Amazon, well you know them :-)
Flogging their model on Bedrock etc. must be part of the plan for Anthropic but AWS investment must surely create tension.
The profit motive is _very_ resilient. Right now I will take it on face value that it's fuel for the ethical mission, but in the future that might change. It's far (far) more likely that the ethical mission will change than the profit motive will change.
More funding doesn't change it, it just raises the stakes.
It is also very telling that a corporation called literally "human-first" is developing a tool to displace said humans completely. Now it is possible that it won't be able to do this, but the problem is that almost everyone including the creators believe that NNs will do it.
Even if you believe in such a soul, does it govern the non-conscious systems that do almost all of the work of thinking and feeling? If not - once again, artificial systems will certainly surpass natural ones.
In other words, it's not a question of whether humans will always be the masters, but a question of whether the "gods" we create will love us or hate us. Anthropic is aiming for the former.
You (and pretty much the entire debate around AI safety) have smuggled in the notion that these AIs even have the capacity to "love" and "hate," and have the agency to perhaps act vindictively. How we get from predicting the most plausible next token to a god-like entity with agency is... not clear at all.
Do people think that we're actually sleepwalking towards skynet? Or are they just saying that so that Sam Altman can get the government to put in rules that bind everyone else while allowing OpenAI to proceed without competition and capture the extremely mundane real market of writing ad copy and such.
It's a metaphor, actually. I guess I needed heavier use of quotes.
The one thing we can be sure of is that if goal-oriented AI exists, it will follows goals.
> How we get from predicting the most plausible next token to a god-like entity with agency is... not clear at all.
It doesn't have to be clear. If it was clear, we'd be working on it. That's... how invention works.
First, that it has arranged to be controlled by a Long-Term Benefit Trust: "an independent body of five financially disinterested members with an authority to select and remove a portion of our Board that will grow over time (ultimately, a majority of our Board)". (https://www.anthropic.com/index/the-long-term-benefit-trust)
Second, that it also published and committed to a Responsible Scaling Policy: https://www.anthropic.com/index/anthropics-responsible-scali...
Disclaimer: I work at Anthropic
So now the gpu farms of meta, google, X, microsoft and amazon are pretty much reserved for large foundation model training of specific models. It was probably a good thing for Anthropic to reserve their spot at the pool, even if as you point out it creates tension with their ethical mission. What options remain for other players who want to compete at this level I wonder?
Strong FTX vibes.
Also why aren’t they a little bit more transparent about the deal? Are they worried GPT-7 from the future is gonna travel back in time to stop them?
And OpenAI's mission was "open".