AI revenue isn't there and might never come, says NYU professor Gary Marcus

In this interview, David Faber speaks with AI expert Gary Marcus about the current state and future prospects of AI technology. Marcus expresses skepticism regarding claims of imminent, significant GDP growth from AI, citing issues like unreliability, hallucinations, and persistent errors in current models like Grok 3. He argues that deep learning alone is insufficient for achieving artificial general intelligence and that current AI models lack the robust reasoning capabilities needed for widespread enterprise adoption. Marcus also suggests that the financial struggles of companies like OpenAI, despite high valuations, may push them towards surveillance as a revenue model due to the high operational costs and reliability issues of their AI products.