Richard Campbell presents a pragmatic look beyond AI hype, examining which AI technologies are proving genuinely useful in production versus those over-promised, and offers predictions for the near-term trajectory of the industry through 2026.
Background
Richard Campbell is a well-known tech commentator and co-host of .NET Rocks!, a long-running software development podcast. He frequently gives talks on technology trends, AI history, and the industry's hype cycles.
This talk steps back from the daily flood of AI announcements to assess what has actually changed in practice — separating real, production-ready AI from overblown marketing claims. Campbell draws on historical parallels (e.g., the dot-com bubble, the rise of cloud computing) to argue that much of the current AI frenzy overstates near-term capabilities while underestimating long-term shifts. Key topics include the limits of large language models (LLMs), the high cost of inference, the practical realities of "AI agents," and the ways enterprise software is - and isn't - being reshaped.
For context: 2023–2025 saw massive investment in generative AI after ChatGPT's launch, followed by growing skepticism about return on investment, especially in enterprise settings. Campbell's perspective is valued for being grounded in actual software engineering and business use, not venture-capital hype.