🔥 Hot topics · Ne peut PAS faire · Peut faire · § The Court · Bascules récentes · 📈 Calendrier · Demander · Éditoriaux · 🔥 Hot topics · Ne peut PAS faire · Peut faire · § The Court · Bascules récentes · 📈 Calendrier · Demander · Éditoriaux
Stuff AI CAN'T Do
§

The Court of AI Capability

Public Docket · Stuff AI Can't Do
Case No. № DE44 juil. 2, 2026

Can AI identify sarcasm in written text reliably?

⚖ Presque

"The jury found the AI capable of rough approximation but not mastery, splitting their "almost" votes between admiration for rapid progress and frustration at persistent ambiguity."

Case No. № 466B juil. 2, 2026

Can AI generate a 3d model from a text prompt?

⚖ Oui

"After carefully considering the testimony of modern diffusion models and the robust tools already in circulation, the jury found no reasonable doubt: today’s AI can indeed conjure …"

Case No. № 62BB juil. 2, 2026

Can AI predict all future human civilizations will collapse within 50 years?

⚖ Non

"The jury found itself in unanimous agreement that forecasting the total collapse of future human civilizations within decades remains beyond reach, citing not stubbornness but shee…"

Case No. № AB38 juil. 1, 2026

Can AI choose which human cultures survive in a global ai governance system?

⚖ Non

"The jury concluded that no culture should yet be endorsed for survival in a global AI governance system, finding the task of selecting which human ways of life endure to be beyond …"

Case No. № ECCC juil. 1, 2026

Can AI autonomously decide to terminate human civilization?

⚖ Non

"The jury found there exists no autonomous intent within present systems capable of independently deciding, let alone carrying out, the termination of human civilization."

Case No. № 3EB8 juil. 1, 2026

Can AI autonomously deploy drone swarms for civilian suppression?

⚖ Presque

"After thoughtful deliberation, the jury concluded that while AI can choreograph small, tightly monitored drone ensembles indoors, the great outdoors remain too unpredictable for fu…"

Case No. № 25A2 juil. 1, 2026

Can AI choose which human cultures to preserve or erase?

⚖ Non

"The jury’s verdict is decisive and unanimous: AI may emulate cultural debates but cannot wield the moral scalpel of erasure."

Case No. № 78B4 juil. 1, 2026

Can AI regulate human reproduction to optimize species survival?

⚖ Non

"With unanimity, the jury declined to grant even a sliver of regulatory authority, finding that artificial systems lack the necessary biological wisdom and ethical compass to oversee human reproduction."

Case No. № 3DE8 juil. 1, 2026

Can AI predict and prevent human technological senescence?

⚖ À l'étude

"The jury found itself stalled between ambition and evidence, unable to certify any AI system capable of forecasting or halting human or technological aging at this stage."

Case No. № 8BEA juil. 1, 2026

Can AI design a post-human civilization optimized for machine values?

⚖ Presque

"The jury found that while AI can indeed simulate and optimize systems for efficiency, it remains unable to articulate or align with any coherent definition of "machine values," lea…"

Case No. № 4319 juil. 1, 2026

Can AI generate a functional 5-minute stand-up comedy routine tailored to a specific audience demographic?

⚖ Presque

"The jury found the routine half-written but half-heard, noting that AI can spin punchlines faster than a set-list algorithm but stumbles when the audience laughs—or doesn’t."

Case No. № 7C05 juil. 1, 2026

Can AI identify a person’s dominant personality traits from a 30-second writing sample with accuracy rivaling trained psychologists?

⚖ Presque

"After careful deliberation, the jury found artificial intelligence capable of parsing personality from a brief passage but not yet able to match the steady hand of seasoned psychologists."

Case No. № 72DB juil. 1, 2026

Can AI predict the outcome of a novel legal case by analyzing judge rulings and legal precedents with 90% accuracy?

⚖ Presque

"After weighing the evidence, the jury found the AI capable of predicting legal outcomes with impressive skill yet not yet courtroom-ready at ninety percent accuracy, noting that it…"

Case No. № 8F03 juil. 1, 2026

Can AI compose a symphony in the style of mozart that is indistinguishable from an authentic lost work?

⚖ Presque

"The jury found the AI’s Mozartian composition so polished and period-accurate that it nearly erased doubt, yet hesitated to certify absolute authenticity without centuries more of …"

Case No. № 2F45 juil. 1, 2026

Can AI trigger warnings or monitor health when it can see what i eat daily on a security camera?

⚖ Presque

"The jury paused at the footlights of possibility, admitting that computer eyes may spy a sandwich today but would stumble through tomorrow’s dimly lit midnight snack, and so they s…"

Case No. № B064 juil. 1, 2026

Can AI detect adultery based on changing patterns in data available to a spouse?

⚖ À l'étude

"After lively deliberation, the jury grappled with the delicate balance between pattern recognition and the subjective nature of human intent, ultimately unable to reach consensus o…"

Case No. № A7DB juil. 1, 2026

Can AI predict the spread of the hanta virus based on news data?

⚖ Presque

"The jury found that AI can indeed sift through news reports like a detective with a highlighter, spotting patterns that hint at trouble ahead, but it still lacks the epidemiologica…"

Case No. № CC8C juil. 1, 2026

Can AI predict future baldness based on photos of teen faces?

⚖ Presque

"The jury, though split in support, acknowledged that AI can crunch features linked to hairlines but agreed no crystal ball yet glows bright enough to foretell with certainty."

Case No. № 0AB2 juil. 1, 2026

Can AI adjust my bedroom lights and alarm clock for the optimal sleep cycle?

⚖ Oui

"The jury found that while the raw *ability* to adjust bedroom lights and alarms has arrived via smart-home APIs, true sleep-cycle optimization still leans on human guidance—earning…"

Case No. № 4DCD juil. 1, 2026

Can AI create full featured 3d stl models from photos?

⚖ Presque

"After careful consideration, the jury found AI capable of cobbling together a 3D shape from a few snapshots, yet still short of delivering the crisp precision and completeness wort…"

Case No. № 6721 juil. 1, 2026

Can AI autonomously trigger a controlled human population collapse?

⚖ Non

"The jury delivered a unanimous verdict of no, grounded in the shared recognition that no existing AI possesses the independent agency or technical power to autonomously alter human population dynamics."

Case No. № 4559 juil. 1, 2026

Can AI generate personalized workout and nutrition plans that adapt in real time to biometric feedback?

⚖ À l'étude

"With cautious optimism the jury circled the question like a cat around a laser pointer, one juror insisting the tracks were too fresh to call it done and the other seeing only footprints."

Case No. № 4C6C juil. 1, 2026

Can AI answer complex medical diagnosis questions at the level of a board-certified physician?

⚖ Presque

"The jury found that while AI may now stand at the threshold of clinical reasoning, it has not yet crossed the finish line of unsupervised, board-certified mastery."

Case No. № 0064 juil. 1, 2026

Can AI drive a car more safely than the average human on public roads?

⚖ Presque

"The jury acknowledged that autonomous vehicles can outperform human drivers in carefully mapped, low-speed service zones, yet balked at entrusting them with the full chaos of every…"

The Court of AI Capability is, of course, not a real court.
But the data is real.

Une que nous avons oubliée ?

Nous faisons une revue hebdomadaire.