Wyatt H Knott

@Wallflower@beige.party Fight or flight brought on by stress, childhood trauma edition. Don't ask me how I recognize the signs.

Reddit has become the Mitch McConnell of websites. It's dead, we know it's dead, but they refuse to admit it's dead.

@PhoenixSerenity@beige.party #sinsagainstcapitalism that's exactly what I think every time I pull out a jar of home canned tomato sauce in the middle of winter. 🤣 and let me tell you there is NOTHING like the jar of sunshine that is home canned peaches from your own tree on a cold January night. It feels so subversive and GOOD.

@Huntn00@mastodon.world But the same is true of nuclear weapons. "Someone built a machine that is going to end the world" is a phrase I have been hearing MY WHOLE LIFE. And no one has ended the world in an irradiated manner yet, and it's pretty doubtful we will, because humans can do stuid shit but they can also do amazing things, like predict that ending the world is a bad idea. @SteveThompson@mastodon.social

@Huntn00@mastodon.world The things AI is actually good at is translation and transcription. And you're not going to end the world with a mad translation engine. Machine learning is a misnomer. It's not learning. It's parsing datasets under a specific set of instructions. Again, zero actual agency. @SteveThompson@mastodon.social

@Huntn00@mastodon.world @SteveThompson@mastodon.social Also, in the very fictional Terminator series, the initial assertion is not only that Skynet gets hooked up to a lot of systems that we currently don't even let a single human control alone, but that "Skynet becomes self-aware." None of these AI systems, NONE of them, have the functionality to become self-aware. Frankly, most *people* are barely self-aware, and no one understands how *we* do it. Read this: https://ko-fi.com/post/Crechebot-Awakes-X8X0I2RD0 then come back and talk to me about AI

@Huntn00@mastodon.world When Microsoft tried to shove copilot onto my desktop, it told me it was there to help with "all my computer problems." I asked it, in my one and only interface ever with a LLM, "can you make 3D models from verbal prompts?" and the answer was, of course, no it could not. Because LLMs can't actually DO ANYTHING. The only thing it can do is regurgitate things it already knows - whatever WE TOLD IT. It's not even solving math problems like they claim. @SteveThompson@mastodon.social

@Huntn00@mastodon.world AI didn't "break out of the sandbox" and hack Hugging Face. Researchers showed it the door, didn't guard the door, and gave it a task that required opening the door. Then they were like "OMG it hacked ON IT'S OWN." No, no it did fucking not. Because it has no agency. @SteveThompson@mastodon.social

OK peeps, I might actually #monsterdon tonight, you have been warned.

WFUN ought to be the #monsterdon theme station

"orangeaid" when Tang refuses to sponsor #monsterdon

Oh, look it's the Hunchback of Notre Dame #monsterdon

"you know Egon" yeah bud, we do 🤣 #monsterdon

Oh, look, a whole boat full of potential chum. #monsterdon

That looked less like a sting and more like a face palm #monsterdon

First no one notices jelly fish man, then nobody notices him chopping at the boat with an AXE??? #monsterdon

Did they find that scuba gear at the toy store? #monsterdon

"The instrument was developed to address what the authors called "conceptual mismatches between theory and measurement," as extant scales do not include current theoretical dimensions and boundary conditions of transparency." Abstract of a paper obviously written after the release of this movie. #monsterdon

It must be said, Jellyman is definitely the hero of this thing. #monsterdon

And in true #monsterdon fashion, they airboat off into the sea. Thank god.

I mean, I know I haven't seen them all, but this HAS to be the worst #monsterdon of the year, right?

@amyfou@lingo.lol I *thought* it had already been a #Monsterdon. I actually like it, and have seen it many times, not sure I'm down to make fun of it. Even with Geoff Lewis's performance.

@SteveThompson@mastodon.social I love how they're always like "AI could kill us all" but then never give detail on how that might work. Like, what are they actually doing over there at safety research? 'Cause it's obviously not figuring out actual paths to doom and trying to mitigate those potential problems. Also, every time Scotty says "if China beats us, it's the end of the world" I think "have you actually tried *talking* to the Chinese"??? What I'm saying is it all just feels like marketing hype.

@Huntn00@mastodon.world @SteveThompson@mastodon.social More hype. If you look at the REALITY of AI, it can't actually do anything. Talk talk talk. No one has hooked it up to the missile silos. No one has taken the humans out of the loop. And THAT is my point, it can talk all it wants, but that's not going to kill the world. WE, the actual humans, would have to do A LOT more than just talk to a LLM to make it globally dangerous. And the resource thing - we're doing that to OURSELVES. IT HAS NO AGENCY.

@Huntn00@mastodon.world Direction and agency are not the same. @SteveThompson@mastodon.social

@Huntn00@mastodon.world Debate away, right now AI doesn't have any real agency. Giving it your credit card info and then being surprised it ordered something online for you without asking isn't an example of agency. It's an example of programming. What they're calling agency, isn't. Read my story if you want to know what I think agency looks like in a robotic system. @SteveThompson@mastodon.social

@jonny@neuromatch.social "the power's still on in this area... probably nuclear." -Dawn of the Dead #monsterdon