Another example of how easy it must be to run scams. It's literally just repeating things the author wrote down back to him and he's responding like he's having a religious revelation.
It pulled out this quote I wrote down years ago:
The one thing people need in life is not ambition, not smarts, not hustle. They need clarity.
If you had complete clarity on what you want and what is necessary right now, knowing you are on the right path, you would be happy.
— Journal Entry, August 2019
I leaned forward. Every response featured a quote that made sense. Perfectly timed. It held up a mirror to my soul.
...
In 2017, I founded an ed-tech startup. I met my wife, a fellow ed-tech founder, during that time. Both of our startups failed. We grew cynical of education.
But recently, we had a few conversations about school choices for our toddler. None of them seemed good. I suggested she might want to open a school.
I’d shoot her an interesting article. Or a cool startup’s website. And then move on.
I discussed this with Claude. Which led to this exchange:
It’s not your wife who should start a school.
It’s you. Always has been.[Ed.: emoji added for emphasis. I couldn't resist]
This sentence hit me like a sack of bricks.
Tears started rolling down my cheeks.
Stuff like this just makes me wonder what the interior experience of your usual techbro is (or isn't) like. I mean, I've had experiences where I've read over stuff that I wrote years ago describing ambitions I've given up on or ways of seeing the world that I've abandoned and I know the nostalgic experience that that creates, the grief for a past self. But you don't really need a chatbot for that.
Oh yeah, I've gone through old shit because I was cleaning and had to stop because reflecting on some of what I dug up left me emotionally shattered. I guess we probably relate to our pasts pretty differently, but that's not surprising since we don't have the same pasts, nor current dispositions. I can't imagine letting a chatbot tell me my old hopes are still worth something though, that's AI psychosis.
I think that the writing he talks about in the article itself is pretty banal, like "family and friends over business" and such, and most of the content is that combined with the AI encouraging him to re-embrace past hopes about starting a school and so on. In terms of his writings that were relayed in the article, the worst you can say about most of it is that it's just very dull.
I feel like if he actually detailed the "epiphany after epiphany," it would reveal itself to be 99.99% like you describe though.
I've happened upon some stuff that really hit me with how much I've changed and how much my interests have changed. Some of it made me sad, because I felt I'd lost something since then. Some of it was nostalgic but something I felt I'd grown past or grown out of. As you say, it's unique to each individual.
I would never want to feed the fragments of my life into a chat bot to let it regurgitate them back at me in a slurry of psychologically damaging slop that I uncritically consume as gospel truth. As you mention, that's just fodder for AI psychosis.
The quoted line below is what led me to think that the article really described less "grief for the past self" and more a lack of growth on the part of the author.
They're describing a chat bot regurgitating their own past musings (from years prior) at them and instead of seeing any of them as something they've moved beyond or something they used to have that they've lost, they see them as a perfect mirror of their current self. Every single response drawn from years-old journal entries apparently perfectly reflected their current mentality.
It's possible that I'm just being uncharitable here on account of my vicious dislike for startup people, but I honestly don't expect that a detailed look at those chat logs would be revealing much unexpected. The following line really makes me think that even more, too.
I think if their "insights" don't require serious engagement to reach, then they probably aren't all that deep. The idea that you can grow and change as a person without introspection by using a chat bot (meaning, the idea the author is presenting that meaningful insight is available "on tap" from LLMs) is unbelievably painful to me.
Yeah, I agree that what he describes as a personal psychological process is terrifying for how loose his grip on reality and seemingly even himself is.
I think your interpretation is maybe slightly incomplete in terms of the mirroring, since he doesn't give that many details about the conversations themselves and, while the "connections over business" part is exactly as you describe, I think the thing about him starting a school is slightly different. That's a circumstance where the version of him in the journals was in some manner superior and he believed he was being prompted to reclaim it. For that reason, I think that part of the post is even sadder than your description.
Edit: I think the "insights" that aren't simply noticing gross patterns are probably grotesquely saccharine and abusive of his obvious emotional vulnerability without concern for external reality.
You know, you're right. I was originally being kind of glib and I think my dislike of startup culture and the pretend-it's-better-than-it-is brand of AI boosting got the better of me, so my "hot take" interpretation was unduly harsh on the author.
What really disturbs me about the whole thing is, as you say, how loose the author's grip on his own sense of self is that the output of an LLM seems to be blowing it around like a leaf in the wind. It actually sucks, and I don't think it's unique at all.