AI has been operating under the "move fast, break things" mindset - they're sucking up everything and then hoping they can fix it all on the backside with iterative refinement. At some point, the future of AI will have to come down to biasing input towards "known good" non-AI-generated content.

Well now the admin is calling it a "glitch" - I am sure Signal hates that take haha.

A fox is in the hen house.

If I am being real, part of it is stubbornness. People are comfortable with their castles and moats, even though there are 300 bridges and tunnels into the castle, even if they spend most of the time shoveling the stables, they're at least comfortable doing that stuff. It's familiar, at least.

Obviously it takes time and resources to adapt to new technologies, but the thing about the new technologies is they often are much simpler and ultimately give you time back. That's why it's debt. Every time you go through some goofy manual process, that's an interest payment.

Yeah, but trying to de-google my phone and browser (using Thorium) has made me realize all the subtle ways they're able to collect "anonymized usage statistics" about my usage/data. Keyboard auto compete, search, voice assistant, genAI assistants, play store, maps, even the launcher... Its so much.

I feel like the rise of AI, coinciding with the rise of authoritarian right wingers, is making a lot of people reconsider security and privacy. For one, I've always been big on opsec but privacy wasn't a huge concern... but AI tools make it a lot more sinister so I'm breaking up with Google.

It's certainly suspicious they would rather the banned and lose it all then sell and take a minority stake.

Web 2.0 was a mistake.

Tulsi Gabbard, obvious Russian plant, is told that she needs to be a little less obvious. #spycraft