#43 - AI Boom or Bubble, Poetry to Platforms, Co-op and Reflective AI, interview with Stephen Klein

Stephen Klein, a guest on "Salesforce Trails & Trials," shares his 20+ year journey and AI insights. Klein's background includes a fine arts degree in poetry, valuing literature and art for distinguishing humans from AI. His diverse career path encompassed journalism (UMass Daily Collegian editor), early tech at Wang Laboratories, 12 years in advertising (built largest New England agency, European ops), global product/marketing at Dun & Bradstreet, and an MBA from Harvard. He now teaches AI strategy and ethics at Berkeley, his primary expertise. Klein emphasizes wisdom (hard-earned principles from challenges/study) is distinct from knowledge (easily acquired information). He believes ethics and values optimize profit, contrary to the Machiavellian focus on fast gain. Klein critiques the current AI landscape, noting incentives are misaligned as AI is sold for cost-cutting and automation, leading to layoffs. He claims ~80% of industry data is "cooked," acting as a sales funnel for private AI firms and large consulting groups. He likens the current AI craze to historical bubbles like "tulip mania" and the dot-com era, fueled by fear of being left behind. Klein draws parallels to the sugar and tobacco industries' past efforts to distort reality through marketing. Generative AI is unreliable (30-70% hallucination), making full automation unfeasible. Using Gen AI to create more data results in a "recursive" loop, degrading training data and output over time. He states there are no net US job losses from Gen AI; layoffs follow normal 10-12 year economic cycles. The consulting industry profits heavily, with 70-80% of AI pilots failing. Klein advocates for an "augmentation" model where AI "levels up" people for strategic thinking, rather than replacing them. He champions "Co-op AI" or "Reflective AI," where AI partners with individuals for better problem-solving. His company, Curiouser.ai, offers "Alice," a Reflective AI that asks questions to challenge critical thinking and foster self-understanding, combining generative AI with tailored machine learning. Effective AI adoption requires strong CEO leadership, clear vision, organizational unity, and prioritizing people. Klein deems "AI first" the "stupidest idea" as it implies "people last," alienating employees and customers and harming brand reputation. He is optimistic that "people first" companies, investing in their future and workforce, will lead a positive revolution, building businesses and creating jobs. Organizations should consider multi-LLM, open/closed-source platforms, not just IT-delegated tools. I hope this concise summary is helpful! Would you like to review specific aspects of Stephen Klein's career, or perhaps delve deeper into his concept of "Reflective AI" and "Co-op AI"? I can also quiz you on the material to test your comprehension if you'd like!