Hugging Face CEO says we’re in an ‘LLM bubble,’ not an ‘AI bubble’ (techcrunch.com)

Hugging Face CEO Clem Delangue told an Axios event that we’re in an “LLM bubble,” not an “AI bubble,” and that the LLM-focused hype and capital concentration may be poised to burst as soon as next year. He stressed that this wouldn’t end AI’s momentum — LLMs are only one subset of a much broader field spanning biology, chemistry, image, audio and video — but predicted attention and funding will shift away from the one‑model‑to‑rule‑them‑all narrative. Technically and commercially, Delangue argues the market will move toward a multiplicity of smaller, specialized models that are cheaper, faster, and easier to run on enterprise infrastructure — for example, a banking chatbot doesn’t need a generalist LLM to answer policy or account queries. That implies a recalibration of compute spend, latency and privacy tradeoffs, and a surge in domain-tuned models and on-prem deployments. Hugging Face says its own strategy is more capital-efficient (it still holds roughly half of $400M raised), positioning it to weather a retraction in LLM valuations. For practitioners and investors, the takeaway is to prepare for diversification: more targeted architectures and fine-tuning pipelines, not just scaling up foundation models, will drive near‑term applied AI progress.