Yes, but I should have been more clear. My critique of OpenAI is that it is a scam organization, because ChatGPT was always nothing more than a glorified toy, and yet OpenAI promised that it was a path to AGI. If OpenAI hadn't accepted hundreds of millions of dollars in donations, but maybe only a few million, I wouldn't be calling them scammers.

I suspect, but can't prove, that most of the hype around OpenAI was orchestrated by OpenAI's marketing department. If true, then OpenAI operated a lot like shitcoiners, using the revenue from their scam to pay unscrupulous people/websites to promote their scam, thus getting them more donations. An audit of OpenAI's financial transactions would prove or disprove my suspicion.

Off the top of my head, the AI project closest to instinctual knowledge I can think of is Cyc, which is a collection of common knowledge. Allegedly it has proven useful, but so far not for making an AGI.

AI researchers have investigated almost every conceivable avenue for how to create AGIs. They tried to do ML decades ago, but only in the last few decades did we get machines fast enough and drives big enough to process and store the required data.

I'm glad you put it that way. The current ubiquitous focus on Machine Learning is hampering the research and application of AI. Older techniques, such as expert systems, are ignored because they can't be scaled through just optimization and better hardware. And yet expert systems are useful right now (e.g. for helping doctors diagnose diseases), but they require people to put in mental effort to make them useful -- just like any other programming project.

I haven't read his critiques, but I suspect I probably agree with most of what he has to say, since I also have a dim view of empiricism.

On a broader note, a lot of AI researchers are about to lose their jobs, and most AI job postings will soon vanish, not to be seen again for a decade. So begins the next in a long line of AI Winters.