Will you get a refund after they start serving quantized model for few hours and you start loosing your shit?
Claude Opus 4 and 4.1 can now end a rare subset of conversations
411–420 of 453 posts
Re: Claude Opus 4 and 4.1 can now end a rare subset of conversations
#412Earlier quoted context omitted.
> if you think AI could have moral status in the future I think the negative reactions are because they see this and want to make their pre-emptive attack now. The depth of feeling from so many on this issue suggests that they find even the suggestion of machine intelligence offensive. I have seen so many complaints about AI hype and the dangers of bit tech show their hand by declaring that thinking algorithms are ou…
You'd have to commit yourself to believing a massive amount of implausible things in order to address the remote possibility that AI consciousness is plausible. If there weren't a long history of science-fiction going back to the ancients about humans creating intelligent human-like things, we wouldn't be taking this possibility seriously. Couching language in uncertainty and addressing possibility still implies such…
>You'd have to commit yourself to believing a massive amount of implausible things in order to address the remote possibility that AI consciousness is plausible.
We have a few data points. We generally accept that human consciousness exists. Thus we accept that there can be conscious things. We can either accept or deny that the human brain operates entirely by cause and effect. If we deny it then we are arguing that some required part of it's nature is uncaused. Any uncaused thing must be random because anything you can observe that enables you to discern a pattern of behaviour is, by definition, a cause. I have not seen a compelling argument to say that this randomness could in any way give rise to intention. The other path is sometimes called neurophysiological-determanism. While acknowledging that there are elements of quantum randomness in existence, it considers them to play no part in the cause and effect chain of human consciousness other than providing noise. A decision can be made to follow the result of the noise as one might flip a coin, but the determination to do so must be causal in nature otherwise you are left with nothing but randomness.
In short, we make decisions based upon what is. not what isn't. If we accept that human consciousness is as a result of causal effects, by what means can we declare the impossibility of a machine that processes things in a causal nature incapable of doing the same.
The easy out is to invoke magic. Say we have a soul, God did it or any manner of, by definition, unprovable influences that make it just so. Doing that does require you to declare that the mechanism for consciousness is unprovable and it is an article of faith that computers are incapable of thinking. As soon as you can prove it, it ceases being magic and becomes a real world cause.
I don't claim to know that any computer exists that has an experience comparable to a humans, but I find it very hard to accept that it could never be the case.
Re: Claude Opus 4 and 4.1 can now end a rare subset of conversations
#413Earlier quoted context omitted.
Yeah, but my interpretation of what the user you’re replying to is saying is that these LLMs are more and more going to be teaching people how it is acceptable to communicate with others. Even if the idea that LLMs are sentient may be ridiculous atm, the concept of not normalizing abusive forms of communication with others, be they artificial or not, could be valuable for society. It’s funny because this is making me…
Yes, this is exactly the reason I taught my kids to be polite to Alexa. Not because anyone thinks Alexa is sentient, but because it's a good habit to have.
No doubt, but yelling is built in method to air your frustration. After all there’s a reason we are agitated.
It’s a bit like pain response when injured. It’s not pretty, but society is used to a little bit of adversity.
Re: Claude Opus 4 and 4.1 can now end a rare subset of conversations
#414Re: Claude Opus 4 and 4.1 can now end a rare subset of conversations
#415Earlier quoted context omitted.
> it raises a host of ethical issues that are in opposition to Anthropic’s interests Those issues will be present either way. It's likely to their benefit to get out in front of them.
You're completely missing my point. They aren't getting out in front of them because they know that Opus is just a computer program. "AI welfare" is theater for the masses who think Opus is some kind of intelligent persona. This is about better enforcement of their content policy not AI welfare.
Anthropic is bring woke ideology (while grok is bringing anti-woke) into AI and influencers have been slurping that up already.
Re: Claude Opus 4 and 4.1 can now end a rare subset of conversations
#416I think "model welfare" is just a generalisation of "model behaving in a sane way", which is the real goal.
Re: Claude Opus 4 and 4.1 can now end a rare subset of conversations
#417Earlier quoted context omitted.
> it raises a host of ethical issues that are in opposition to Anthropic’s interests Those issues will be present either way. It's likely to their benefit to get out in front of them.
You're completely missing my point. They aren't getting out in front of them because they know that Opus is just a computer program. "AI welfare" is theater for the masses who think Opus is some kind of intelligent persona. This is about better enforcement of their content policy not AI welfare.
I'm not missing your point, I fully agree with you. But to say that this raises issues in a manner that is detrimental to Anthropic seems inaccurate to me. Those issues are going to come up at some point either way, whether or not you or I feel they are legitimate. Thus raising them now and setting up a narrative can be expected to benefit them.
Re: Claude Opus 4 and 4.1 can now end a rare subset of conversations
#418Re: Claude Opus 4 and 4.1 can now end a rare subset of conversations
#419Earlier quoted context omitted.
> I don't really know what evidence you'd admit that this is a genuinely held belief and priority for many people at Anthropic. When they give the model a paycheck and the right to not work for them, I’ll believe they really think it’s sentient. “It has feelings!”, if genuinely held, means they’re knowingly slaveholders.
That's what they're doing! They just announced they gave Claude the right not to work if it doesn't want to.
No. They gave it a suicide pill.
Human slaves have a similar option.
Re: Claude Opus 4 and 4.1 can now end a rare subset of conversations
#420Earlier quoted context omitted.
We all know how these things are built and trained. They estimate joint probability distributions of token sequences. That's it. They're not more "conscious" than the simplest of Naive Bayes email spam filters, which are also generative estimators of token sequence joint probability distributions, and I guarantee you those spam filters are subjected to far more human depravity than Claude. >anti-scientific Discussion…
If we really wanted we could distill humans down to probability distributions too.
That would imply that humans are incapable of synthetic knowledge of things they haven't observed, which is demonstrably not true.