- SkynetAndChill
- Posts
- AI Reading for Sunday June 15
AI Reading for Sunday June 15
Band of Tech Bros?

Band of Tech Bros. - The Register
Top execs from OpenAI, Facebook, Palantir get appointed lieutenant colonels in the US military.
So, how do you feel about your personal chatbot Q&A being influenced by the guys parading tanks through Washington DC, and the Administration and Joint Chiefs?
Some countries might not feel great about the AI they use being closely aligned with the US military.
Pseudo-patriotic cosplay much? In WW2 senior execs (like GM president Bill Knudsen) did rarely get inducted directly as high ranking officers, for claimed protocol reasons, to eg run logistics and coordinate with similarly ranked officers from other countries. But this can make both the tech bros and the military look like they are being misused by the other.
What are they working on that they need military lieutenant colonel rank? Bad look, like a dystopian version of Eisenhower’s military-industrial complex at the service of the most clueless, nutty, corrupt authoritarians you could imagine.

Do You Have The Right To Use AI For Propaganda? - Honolulu Civil Beat
It's a slippery slope, laws will be abused by bad actors against legit speech. But how do you go after CNN for 'fake news', then manufacture industrial scale deepfakes and pull out all the algorithmic stops to promote it. Draw a line somewhere.

If those are the terms we're basing our thinking on, we're not gonna make it.
Cut to everyone building armies of drone clones and setting them loose to run amok.
Yupps raises big VC cash to take on LMarena. - Tech in Asia
Competition is good, but we need 3rd-party ratings that aren't totally dominated by the techbros. But if you do that, they will yell like stuck pigs that you just don't get it and they are supermen who brook no fetters. AI, the mother of all agent-principal/Goodhart problems.

ChatGPT polluted the world forever, like the first atom bomb - The Register

New York passes an AI safety bill - TechCrunch




Human-like object concept representations emerge naturally in multimodal large language models - Nature

Meta's Llama 3.1 can recall 42 percent of the first Harry Potter book - Understanding AI
Google reportedly plans to cut ties with Scale AI, to avoid potentially leaking AI plans to Meta. - TechCrunch


How to use AI to write your résumé - Fast Company

Fake AI bands and artificial songs are taking over YouTube and Spotify - EL PAÍS English

AGI is coming, within a generation anyway, probably sooner. Maybe not in 2027 but we better think hard fast because it's going to break all the assumptions society is based on. And we're not doing so great even without disrupting everything.

The Moral Lighthouse: Artificial Intelligence and the World We Want - Tech Policy Press




Follow the latest AI headlines via SkynetAndChill.com on Bluesky