AI itself won’t destroy humanity but AI companies seem to want to have a go. They really need to be held accountable for what they do and stop being able to hide behind “the AI did it” like LLMs have any independent agency.
“My agent hacked hugging face” is almost certainly not a civil legal defence in the US, and I wonder why so many people seem to suggest that it is/was (willful ignorance, is my suspicion).
My bet is that OpenAI offered to compensate HF in one way or another (or already does) and an agreement was made.
OTOH a criminal case would likely be an uphill battle, unless criminal negligence could be established.
That’s why Nvidia bought Hugging Face; to stop HF from suing. Nvidia sacrificed a few days of profit to keep the AI bubble inflated and kick the can down the road.
I present you with your official tin foil hat. May you wear it proudly!
Yep, it is well known that AI can/will act unpredictable, if you fail to have operational restrictions that the AI can’t bypass, then you have to be responsible.
An AI/LLM should not have any generic outside access to anything.
And any access should be limited to the bare minimum to do it’s job.
You want it to have access to your calendar?
Ok what level of access do you mean?
Read Free/Busy?
Read the title of meetings?
Read the contents of meetings?
Accept/decline meeting requests?
Make new meetings?
Edit existing meetings?
Delete meetings?
And so on, one of the worst things we can do with AI/LLMs is to assume what it will do and act on that rather than to actually set real rules for it.
AI/LLMs does not understand reason, it doesn’t understand unwritten rules, every rule has to be set up outside of the AI/LLM chat, and use the philosophy that anything not explicitly allowed is blocked.
It doesn’t understand or follow written rules, never mind unwritten ones.
That is exactly what I wrote, I pointed out that the the rules defined for the AI/LLM has to be external of the AI/LLM, like a network firewall but more sophisticated.
It’s not exactly what you wrote, you only explicitly wrote it doesn’t follow unwritten rules. I think we are both being pedantic over the basic premise though, LLM output dumb.
Fair point!
I am glad we both see the extreme caution we need to treat AI/LLMs with.
Wait, I just think I figured something out…
What is the deterrent for humans to not break rules?
We don’t want the consequences that would impart.
An AI or LLM is not advanced enough to understand consequences these days, so they don’t matter, other than to act as an input to compose a suitable reply according to the LLM.
I have known this logically for a long time, but this is the first time I made the actual intellectual connection and actually understood cause and effect.
Probably old news, but having a moment of realization like this is always fun and interesting.
It’s been obvious for a while that scaling has hit diminishing returns (‘let’s voluntarily slow down’ like they can actually go faster) and the ‘just one more datacenter bro’ has stopped working.
Much smaller models have proven adequate for what the technology does well and the only thing stopping people moving to them wholesale is (probably illegal) market cornering. There is no moat but what they make.
At least it’s looking like there’s a good chance they won’t get to IPO and steal pension funds and crash economies as badly as they might have. SpaceX actually has worthwhile technology mixed in but xAI still managed to drag it down. Anthropic and OpenAI would go over like a lead balloon now and it’s only going to get worse as they continue to bring nothing new to the table.
They see the bubble collapsing and want to postpone the inevetable by ,voluntarily slowing" development.
It’s just the responsible thing to do, create panic to use as a smoke screen for all the other shit you’re hiding.
I’m sure there’s corporate jargon for doing this, but professionally
Proactive stakeholder management, perhaps
anthropoic wants to be ahead of the curve thats why, slow down so they can get someone else to hold the bag. they cant outwardly stop building DATACENTERS either.
