CNBC reported on September 18 that at Dreamforce the AI safety debate met reality, with business leaders saying older AI models are enough for their needs. The remark contrasted with the same conference's headline moment, when CNBC reported that Nvidia CEO Jensen Huang diverged with the CEOs of Anthropic and OpenAI on AI safety.
The strategic implication is that for many enterprises the binding constraint is integration, data and process, not the newest model. That favors vendors that sell workflow and governance over those that only sell raw capability.
Source: CNBC