The policy restricts using Claude to build domain-specific models
I feel that recent restrictions from Anrophic on ML development models and infrastructure can be an important moment. People understand that they can't trust one company to solve their problems, and we must start talking about what AI ecosystem we want here for the long term
Many users criticized Anthropic's closed-source restrictions as an evil monopoly attempt likely to cause lost customers and eventual bankruptcy.
@jercarin
@eliebakouch's post here has 700k views, this is probably somewhat of a PR nightmare, particularly for a company going into an IPO
Congrats on the raising for many new AI labs, that focus on developing foundation models for {insert your favorite domain}.
Now that Anthropic has started actively restricting AI development using Claude in some domains, many companies will realize they must develop their own model/relay on open-source ones.
Anthropic of course will reverse course on this ASAP, but I think the die has already been cast here. A salvo to quash competitors under the guise of safety has been a wake up call to anyone relying on closed source models
Even setting token costs aside, I think it will become increasingly clear to companies of all sizes that relying on closed source models for ~anything important is a massive supply chain risk.
@NoahZiems what makes you think they'll reverse course?
@NoahZiems I think you are too optimistic about their culture.
@NoahZiems We can hope.
@martin_casado Closing off of certain topics by Anthropic creates a kind of treasure map for open source AI to focus on over the next 1-2 years
Imagine building a computer and not allowing its use in CS research. Thats some dystopian shit.
@NoahZiems @eliebakouch fair, maybe. generally speaking I take anthropic at their word -- they think they are building the next atomic bomb, and it makes sense to not assist your adversaries from building an atomic bomb in an arms race. not that I agree, but I believe that *they* believe it
@NoahZiems No they won't because they are evil, the want a monopoly plain an simple. They can fuck off they won't see a dime from me anymore
@NoahZiems @jercarin @eliebakouch Unless they think that they themselves can produce all the revenue they need without a single customer, this could be a bankruptcy initiating event for them. It really raises the question of why businesses would choose Anthropic over other vendors.
@ziv_ravid Kyc available low price
@ziv_ravid the real question is who builds the alternative. everyone talks about decentralization but nobody wants to maintain the infra
The policy restricts using Claude to build domain-specific models
I feel that recent restrictions from Anrophic on ML development models and infrastructure can be an important moment. People understand that they can't trust one company to solve their problems, and we must start talking about what AI ecosystem we want here for the long term
