It's interesting (read: wrong) for an AI company to bother writing the user interface for their web application. This was a failure of integration testing and

It's interesting (read: wrong) for an AI company to bother writing the user interface for their web application. This was a failure of integration testing and defensive design, whether the component was open-source or not. There's no reason to believe that an AI company would have the diligence and experience to do the grunt work of hardening a site. But management obviously understood the level and character of interest. Actual users include probably 10,000 curiosity seekers for every actual AI researcher, with 1,000 of those being commercial prospects -- people who might buy their service. This is a clear sign that the managers who've made technical breakthrough's in AI are not capable even of deploying the service at scale -- no less managing the societal consequences of AI. The difficulty with the board getting adults in the room is that leaders today give the appearance of humility and cooperation, with transparent disclosures and incorporation of influencers into advisory committees. The leaders may believe their own abilities because their underlings don't challenge them. So there's no obvious domineering friction, but the risk is still there, because of inability to manage. Delegation is the key to scaling, code and organizations. "Know thyself" is about knowing your limits, and having the humility to get help instead of basking in the puffery of being in control. This isn't a PR problem. It's the Achilles' heel of capitalism, and the capitalists in OpenAI's board should nip this incipient Musk in the bud or risk losing 2-3 orders of magnitude return on their investment.