Jim Perrin

@jperrin.bsky.social

Something something at Microsoft something something. Life coach for terrible ideas

oh my goooood the eye-roll convulsing my face hearing the news cover openAI’s “security incident” CLEARLY using a bunch of language from their own press releases about its “autonomous agent” “going rogue” JUST SO THEY CAN HINT MORE THAT IT’S ACTUALLY FULL ARTIFICIAL INTELLIGENCE (it is not)

In 60 days or so, we’ll face an election we can’t successfully win without a landslide. Any low-percentage win and they’ll do what we know they already know exactly how to do, and more. We’ve all seen it happen. So we need that landslide victory. We all must vote.

Budgets are intentionally unrealistic on the low side to get projects approved in the first place. Sales targets are intentionally unrealistic on the high side to get projects approved in the first place. This guarantees an unhealthy ecosystem. It's bogus numbers all the way down.

Jason Schreier@jasonschreier.bsky.social · last mo.

FWIW I think it's a little strange that "Arkane's Blade is late and over budget" is becoming a lingering news story when the total number of big video games that were *not* late and over budget could be counted on one hand

Also if anyone wonders if SpaceX revenue growth will make the difference: well MSFT has managed to double revenue every ~5 years for THIRTY YEARS to get to this position. So doubling six times. And mostly with a relatively capital light model? Good luck matching that!

NEW: malware developers added nuclear & biological weapons text to to their spyware. Goal? To trigger LLM safety refusals... so that their spyware wouldn't be analyzed by an AI security scanner. Cleanest practical example I can think of for why over-indexing on first order "safety" is risky. 1/

Bild