ChatGpt already is multiple smaller models. Most guesses peg chatgpt4 as a 8x220 Billion parameter mixture of experts, or 8 220 billion parameter models squished together
ChatGpt already is multiple smaller models. Most guesses peg chatgpt4 as a 8x220 Billion parameter mixture of experts, or 8 220 billion parameter models squished together
Let them Fight
My one dark hope is AI will be enough of an impetus for somebody to update DMCA
Flat epistemological statements like this are why I feel like more STEM people need to take Philosophy.
> pay once, get access to everything everywhere
> thinks about Elsevier
OH GOD PLEASE NO
That doesn’t seem to be the same article
Turns out that whole idea of women being the primary bearers of hundred of years of exploited reproductive labor might have had some weight to it, huh.
All that labor being redirected into “L’economie” means that, at base, you’ll have less children.
This is interesting but I’ll reserve judgement until I see comparable performance past 8 billion params.
All sub-4 billion parameter models all seem to have the same performance regardless of quantization nowadays, so 3 billion is a little hard to see potential in.
Those cost efficiencies are also at the expense of the Chinese government. The massive investment is all part of their green revolution policy package.
It’s why Solar cells are also incredibly cheap to produce in China, and why they’re also mostly sold in China.
I seriously doubt the viability of this, but I’m looking forward to being proven wrong.
The OSI just published a resultnof some of the discussions around their upcoming Open Source AI Definition. It seems like a good idea to read it and see some of the issues they’re trying to work around…
https://opensource.org/blog/explaining-the-concept-of-data-information
I would recommend instead to use the AI Horde: https://stablehorde.net/ It’s a collection of people hosting stable diffusion/text generation models
There’s also openrouter which can connect to ChatGPT with a token-based system. (They check your prompts for hornyposting though)
It helps differentiate between GNU/Linux users and the five people who use GNU/Hurd
They should probably do the same with cascade, seeing as that license seems to be more restrictive as well
Judging by my bank account I’m transitioning to non-profit status as well.
I think the type of situation where you start a business under your own labor, and then hire others to exploit their labor is possible in a well portioned socialist economy.
You wouldn’t be able to hire “just a dishwasher.” You’d enter into a mutually and equally benefitial partnership with somebody who you would share labor and extract the revenue from that labor equally.
The exact type of that relationship and the responsibility details are largely immaterial to whether it’s exploitative at that point.
You kind of have to presuppose that in an equal partnership that one of the partners is not being exploited. Anything else is just kind of a partnership in name only.
Yes of course, there’s nothing gestalt about model training, fixed inputs result in fixed outputs
See you guys in 2040