Since I'm dealing with how good claude code is, I gave a research task to claude to help my in-laws (who are psychologists) understand the coming storm at a more visceral level.
AI produced this: https://nexivibe.com/future.mental.health.web/index.html
I can't argue with the urgency. I then had claude produce a prompt based on this that I could feed Grok as a project to do therapy, and the result was shockingly good.
Obviously, Dr. Oz is going to be a political lightning rod for some people, but if you factor that out and try using AI for therapy... this is coming. This is happening right now.
It's unfortunate that there are not citations for some of the numbers. I found this piece compelling:
> Marcus. He is twenty-eight and did two tours in Iraq. He has nightmares four to five nights a week. He drinks to fall asleep. His girlfriend left because she couldn't live with someone who woke up swinging. He tried therapy three times. The first therapist was a civilian who asked how combat made him feel. The second used Cognitive Processing Therapy but sessions were every two weeks, and when she said “trauma narrative,” his chest closed. The third was private practice, $160 a session. Marcus started to describe what happened in the house in Mosul — the one with the family in the back room — and the therapist's face changed. A microexpression, less than a second. Marcus caught it. He stopped talking. He never went back. The 73% problem, made human. Marcus has things inside him that are killing him slowly, and he has never found a room safe enough to say them. Not because the therapists were bad. Because the therapists were human, and the things Marcus needs to say are the things that change how a human looks at you.
Just 9 parts plus foreword and appendix. It’s cool that it could generate this but it’s going to miss 99% of the audience. I intend to be constructive, not dismissive of your efforts.
I think there is a coming storm as you say: the disinformation crisis on overload. Even what you present makes my eyes roll as a more competent person in the room. It's not an impressive website, it's a carbon copy of plenty of sites that exist.
It is full of text that is hard to decipher as being meaningful in any way, and that has unfortunately been true of LLMs for a long time, they are great at bullshitting.
Yet, to determine if the content makes any sense takes a substantial effort, an effort I am highly confident you did not do.
Still, I believe you're right, this will be enough. This will be "not everything you read on the Internet is true" x1000.
Sure, the results CAN be good. The real question is if your health is one of those cases you want to roll the dice with. I don't, but that's up to everyone.
Tesla fsd was the same thing. You needed to have people to try it. Some paid for it with their lives. Will it make progress, sure. Do I want to be the one that dies to make that progress? No