Open AI has established itself as a market leader in LLM applications, but that dominance is not guaranteed. Especially with the moat being drained by open source, the erstwhile company is leading the charge to establish regulatory barriers.
What Mr. Altman calls for is no less than the death of open-source implementations of AI. We can, do, and should adopt AI governance patterns. Regulatory safeguards are absolutely fine to define and, where necessary, legislate. Better would be a regulatory agency with an analogous knowledgebase of CISA. But what will chill startups and small business innovation completely in using AI to augment people is a licensing agency. This is fundamentally different from the export restrictions on encryption