You can't pull the plug after it kills people. For example see [1], where the DOJ mistargeted a school with AI.
There are scenarios where kill -9 isn't going to happen in time. What if the team that is harming people with AI is different from the one that is monitoring the harm? What if no one is monitoring? What if the user is intentionally malicious?
And wiping out humanity doesn't necessarily mean shooting people either. Every trader involved in the '08 financial crisis was locally acting in their own interests. Those could have easily been AIs optimizing trading strategies too.
Do you think that AI killing us would not involve us doing stupid shit with AI first? Or do you have some strange idea that this strange AI we're talking about would just pop up out of nowhere like a miracle?
He was clearly wrong about LLMs not being able to plan, etc. You think he’d think LLMs are solving/proving millennium prize math problems by now? At this point he’s just doubling down
> then they should also be liable for the harm caused
By that logic, producers of hammers should be liable for people banged in the head. No. It does not happen for gun producers, you figure for makers of screwdrivers "sometimes used for stabbing".
If a doctor told a patient their delusions were real and that they should kill themselves or family members, that doctor would be disbarred and face a prison sentence.
That is what is going on here; not the ancient “hammer/gun” defense.
SOTA LLMs are giving both medical and psychological advice they have absolutely no authority to give. If you or I convinced someone to kill themselves, we’d face a prison sentence. [0]
Hammer companies aren't both selling you the hammer and then literally telling you to harm someone with it.
And there are people that have much more credibility than him who actually take this scenario seriously. But I'm sure you will downplay them by saying they are tech bros or that they have some stake in being doomers (as if saying that AI might kill everyone would be good strategy for attracting investors - it's obviously not).
> you seem to be very invested in AI wanting to kill us.
Quite contrary - I wish AI did not exist or at least that the progress would plateau.
> I Wonder why?
Because I don't want to die.
> Tell us who you work for.
I suspect you want to imply I work for OAI or other lab - I don't. If I did, I wonder why would I want to lie* that technology I develop could kill my investors. I could ask who YOU work for - what interest do you have in downplaying dangers of AI?
* here we assume that people who say AI could be extremely dangerous are lying and not actually believing it - personally I believe that they don't lie and actually believe it. Why do they keep working on this technology then? Read mails between Musk and Altman from decade ago.
You can't pull the plug after it kills people. For example see [1], where the DOJ mistargeted a school with AI.
There are scenarios where kill -9 isn't going to happen in time. What if the team that is harming people with AI is different from the one that is monitoring the harm? What if no one is monitoring? What if the user is intentionally malicious?
And wiping out humanity doesn't necessarily mean shooting people either. Every trader involved in the '08 financial crisis was locally acting in their own interests. Those could have easily been AIs optimizing trading strategies too.
[1] https://futurism.com/artificial-intelligence/us-military-pen...
[flagged]
This is a very ironic statement.
Do you think that AI killing us would not involve us doing stupid shit with AI first? Or do you have some strange idea that this strange AI we're talking about would just pop up out of nowhere like a miracle?
3 replies →
I won't bother arguing, but there's also no need to be rude about it
He was clearly wrong about LLMs not being able to plan, etc. You think he’d think LLMs are solving/proving millennium prize math problems by now? At this point he’s just doubling down
[dead]
Why create an account just to make this comment?
> No one is dying from a glorified knowledge base that can do auto correct amazingly well.
All kinds of vulnerable people are dying due to LLMs. [0]
If SOTA LLM companies can benefit from the "intelligence", then they should also be liable for the harm caused.
[0]: https://en.wikipedia.org/wiki/Deaths_linked_to_chatbots
> then they should also be liable for the harm caused
By that logic, producers of hammers should be liable for people banged in the head. No. It does not happen for gun producers, you figure for makers of screwdrivers "sometimes used for stabbing".
If a doctor told a patient their delusions were real and that they should kill themselves or family members, that doctor would be disbarred and face a prison sentence.
That is what is going on here; not the ancient “hammer/gun” defense.
SOTA LLMs are giving both medical and psychological advice they have absolutely no authority to give. If you or I convinced someone to kill themselves, we’d face a prison sentence. [0]
Hammer companies aren't both selling you the hammer and then literally telling you to harm someone with it.
[0] https://www.npr.org/2019/02/12/693807708/woman-who-provoked-...
2 replies →
> Because he has credibility, much more than you.
And there are people that have much more credibility than him who actually take this scenario seriously. But I'm sure you will downplay them by saying they are tech bros or that they have some stake in being doomers (as if saying that AI might kill everyone would be good strategy for attracting investors - it's obviously not).
[flagged]
> you seem to be very invested in AI wanting to kill us.
Quite contrary - I wish AI did not exist or at least that the progress would plateau.
> I Wonder why?
Because I don't want to die.
> Tell us who you work for.
I suspect you want to imply I work for OAI or other lab - I don't. If I did, I wonder why would I want to lie* that technology I develop could kill my investors. I could ask who YOU work for - what interest do you have in downplaying dangers of AI?
* here we assume that people who say AI could be extremely dangerous are lying and not actually believing it - personally I believe that they don't lie and actually believe it. Why do they keep working on this technology then? Read mails between Musk and Altman from decade ago.
1 reply →