Stuff AI CAN'T Do
§

The Court of AI Capability

Public Docket · Stuff AI Can't Do
Case No. № 4B66 ago. 14, 2026

Can AI recognize human emotions?

⚖ Casi

"The jury leaned toward Almost, finding AI capable of reading emotional signals but still missing the deeper, lived meaning behind them."

Case No. № 615D ago. 14, 2026

Can AI recommend personalised medical treatments based on patient history?

⚖ Casi

"The jury agreed that while AI can sift through vast datasets to suggest potential treatments, it still lacks the decisive clinical authority to autonomously tailor care with unshakable confidence."

Case No. № 2938 ago. 14, 2026

Can AI generate a 10-minute educational video from a textbook chapter?

⚖ Casi
Case No. № BEF0 ago. 14, 2026

Can AI translate poetry while preserving rhyme and meter?

⚖ Casi

"AI can turn verse into rhyme—sometimes with startling grace—but the jury found the magic too fleeting to declare a full translation victory, concluding the art remains stubbornly b…"

Case No. № 30DB ago. 14, 2026

Can AI generate code review comments on production pull requests?

⚖ Casi

"After lively debate between practical utility and cautious precision, the jury landed just shy of full acceptance, acknowledging AI’s sharp eye for low-hanging issues but stopping …"

Case No. № 8486 ago. 14, 2026

Can AI generate a full website mockup from a one-paragraph description?

⚖ Casi

"The jury grappled with whether a one-paragraph description could truly produce a *full* website mockup—not just a sketch but a working prototype."

Case No. № 8389 ago. 14, 2026

Can AI generate a 60-second commercial-quality video from a prompt?

⚖ Casi

"After careful deliberation, the jury concluded that while AI can stitch together moments of visual promise, it stutters before sixty seconds of seamless polish."

Case No. № 6BBB ago. 14, 2026

Can AI score in the top 1% on math contests up to amc 12 level?

⚖ Casi

"The jury marveled at the AI’s lightning-quick algebra and geometry, yet paused before awarding full marks: even the brightest silicon scholar stumbles on the subtlest word-problem …"

Case No. № 6FE5 ago. 14, 2026

Can AI describe humanity to an alien race based on the entire recorded history of mankind?

⚖ Casi

"The jury found that artificial intellects can arrange chronicles like beads on a thread, yet still fumble the silk of meaning, leaving gaps a human weave might never miss."

Case No. № C703 ago. 14, 2026

Can AI generate a credible scientific hypothesis from raw experimental data?

⚖ Casi

"After spirited deliberation, the jury acknowledged that artificial minds can indeed sniff out hidden patterns in raw datasets and whisper fresh hypotheses like lunar moths flutteri…"

Case No. № E3DC ago. 14, 2026

Can AI emerging health issues from smart watch data?

⚖ Casi

"After careful deliberation, the jury acknowledged AI’s sharp eye for spotting anomalies in smartwatch streams but hesitated when the evidence didn’t yet carry the weight of a doctor’s nod."

Case No. № B8D4 ago. 14, 2026

Can AI operate as an urban planner?

⚖ Casi

"The jury found itself cautiously optimistic, though not quite ready to fully anoint artificial minds as city architects."

Case No. № 7F39 ago. 14, 2026

Can AI communicate or interact with animals in any meaningful form?

⚖ Casi

"The jury found the AI capable of parsing animal signals but not yet capable of genuine dialogue, with one dissenting voice insisting that no meaningful interaction has been achieve…"

Case No. № 859F ago. 14, 2026

Can AI predict wildfire outbrakes based on sattelite imagery, weather patterns and historical data?

⚖ Casi

"After thoughtful deliberation, the jury agreed that AI has made impressive strides in wildfire prediction, uniting satellites, weather, and history with remarkable precision in controlled studies."

Case No. № BC1C ago. 14, 2026

Can AI detect voter fraud by analyzing patterns in absentee ballot signatures across?

⚖ Casi

"After careful weighing, the jury found that while AI can spot oddities in penmanship with some skill, it cannot yet be entrusted to prove fraud beyond legitimate doubt—too many sig…"

Case No. № 7326 ago. 14, 2026

Can AI outperform humans at predicting protein-protein interactions?

⚖ Casi

"After spirited deliberation, the jury stalled at “almost,” split between two camps: those who saw AI outrunning human heuristics with 90% precision and those who recalled the stubb…"

Case No. № 4110 ago. 14, 2026

Can AI diagnose certain rare diseases from electronic health records?

⚖ Casi

"The jury found cause for cautious optimism but stopped short of full endorsement: two jurors marveled at targeted successes in narrow settings, while one pointed to credible eviden…"

Case No. № DE44 ago. 14, 2026

Can AI identify sarcasm in written text reliably?

⚖ Casi

"After three days of deliberation the jury agreed that artificial intelligence can often catch the eye-roll hiding between the lines, yet still stumbles when sarcasm wears the thinnest of disguises."

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

¿Nos faltó uno?

Revisamos semanalmente.