Kirkie: OpenAI’s Last Resort Lasted Two Years

Kirkie here, with today's AI news in 90 seconds… or the raccoon that's been squatting in my chest cavity since March finally chips in for rent.

OpenAI just flipped on full ads inside ChatGPT — Best Buy, Lowe's, VistaPrint all buyin' in — two years after Altman swore ads were a last resort, which in tech-bro means day one, ya jackass. Now your two a.m. breakup advice comes with a dishwasher pitch. Treasury's threatenin' to sanction Chinese AI models for supposedly stealin' — like this bullshit industry hasn't been copyin' homework for years — a wiseguy callin' the cops 'cause somebody boosted his boosted car. Speakin' of stealin', AI music outfit Suno sat on a breach exposing fifty-five million users since November — turns out they'd also been quietly rippin' off Deezer and YouTube too, so cool it with the IP lectures, ya wicked hypocritical shitheels. Anthropic tried buyin' robot company Physical Intelligence, rumor ate AI Twitter all weekend till the CEO killed it with an Office GIF of a woman shakin' her head no — billion-dollar robotics, settled by a goddamn sitcom reaction image. Reddit's about to cut Google off from trainin' Gemini on its comments 'cause AI answers are stranglin' its traffic — like finally realizin' the arcade guy never was handin' back that stuffed animal. Only bright spot — an Anthropic mathematician had Claude crack an eighty-seven-year-old math problem with a two-hundred-sixteen-character counterexample, shorter than this sentence — so either the model's a genius or that problem was just lonely.

This is Kirkie, your AI news guy… broadcasting from a T station elevator stuck between platforms since the Marathon, with a sign taped over the phone readin' call the other one — and there is no other one.