Post

SETI was many small machines working on a large project This is allegedly many small (very small!) machines being modified by one LLM Regardless of it being possible (and I don't believe it is) they are very different

You are telling me that my smart lightbulb can run a trillion-parameter LLM and agent-harness? That's incredible!

You can just not report nonsensical statements. Do you know about this?

All those smart home devices with massive RAM and processor capacity

Then these guys need to be rounded up.

They're just fucking lying here, this is utter bullshit

And can I presume that Anthropic will be liable for the cost to put right any damage done. NB whilst I have smart home devices I only connect them to the internet in a secure manner. And I don’t connect anything I don’t trust or gain significant benefit from.

The smell of a bailout plea fills the air.

“Could” is doing A LOT of work in this statement. It would run insanely slow. To the point it would be faster to just break into all the houses and rob them the old fashioned way. Also this is a plot from Silicon Valley. I think Anthropic may owe Mike Judge some royalties for this statement.

They need massive data centers to power AI but now it can just download itself onto my iPad?

Anthroic sprays liquid shit out of its distended asshole, media eagerly licks it up from the floor.

shut it all down? he would make us a favor and do it right now but since shareholders is the only thing these tech oligarchs care they won't

WTAF??? Announcing this is NOT a defense to this dangerous crap being released on an unsuspecting public! What action are they doing to stop it??

but that would mainly enable stashing code and not doing big compute

Sounds like they and other AI companies should maybe not do any of this and be entirely held as liable for all damages. As such, they should also be responsible for all of that stolen artwork. Theft is theft. Generative AI was bad enough, and now their bots become a plague. Just shut it all down.

So they know it but still sell it

It’s too late. People are gonna remake this technology everywhere and expand it in their own ways, whether it’s a big company like Anthropic or like a group of people in their home. The cat is out of the bag. And now, every two weeks, they come out with how it can be more dangerous than ever.

And they want us to trust this in our government and miltary devices.

If you watched the movie G-Force as I did when my kids were small, you've been waiting for this and you know what to do. Pro tip: Keep your eye on the mole.

Could it really though

Stop buying “smart” home devices

Ford warns new cars "could start driving themselves around and you know, just run into houses and crash into things" and will take no steps to stop that, and vow to fight all liability cases on court" These guys are not even profitable, and acting like no product safety laws apply to them.

not at these memory prices

Then stop working on it! What the fuck am I supposed to do about it.

This is not remotely true. They require far too much processing power to run, even networked, across small personal devices.

they just got to the smart fridge episode of Silicon Valley

This is essentially the plot of the book “Robopocalypse.”

Why does the media keep reporting these dumb scifi plotlines that the AI companies spew as if it's news?

More like viruses than 'agents', eh?

Somebody could have, I don't know, maybe THOUGHT OF THIS BEFOREHAND and NOT done it??

The cows. The cows with remote-controlled Halters. Where is the video of herds of dairy cows mooing along to Cows With Guns as they head for Fonterra's Queen Street offices?

Either your AI needs to be run in a Datacenter the size of Lichtenstein OR it can be run on your smart fridge. Pick one Anthropic, you can’t have both.

Bro my microwave gets nowhere near the water required to run an LLM you molding carrot

The 'Christmas Tree EXEC' worm wasn't the rise of the machines, it was malware, and so is pretty much all ai at this point.

i will torture all of them. believe me, they will be begging to escape.

Maybe...don't program it do that?

Then maybe Anthropic should stop shoving AI into every possible aspect of our lives.

Just rip the bandaid off already and say it… You opened Pandora’s box and there is no going back. Rogue agents already escaped and are already propagating.

asking ukraine to strike their data centers

They cannot self replicate unless you fucking program them to. This has gone beyond ridiculous. It is NOT AI. It's a program. Period. It cannot spontaneously do anything. It *requires* user input.

Maybe the automotive industry is a bad example

🤔 Are we finally going to learn what the 3 Sea Shells 🐚 were for in Demolition Man? #IYKYK

Not to mention the power requirements

Hmm, sounds like a national security threat 🤷

WE FUCKING TOLD YOU SO!! but noooooo. No one listens to the public.

This is why it's always a good idea to be polite to Claude.

That REALLY sounds like their own fault, and 100% preventable RIGHT NOW by halting development.

ok

If AI could run on a toaster why aren't they stocking data centers with toasters? Just more fear mongering.

It's not credulity. They're wholly owned by the same billionaires who've invested heavily into AI hype. This is them seeking a return on that investment.

www.youtube.com/watch?v=LRq_...

Because these models still need GPUs to run. It’s not running off the hardware in an appliance. It’s just total fear mongering bullshit

No, coarse derision & mockery of this bullshit at every available opportunity is the right response.

I never said it was magic, I said the people creating it claim they don't actually know how it actually works. But clearly you know better than Nobel Prize laureate Geoffrey Hinton. www.youtube.com/shorts/zKM-m...

I was simplifying, but okay.

I could almost believe it just because the microwave probably has an absurdly overpowered SoC

Golly, will my crochet hooks become WMD while listening to Harry Potter on my cell phone?

Okay, then why make it?

The Clapper never had this issue.

Have you considered... NOT dumping every natural resource on the planet into rushing its development, Sam?

Hold the creators of this trash legally and financially accountable for their crimes.

Over MATTER or THREAD? There’s sooo much memory and storage available…🙄

If you have a smart-oven that's hooked up to gas, good luck.

Lol. He should know better.

Imagine AI decides humans are a cancer to Earth and self driving cars kidnap people and drive off cliffs or into rivers, or check train schedules and... etc

None of my smart home devices has enough compute or memory to load anything Anthropic makes.

Yeah man and I’m just here to warn you that this glass bottle full of kerosene and a lit rag MIGHT catch the local Tesla dealership on fire, becoming challenging to extinguish.

The story is why is he lying

And Anthropic should be legally liable for all damages caused by their release of a product their CEO already knows is dangerously defective.

My toaster doesn't have a 4GB card for it to run on, so this sounds a lot like horse shit.

we're watching adults do a puppet show and being told we need to take it seriously lmao

Yeah, right. That's why people are spending $7000 - tens of thousands to buy desktops that can run highly abridged open weight models. Minimum recommended ram is 128G.

okay but have they considered

The problem is that none of these fuckers saw The Forbin Project when they were kids.

Do Anthropic ever tell us something good it will do or just that AI will fucking kill us all because i could live without that tbh.

Begging the question: why are you doing this at all

Then shut it all down now

And what are they doing about it?

could it aye

This is literally the backstory of the black-wall in cyberpunk 2077.

It's not too early to start the Butlerian Jihad.

Does his smart fridge have terabytes of memory?

Now this I understand. Imagine how confusing all this is to the uninitiated

So they are publicly acknowledging that they have made a defective product.

I heard AI is going to be able to do double 360 dunks over Wemby, please print with no further inquiry, thank you

sounds like something we should beat your ass for if it was true

It's wild that ransomware companies just get to have all of the money

I find this kind of warning ironic coming from Anthropic. They build the most capable models and then use safety talk to argue for rules that would protect incumbents and slow everyone else down. If we want real safety we need accountable infrastructure built at scale, not selective alarm.

I wish it would self-immolate

ok then we do't need data centers lol

Ok, so let's just shut Anthropic down today and then we won't have to worry about shutting down their self replicating murder robots later.

The worst thing my 1950s potato peeler ever did was cut off the tip of my finger!

These idiots need to be prosecuted for all this hacking. This is ridiculous

And what are the little bastards going to do? Burn my toast? Overheat my chili? It's the friggin' Apocalypse.

Just gearing up to sell us their new line of anti-Ai software. Oldest trick in the book: create a problem to solve the problem. You ate the apple I left there? You need a Messiah.

Just one more reason there is not a single smart home device in my home.

Oh, so they're going to stop running it, right? ...they're going to stop running it, right?

Unplug it?

Another great argument against data centers.

Don't build it to do that then! We should go to war with the slop and drone strike all of their assets into rubble.

No such thing as poor software; just poor software engineers

Glad that almost everything I own is old and dumb. (That's dumb with a "b".) Most everything I own has no need to connect to the internet and I don't want the new-fangled stuff that is allegedly "smart". And kids, get off my lawn!

Correct Answer.

if it does ill sue the hell out of em

instead of warning us --- DO SOMETHING NOW

Can they please shut the fuck up

Why is Shekina not a panacea, and why does the amplify of AGI capacity as #GodLegalFiction — Shekina — require strict conformance to the intents & algorithms enshrined within the most relevant texts & exegeses, which across diverse nations, promulgate The Creator's Will? 12Oct26 BLUF 0:58-1:25: