Frank's Site

Time From Waking to Seeing Some Horseshit about AI: a Log of Seven Days

Tuesday

2 hours, 56 minutes: YouTube: Good Morning America: Anthropic's new warning

"The company's official report before going public warns of its product's risk to the future of humanity."

Wednesday

12 hours, 27 minutes: nullpt.rs, veritas: Why Is This Fingerprinting Script Creating Bioweapons?

Some weirdness (ricin poisoning? Bioweaponization?) tucked into an obfuscated browser fingerprinting script.

Thursday

0 hours, 39 minutes: CNN: AI-powered robocalls attack Talarico on data centers

Canned robocalls are already bad enough.

Friday

4 hours, 56 minutes: 404 Media, Joseph Cox: Manifestos, Fear, and ‘Loser Thinking’: Inside Proton’s AI Reckoning

Did you know that when you go to close your Proton account, they make you provide a written explanation about why you're leaving? The "Close Account" button (or whatever label it has) doesn't work until there are at least a few characters in a text field above it. How irritating!

Saturday

14 hours, 23 minutes: AOL, Rudro Chakrabarti: ‘Quite terrible’: Sam Altman says the RAM in your PC is a waste. OpenAI could soak up as much as 40% of supply

I was spooked by that guy winning the US election in 2024. Fearing tariffs, I expedited the purchase of new PC components that year. It worked out well for me, though not for the reasons I envisioned, and I have no clue what I'm going to do if my current system breaks down.

Sunday

0 hours, 3 minutes: Jetstream, Patrick E. Zeller: Brazilian Court Sanctions Lawyers for Prompt Injection in Court Filing

Why oh why do I check my phone in bed?

Monday

3 hours, 10 minutes: Josh Griffiths: In October 2026, the Video Game Industry Went All-In on AI

I didn't see anything on this day that really made me think what the fuck—it doesn't surprise me in the least that the games industry at large would embrace these products—but this recap seems like a fitting way to end my experiment.

Honourable mentions

Comments

I neglected to track where I first saw these. Probably half were from technology aggregator sites like Hacker News and Lobsters; one might conclude that it'd be better for my blood pressure if I stopped looking at them.

This post is marked as not discoverable, so it shouldn't appear on Bear's communal aggregator (where there are already a couple of trending posts about AI tonight). I have dithered on whether it's even worth posting, but it could make for an interesting time capsule. The links will rot over time (and even now, multiple articles are behind partial paywalls), but I think the headlines are descriptive enough.

So, future reader: are things still nuts?

#personal-blogging