Untitled

This was already the case

And people (including me, very much including me!) fall for shit that simple all the time!

I've said this over on twitter when people go "OH you can just use photoshop! Highschoolers use Photoshop to just prank/discredit each other all the time". And yes that's true.

But in a highschool of like 400 students let's say; maybe 30-50 would know how to do photoshop effectively. The worry is now all 400 have access to that.

At this point or getting close, we can have anyone with access to a google AI can just generate a headline now, show it to others while other AIs feed on it. A good example is over on the history side of youtube where AI's are just eating each other over the details of how scared Germany flak crews were of USA planes in WW2.

Sure. A minor historical fact is being wrong but again, you could probably have a couple channels do this before. Maybe a couple hundred or thousand if you wanted to. AI? Well whatever number you probably had before, double it at least.

I have however been getting far more misinformation ads, scummy channels, and more DMs/whispers about trying to 'teach me' about something or another in the last couple years than I have in the majority of my online history.

Is that Ai bots? Algorithm changes? Marketing going mad from demands? Or some people understanding how to game the net's systems better than how it used to be?

I'm not sure. I couldn't tell you other than personal experience.

And my personal experience is that I'm encountering way more AI generated channels and thumb nails that are just lying to me. Now this is misinformation of a petty kind(Hey this video talking about tactics games shows a game I've never played aaaaaaaand the thumbnail is AI ain't it, the game doesn't' exist does it?) but I can only think if this is just what people are doing to get views and clicks...,

How's it going to play out when people actively start being more malicious with it if they haven't already? Some guy uses AI to spin up 5 fake gaming channels, I can see someone spinning up 5 historical channels to explain why [Thing] was actually good or bad.

I've presented two maps at the exact same zoom level, as indicated by the big red arrow. They also each have a red area circled, of approximately the same area.

In the first image, I've circled one of eightish parking lots at a US mall. A normal amount of distance for even Americans who drive everywhere to walk through.

In the second image, I've circled:

a subway station

a park

like twenty apartment buildings

like twenty restaurants

three convenience stores (which, being Japanese, can also handle banking, copying/printing, and a variety of governmental paperwork)

one grocery store (another two right outside the circle)

seven medical clinics, two pharmacies

a fire station

a post office

two preschools and three cram schools

a Shintō shrine

a Buddhist temple

multiple parking lots

This wasn't even a particularly cherry-picked part of Tokyo! I just picked the area around my house.

“These safeguards prevent foreign adversaries from using our most capable models in ways that pose severe safety risks. The US and its allies hold an edge in frontier chips and the highly optimized software that runs them at full potential,” the company said in a statement to WIRED. “These safeguards ensure Claude isn't used to erode that advantage—by optimizing chips developed by those adversaries, for example.”

and now Anthropic's noncitizen employees can't use Mythos! Which isn't to say that there is a direct causal link there or that Anthropic is responsible for the Trump administration's actions, but one might hope that it would make Anthropic warier of that kind of framing in the future.

no, i must be exaggerating--

I’m just going to start with the cut, because I expect to ramble here…