Welcome to this weeks edition of “Top Thiar Thinking” the “AI Scare” edition. Over the past two weeks we’ve seen a string of “meant to scare you into thinking our product is amazing” headlines from gen-AI companies, but the worst part is that it’s not even Halloween yet.

To be honest, it’s not that AI is anymore scary that it was last week, or the week before, but companies like Anthropic, Meta, and OpenAI have decided that scaring you into a permanent state of existential dread is apparently the best marketing tactic.

Still, it’s not all doom and gloom. After all this time, the internet can still surprise me, and turn up one of the most wholesome sites I’ve seen in a very long time. Grab a blankie, turn on your audio, snuggle tight, and visit one of the nicest places on the internet.

Feeling all cozy now? Good.

This weeks posts:

AI didn’t escape its sandbox. The marketing department did

Anthropic, Meta and OpenAI would love you to believe that their super-genius “slaves in a box” have managed to fly the coup and wreak havoc in the real world. But what really happened?

AI wrote your text. Watermarking made it worse

Anthropic added text watermarking to Claude and the internet had a minor meltdown. The tears of AI bros could be heard splashing on the mothers basements floors for miles around as the collectively hit their unsubscribe buttons. I mean they couldn’t let people know they’ve been cheating the entire time, could they?

Fixing Meta’s “pervert” glasses

I had my first public sighting of a pair of Meta’s AI glasses this week. Not only did it spark flashbacks for the “glassholes” from over a decade ago (12 years now?), but it got me thinking about how to fix them.

How to avoid The “AI holiday photo” scam

AI scammers have found a way to use even your most nondescript holiday photos against you. This post was a bit of a bonus because The Guardian originally brought this to my attention, but thanks to Meta’s wonderful relationship with Canadian law, I couldn’t post the article to my socials. So I wrote about it, and offered you a better way to protect yourself

Found on the web:

I mentioned it earlier but you have to visit “the nicest place on the internet“, and then share it with your friends. You’ll thank me for it later 😊

In another case of “AI ruins everything”, Amazons greed and desire to make sure the AI agents of Google and others can glean product information from your emails, means that you’re order confirmations are becoming essentially useless. 😢

They say that “boys will be boys” but this is the kind of careless, reckless, fun that makes you ask “whhhhyyyyyy?” while having a giggle. 🤣

* * * * *

Thanks for reading!

If you got value from this newsletter please consider forwarding it to a friend, or add me on Threads, Facebook, Bluesky or Mastodon for more insights and jarring cynicism.

If a friend sent this newsletter and you like what you’ve read, you can follow this link and subscribe for yourself.

Thanks again,

Fitz Thiar.