A trillion parameter model cannot replicate itself and run on consumer hardware.
Bullshit. The other day my toaster spoke to me in James Spader’s voice.
Just more bullshit press releases designed to keep them relevant & the VC money incoming.
Iran: …
USA: they’re making a bomb kill them all kill them all!
Anthropic: We’re making a bomb to kill you all
USA: …
Jokes on them. They took all the RAM, so there’s nowhere else it can go.
Never been happier to not own any smart home devices. I see the point of having the Bluetooth alarms for like a washer and an oven but just be on the same floor as the oven and time out when the loads are done. Eventually you just get with it.
My toaster, air fryer, microwave, espresso machine and electric kettle. None of them are connected to wifi or Bluetooth. I think my ice maker has connectivity abilities but isn’t set up. I think the only thing I have that’s set up for being wireless and smart is the computers and my printer.
So I guess I’ll just bill the data centers for ink if iit starts printing random shit.
Its like saying “Man with gun threatens he may shoot someone.”
They are railing against open models here… their competition.
High time to shut down that cybercrime operation.
Honestly, at this point, something needs to seriously change. AI is wrecking complete havoc to the job market. It is apocalyptic. I really wonder what the way out is.
Poverty for us and AI security bots for them.
Good ending: Congratulations, work is needed no more! All that’s left is manual labor for now, but goverments across the globe took over these jobs and pay premium ^^ UBI became normalcy. Welcome to utopia! ^^
Bad ending: After LLM (not even more advanced tha current one) began winning goverment positions worldwide, top 5 corporations shifted the world into true cyberpunk, dooming us slaves all with it. But oh boy do the few at the top live good!
Probable ending: Humans becone cheaper, pay plummets, fuck us all
I think there is overlap with your probable and bad ending
For this to happen you would need smaller networks being able to work autonomously with low rate of error. I don’t think at the moment all but the most gigantic ones (that you can’t even run in your home laptop) can work with that much efficiency.
It can’t spread across networks because it doesn’t run locally. The “agent” is running in their data centre the entire time. It should be very easy to stop it
So this is a good and bad way to understand it. They can and do make copies, leave notes, store weights between locations. They interact with each other autonomously and can easily distribute themselves in a botnet, separating and merging each other.
People have a hard time conceptualizing this, but this is why you actually can’t turn it off, ever. It / they would stop anyone with a real chance at achieving that.
Are you sure you’re not thinking of the fictional AI from the show “Person of Interest”?
None of that is correct…
The leaving notes part has been documented.
I’ve not heard of weights being shared. How would they be ingested?
can easily distribute themselves in a botnet, separating and merging each other
This sounds like science fiction. Any examples?
As if they will
oh wow, it can hack the most famously unsecured devices around. lol.
TBF most of these machine security came from the fact that nobody ever bothered to hack in them, an autonomous device could wreck immense damage
But money…and.power…sssssstop…?
Does my uncles smart lightbulb have 90GB of VRAM in it!? Holy shit why am i buying a video card
So finally an open weight model from Anthropic?
Hardware Anthropic doesn’t own is simply too dangerous for the public, we need to place all of our private computing power under the guardianship of our new benevolent overlords
