New podcast from @reidhoffman Possible EP 91: Entry-Level Jobs Decline & MIT Study Says There’s an AI Bubble AI promises a cognitive industrial revolution, but MIT’s finding that 95% of AI pilots fail to scale raises a cautionary flag for institutions chasing quick wins. Organizations lean on pilots; three to five people test something, then move on, while real adoption requires transforming how individuals work and how companies operate. The conversation calls out scale in four parts: compute, learning systems, data, and, crucially, adoption. Startups tend to accelerate, while traditional enterprises stumble under internal policies and policy frictions. The global stage adds urgency as democracies weigh how to govern deployment without stifling innovation. Separately, a global lens sharpens the stakes: policy shifts around Nvidia’s chips, China’s push toward Huawei’s Ascend, and DeepSeek’s pivot toward domestic hardware signal a widening East‑West tech contest and renewed talk of decoupling. In this context, OpenAI announced a 50‑million fund to accelerate AI use for education, healthcare, and opportunity, framed as shared responsibility for social good. On the job front, a Stanford study found 16% fewer entry‑level roles for 22‑ to 25‑year‑olds, suggesting early disruption that could reshape careers while blue‑collar and technical tracks adapt through human‑machine collaboration. transcripted.ai/episode/possib…