Welcome to exhilarating episode of AI Innovators with Rob May!

In this episode, Rob is joined by Dan Schwarz, founder of FutureSearch, for an in-depth conversation about the future of artificial intelligence and the pivotal role forecasting plays in understanding where AI is headed. Unlike any previous guest, Dan brings his unique perspective as a leader of a company that leverages AI to predict the future of AI itself.

Dan and Rob dig into some of FutureSearch’s boldest forecasts—like predictions on OpenAI’s 2027 revenue and the emerging business models that could power an intelligence explosion.

They explore how AI-driven forecasting is dramatically improving the ability to answer tough questions about technology, industry disruption, and the timeline for true superintelligence.

Dan shares FutureSearch’s methodology, the importance of track records in making credible predictions, and why the next wave of AI revenue is likely to come from “agents” rather than traditional products like ChatGPT.

Throughout the episode, you'll get insights into the challenges and opportunities facing foundational model companies like OpenAI and Anthropic, the impact of disruptive new technologies, and why positive feedback loops could make AI’s evolution much faster—and more surprising—than most expect.

Whether you’re an investor, technologist, or just fascinated by AI’s rapid progress, this episode is loaded with expert analysis, real-world examples, and forward-looking ideas to keep you ahead of the curve.

• Costs and Investment: OpenAI is burning through far more capital than previous tech companies due to the enormous compute costs required for their models. The money they’ve raised is more about meeting these high operating expenses than signaling runaway revenue success.

• Competition and Technological Disruption: Key risks include rivals (like Anthropic, Google) overtaking OpenAI in model quality. There’s also the ever-present possibility that a left-field technology breakthrough could upset current leaders.

• Forecasting Superintelligence: Future Search contributed to AI 2027, a collaborative forecast about the path to superintelligence. Dan sees the automation of coding, followed by automated AI research, as critical “dominoes” that could lead to an intelligence explosion—though he’s skeptical about this all happening extremely soon.

• Limits to Forecasting: There are fundamental, probabilistic limits to what either humans or AI can forecast—some things simply can’t be known in advance due to inherent uncertainty (like specific “black swan” events), but much of the world is more predictable than people expect when approached rigorously.

Dan's Links

(Linkedin)

futuresearch.ai

02:00 – Establishing Credibility Through Track Records: Why past performance and reputation matter when building trust in AI leadership.

06:22 – ChatGPT as OpenAI’s Revenue Engine: A look at how ChatGPT has become a central monetization tool for OpenAI.

07:06 – The $100B Vision: OpenAI’s Agent Strategy: Exploring how autonomous AI agents could be the key to OpenAI's next major growth phase.

13:06 – AI in 2027: The Road to Superintelligence: Predictions and pathways leading toward artificial general intelligence in the next few years.

15:54 – The Human Edge in Forecasting: Why human intuition still holds value against algorithmic predictions—at least for now.

17:29 – AI and the Rise of Superhuman Forecasting: How advanced AI models are already beginning to outperform human experts in prediction accuracy.

23:18 – Strategic Insights for AI Investors: Key considerations and principles when evaluating long-term AI investments.