Forgot your password?
typodupeerror

Comment Re: US Bubble AI Economy Pops - Chinese Cheapness (Score 1) 109

They can be called in the same way, but not all of them are as good at working in a particular harness as each other. Anthropics models have been trained to call tools in a certain way, the chinese models may or may not work in a way that is compatible. Tool calling ability (not just whether or not you can do it, but how and when it decides which tools to call and why) differs WILDLY between open models.

Comment Re:Respecting copyright is an important part of FO (Score 1) 109

The whole process that split AT&T's System V and BSD should bear some weight here, at some point there was an agreement that, once BSD rewrote the few offending portions, AT&T had no claim anymore.

Frankly I'm surprised that the settlement between SCO and IBM didn't include verbiage that this was a done deal with no right for any successor-entity to bring this up again.

Comment Re:This is the plot for "The Blob", isn't it? (Score 1) 59

There are tons and tons of pathogens with high mortality rates without medical intervention. There are tons of pathogens that only see minimal death rates without active medical intervention because vaccination reduced the penetration that those pathogens have into the community and may have even forced evolution for increased transmissibility in lieu of virulence in order to spread at all.

Comment Re:Don't look! Don't look! (Score 5, Interesting) 97

Damn, I looked. Who else would be self important enough to continuously log their location? And then stupid enough to rob a bank?

Just because someone is stupid doesn't mean that they aren't subject to specific protections under law.

Ernesto Miranda, for whom the Miranda Warning is named, was by accounts a terrible person. Miranda's conviction was thrown out on those technical grounds that his confession should not have been permitted, then he was retried and convicted of the crime without his confession as evidence. Once he was released from prison he died in a bar fight.

The point of protections are that they apply to everyone, guilty or innocent, and are supposed to regulate the way that the legal system all the way from the patrolman to the attorney general behave. That doesn't mean that criminals aren't still criminals, but it does mean that the government has to provide proper justification for its actions against persons. If someone really did commit a crime then the government should be able to show cause, and this keeps everyone else from being scrutinized when the government has no business scrutinizing.

Comment Re: They can only self-improve if they are capable (Score 1) 216

They are. It's just not happening on the versions we have access to. Look at Sakana. RSI just involves defining an experiment and parameters, and varying the values of the parameters, trying out a training run, and record the results. If it works well, use the results as part of the base model for the next iteration. All of that can be automated. I have no idea why people are saying they can't be. If people iterate on them and change one variable at a time, and use that as a means to improvement, an AI can be tasked to do the same. And they can also be trained to design the experiments themselves.

Slashdot Top Deals

10.0 times 0.1 is hardly ever 1.0.

Working...