think that we'll be running packs of goo in our PCs any time soon. But I do think some entirely different way of looking at the problem will emerge that will reduce computational requirements by many orders of magnitude. And it won't involve gigantic statistical engines trying to find the best average response to a question.
There are 86 billion neurons in the human brain and 16 billions in the cerebral cortex. More importantly, there are 100-1000 trillion synapses, and we know that parts the dendritic tree do independent computations as well. The computations are not synchronized with a global clock, but 1 ms events correspond to a kHz refresh rate and we know spikes can do temporal coding. Meanwhile, the best we can do is Cerebras CS-3 and with WSI at 5 nm there isn't much more where that came from.
Try doing the math on how many CS-3 you'd need to represent and refresh the above.
It could be TurboQuant or something like it. The biggest detraction to local LLM models is being able to close the gulf between obscenely-expensive 512GB NPUs, to house the 230GB uncompressed models (+ context), and more common 24GB GPUs. Quantized 15-18GB models are already working pretty well, but context size is still a bit of a problem.
Of course, the whole industry need to ramp up memory production and wrestle duopolies from the few that can make the raw silicon. It was pretty fucking pathetic that parts of the PC industry decided to leave these silicon processing weaknesses in various places. Large corps could have easily jumped into the industry and made bank in the long-term, but that would require not funneling into short-term quarterly profit bullshit.