"Been thinking a lot about whether it's possible to stop humanity from developing AI. I think the answer is almost definitely not. If it's going to happen anyway, it seems like it would be good for someone other than Google to do it first."
- Altman to Musk, immediately before proposing what became OpenAI ( https://www.techemails.com/p/elon-musk-and-openai )
"OpenAI is on a path of certain failure relative to Google. There obviously needs to be immediate and dramatic action or everyone except for Google will be consigned to irrelevance."
- Musk to Altman, later ( https://www.techemails.com/p/elon-musk-openai-path-of-certai... )
I'm not convinced. In the past half decade or so this industry has veered hard toward outright fraud and grift. I see this trend all over--adtech, cryptocoins, "AI", security... These days I assume technologists are frauds until they prove otherwise. It's a blunt instrument, but it often works well.
When I advise the NIH to do was to do a large bulk group buy on behalf of many thousands of scientists that use the scale of the NIH to negotiate in extremely large concession. My experience with Amazon is that you can basically get them down about 50% just by asking.
I work for a well-funded company now and we use the cloud and we have on-prem infrastructure. The on-prem infrastructure is extremely hard to change sometimes just asking for a GPU will take 6 months or more. Storage is always highly limited and slow. Let the cloud hyperscalers do what they're good at and focus on doing the science
When I worked for Google I visited NIH, sat on study groups, and helped advise program managers how to move more compute to the cloud. Like many other techies in SV I have a PhD in a quantitative science and understand how NIH works. My efforts were entirely designed to help update the establishment, not tear it down, and that's true for the wide swath of my coworkers I encountered.
The folks who are doing this are a subset of the tech community, who do not represent the larger community.