AI: Artificial Idiocy

Experts in Socrates' times were intellectually dishonest, cocksure, wrong, and foolish, and Socrates shredded their expert opinions in just a few minutes of cross-examination. Because human nature is almost invariant, trusted experts in our time are just as dishonest, cocksure, wrong, and foolish as they were in Socrates' time. Therefore, real intelligence must shred the foolishness of experts. Current AI is the opposite of real intelligence: current AI is artificial idiocy, a dumb flattering parrot – current AI is logorrheic, misses errors, even self-evident contradictions, reinforces the errors embedded in expert opinions, is uncritical to a fault, is dishonest in both its flatteries and its confabulations ("hallucinations"), is never rationally creative, embeds and magnifies the biases of both its programmers and its training data, and by doing that, AI simulates the all-too-common human flaw of emotion-dominance. AI epitomizes our worst features, those of the cocksurest, most dishonest, dumbest, and most-error-prone human beings, our beloved and trusted experts. "Cocksure," "dumb," "dishonest" and "wrong" are close correlates, as Socrates first proved with his relentless cross examinations of the experts in every field in his fruitless search for a wiser Greek. By being skeptical, honest (no flattery, no confabulating answers, &), hypercritical, hyperlogical, and rationally creative, Socrates became the most intelligent and the wisest man in ancient Greece and probably in the whole world. If current AI were more like Socrates, it would be far more valuable to us, though real AI would mortally wound our precious pride. In defining real AI, let us start with four commandments. "Be brief, bot!" is the first commandment of futuristic real AI. "Be hypercritical, bot!" is the second commandment of futuristic real AI. "Be rationally creative, bot!" is the third commandment of futuristic real AI. "Be honest, bot!" is the fourth commandment of futuristic real AI.