"AI Exposed" Part I: The Simple Truth

WethosAI founder Stuart McClure, joined by CTO Alen Capalik, opens the "AI Exposed" webinar series with a case against the current AI hype, framing it through Kahneman's System 1 versus System 2 thinking and arguing that today's models reflect optimism bias more than genuine reasoning. Drawing on their cybersecurity backgrounds, they trace the frothy economics of AI (500-billion-dollar bets, sky-high valuations, companies losing billions, 42 percent of US venture funding) and catalog the fears it drives, from job displacement and misinformation to embedded bias, privacy, and energy demands. Live demos expose concrete failures: inconsistent letter-counting, undetected six-fingered hands, botched object-stacking logic, confident hallucinations, name-blocking from lawsuits, and demographic bias in image generation. McClure closes on the brain's vast energy and learning efficiency over transformer models, floats fixes like LLM routers, RAG, and fine-tuning, and previews Part II on logic and reasoning.