On: U.S. appeals court upholds designation of Anthropic as supply chain risk

[gwerbin]

But that is not at all a reasonable fear. Is it really reasonable to believe that Anthropic, after receiving a government contract, would then proceed to sabotage their own product to not function as contracted? That seems like an utterly ridiculous claim to me, nothing close to "reasonable". There is no charitable way to view this designation except as political punishment and/or as a favor to Altman and Musk.

[chrisjj → gwerbin]

> would then proceed to sabotage their own product to not function as contracted?

Claude terms here: ANTHROPIC EXPRESSLY DISCLAIMS ALL IMPLIED WARRANTIES, INCLUDING WARRANTIES OF MERCHANTABILITY, NON-INFRINGEMENT, AND FITNESS FOR A PARTICULAR PURPOSE

Like most so-called AI, the Claude program is inherently unreliable. I doubt Anthropic would ever agree to "function as contracted".

[fn-mote → chrisjj]

The point to me is that the contract would include the understanding that the system was not 100% reliable.

[chrisjj → fn-mote]

OK, but even so, disclaiming FITNESS FOR A PARTICULAR PURPOSE lets Anthropic disable any function it wishes.

[gwerbin → chrisjj]

But that has nothing to do with the supply chain risk designation, it's just a DoD requirements mismatch.

[chrisjj → gwerbin]

My response was re the not functioning as contracted issue, not the supply chain risk. Having said which, I can well imagine possible cause and effect here.

[ExoticPearTree → chrisjj]

That means that the software was not tested for every scenario. Not that the user cannot try to use it in a way it was never intended to.

If, on the other hand, Anthropic deliberately blocks the user from doing something, then it is sabotage.

[chrisjj → ExoticPearTree]

> That means that the software was not tested for every scenario.

Actually I think it means what it says.

> Not that the user cannot try to use it in a way it was never intended to

No-one said it did. User can try anything he likes. That's beside the point.

> If, on the other hand, Anthropic deliberately blocks the user from doing something, then it is sabotage.

Well, consider CSAM generation. Anthropic is free to block whatever it likes.

[gwerbin → chrisjj]

Oh please. Then any MIT licensed software is also a supply chain risk.

[chrisjj → gwerbin]

I didn't mention supply chain risk.

My point is about function as contracted.

[ExoticPearTree → chrisjj]

Again: if Anthropic deliberately blocks a user from ordering it to prosecute a target, then it is degrading the software capability on purpose.

[chrisjj → ExoticPearTree]

Yes, but degrading capability on purpose is entirely permitted by a No Warrantee contract, and hence not at all the sabotage their own product to not function as contracted as feared by to DoD. Unless there's a very different and unusual contract in this case.

[ExoticPearTree → chrisjj]

I don't think you understand what "No warranty" means in a software license. It means Anthropic cannot be held liable for results if the software misbehaves and instead of bombing a terrorist compound it bombs a wedding (I chose this example because the US has a tradition to bomb weddings in Iraq and Afghanistan). But in no form does it mean "yeah, I don't feel like doing what you prompted me".

Hence, Anthropic sabotaged the models on purpose to not respond properly to all prompts.

[chrisjj → ExoticPearTree]

> It means Anthropic cannot be held liable for results if the software misbehaves

Actually I think it means what is says - which does include no warrantee of FITNESS FOR A PARTICULAR PURPOSE. This lets Anthropic block any purpose it likes.