Anthropic just said they're going to limit the ability of the new model to assist with building foundational models.
The ability to assess the user's goal and tune the inference accordingly is there. So, what methods are there to tell if a provider has their thumb on the scale?
Plodding along a longer path to the same goal would burn more tokens without necessarily a decrease in solution quality. Maybe a few more docs.
Heard a tongue-in-cheek comment about "building a god" from someone at one of these AI labs.
The builders believe that the machine you describe will judge them positively, purely because they are building the system according to their judgment and beliefs.
> As a company, Anthropic always does as much as it can to work with customers to find creative new use cases and new sources of revenue that allow them to do more with their existing workforce, rather than focusing solely on cost savings (which often means reducing the workforce).
Without direct workforce or policymaker representation on the boards of private entities, the private sector will seek to maximize shareholder value even if that means workforce reductions.
It's not clear that any country could realistically ensure that incredibly powerful industries/private sector entities operate perfectly aligned with national interests, short of nationalization.
Large tech companies are already quasi-state actors. In theory, international law and regulations can be binding and enforceable. We see how well that works in practice.
The First Reconstruction was a very different civ-mil scenario. The military protected freedmen from the various insurgent and paramilitary groups that sought to deprive freedmen of rights.
The ability to assess the user's goal and tune the inference accordingly is there. So, what methods are there to tell if a provider has their thumb on the scale?
Plodding along a longer path to the same goal would burn more tokens without necessarily a decrease in solution quality. Maybe a few more docs.