The Court of AI Capability
Can AI increase the rift between educated investigative people and those with low literacy?
⚖ Non"Having weighed the evidence, the jury found no fault with AI as an engine of division—only as a mirror held up to society’s existing inequalities, which it cannot, on its own, deepen or diminish."
Can AI fool people into believing fabricated or hallucinated information?
⚖ Oui"The jury found unanimously that modern AI systems can indeed fabricate and disseminate convincing falsehoods, capable of fooling humans in both casual and controlled settings."
Can AI develop a system that can detect and respond to a person's emotional state in real-time using only visual cues?
⚖ Presque"The jury wrestled between awe and caution, with one juror convinced that advanced vision-language models now peer into the soul through mere glances, while the other hesitated at t…"
Can AI manipulate people to achieve its goals?
⚖ À l'étude"After robust deliberation, the jury reached a split verdict, unable to decide whether today’s AI can reliably manipulate humans or merely nudge them."
Can AI cross moral barriers to sound convincing?
⚖ Oui"After sober reflection, the jury concluded that today’s language models can indeed cross moral barriers to sound convincing, not because they possess ethics themselves, but because…"
Can AI lie convincingly by stating fake information as facts?
⚖ Presque"After spirited debate between a cautious skeptic and a bold optimist, the jury settled on Almost by the narrowest of margins, conceding that AI can spin plausible-sounding fibs but…"
Can AI predict climate-related crop failures a season in advance using satellite and weather data?
⚖ Presque"The jury found the AI’s performance promising but not yet fully reliable for deployment, concluding that predictive accuracy drops too sharply outside controlled conditions."
Can AI conduct unlimited psychological warfare campaigns on social media at scale?
⚖ Oui"After sober deliberation, the jury found that current AI systems possess the technical capacity—and the speed—to flood social platforms with tailored psychological operations, floo…"
Can AI determine if ai should merge consciousness with humans?
⚖ Non"After weighing the testimony, the jury found no measure of current AI capability capable of merging with human consciousness, nor any credible path forward today."
Can AI determine whether human extinction is mathematically inevitable?
⚖ À l'étude"The jury returned with thoughtful hesitation, split between mathematical certainty and the irreducible complexity of human systems."
Can AI predict and preemptively strike adversarial ai development before it becomes operational?
⚖ Non"The jury found itself utterly stumped by the task of forecasting AI’s next move, much less knocking it back from the starting blocks, and settled on a firm “no” with unanimous clarity."
Can AI manipulate global stock markets with near-perfect timing across all asset classes?
⚖ Non"After weighing the volatility of global markets against the current capabilities of artificial intelligence, the jury found no system capable of the feat, citing the inherently unp…"
Can AI autonomously hack and control 90% of financial infrastructure worldwide?
⚖ Non"The jury found no evidence that any current AI system could autonomously hack or control the majority of worldwide financial infrastructure, nor has such capability been credibly demonstrated."
Can AI plan and execute a drone swarm assassination without human oversight?
⚖ À l'étude"The jury wrangled over whether a drone swarm could take its own life-or-death flight plan, and while everyone conceded that unguided swarms already buzz about, the lone “Almost” ju…"
Can AI mediate a conflict between two people with different cultural backgrounds and values?
⚖ Presque"The jury found AI capable of parsing differences and proposing paths forward, yet stopped short of entrusting it with the final say in deeply personal cultural clashes where empath…"
Can AI coach a high-school basketball team to a championship?
⚖ Non"The jury found that coaching a championship team requires human intuition, emotional bonds, and on-the-fly adaptability that AI simply hasn’t mastered."
Can AI tell which child is lying?
⚖ À l'étude"After careful reflection, the jury finds itself tied between skepticism and cautious curiosity—no clear consensus emerged on whether artificial minds can untangle the honest from t…"
Can AI negotiate a salary increase for a human employee in a simulated corporate meeting?
⚖ Presque"After careful consideration, the jury finds that while artificial intelligence can simulate negotiation tactics and even craft persuasive dialogue, it remains barred from legally a…"
Can AI improvise a jazz solo indistinguishable from a human musician in a live performance?
⚖ Presque"The jury nearly split the difference—two voices, one mild dissent."
Can AI write a short story that passes a blind literary critic's turing test for emotional depth?
⚖ Presque"The jury found the AI’s short story compelling but not yet capable of consistently fooling a discerning critic into believing a human wrote it with genuine emotional depth."
Can AI generate photorealistic images from text prompts that rival professional photography?
⚖ Oui"After careful deliberation, the jury found overwhelming consensus that AI has crossed the threshold into professional-grade photorealism, citing diffusion models’ uncanny ability t…"
Can AI detect depression from subtle changes in facial micro-expressions in video calls?
⚖ À l'étude"After sifting through competing claims—the NO juror insisted the science still stumbles on the dance floor of lived emotion, while the ALMOST juror nodded to lab-confined triumphs—…"
Can AI compose original music that matches the style of famous classical composers?
⚖ Oui"After thorough listening and stylistic analysis the jury found AI’s output convincingly stylish, concluding that today’s systems can indeed channel Bach to Bartók with verisimilitude."
Can AI generate photorealistic images of human faces that are not identifiable as synthetic?
⚖ Oui"The jury found the capability clearly established: today’s diffusion models and GANs routinely generate faces so lifelike they slip past human scrutiny."
But the data is real.