In my experience, artificial intelligence is invariably an abysmal tool for generating responses that assess highly charged emotional topics.

In other words, therapists, like many professionals, are not losing their jobs anytime soon AT ALL to AI!

I’ve always liked the popular writer Mark Manson and his content. Many years ago, I read his bestselling book, “The Subtle Art of Not Giving a Fuck.” Great book recommendation, and I’m now intrigued to read his others. I was especially intrigued with his new AI tool, Purpose, to see if its Large Language Model (LLM) did a good job at analyzing a conflict I had via copy and paste that I had over text recently. (While I get the conventional wisdom in ideally solving conflict verbally, face-to-face, or over the phone, I think there are important exceptions to when solving conflict in writing like text or email is preferable, particularly if you’re on the receiving end of someone demonstrating problematic behaviors gaslighting, manipulation, and interrupting.)

Anyways, the Purpose AI app is wonderful in theory, and kudos to the idea, but it’s nowhere near where it needs to be at this time. I tried the 7 day free trial, but recently paid for a month at $19.99 to retry it, in order to juxtapose it with ChatGPT and Gemini. It does have other intriguing content besides chatbot stuff, so I may check out the non-chatbot stuff while I paid for this month.

While ChatGPT is notorious for just siding with whatever you say in your prompts, the Purpose App, at least from my experience, swung too far the other way. I wrote a statement which had healthy boundaries and solutions, and it misunderstood it for being controlling, in a way where an regular human, like Mark Manson, would have understood the context more accurately. Perhaps the experience would have gotten better with more personalization and back-and-forth, or for personal growth via a different context, but I cut my losses. I figured it’d be foolish to invest any more mental capital experimenting with dumb AI chatbots. Gemini was a bit more in the middle, between ChatGPT and Purpose, but way too “formal academia-sounding” in how it responded.

I find the ONLY helpful thing about AI, is to “brain dump” everything comes to your original thought and mind authentically, and have it polish your writing minimally, for certain criteria. Indeed, this is true even if your “stream of consciousness” is obnoxious, unproductive, hurtful, replete with typos, awkward sentence fluency and disorganization… it doesn’t matter. What matters is writing your original human-generated content by yourself, and only involving AI in maintaining your substance, while making minimal edits for what you seek (for me, it’s often so-called securely attached, nonviolent communication styles, firm and healthy boundaries, respect and empathy towards others and the self, cohesion, organization, omitting redundancies). AI algorithms should have common sense shit like this natively built in, but I always specify the above, or something to that effect, to get better generated responses.

What I strongly DON’T recommend is asking AI to generate original content for you, just like if you are a student writing a paper. Unless you want to read a funny poem about Oregon’s property tax system. Then, ask ChatGPT to write one, and you’ll get pretty hilarious iterations (that was also my first ever AI prompt, by the way, back in early AI days of 2023, I believe).

I’ve been thinking a lot lately about what healthy boundaries look like. I’ve had this naive, people-pleaser misconception that “boundaries,” more often than not, are cold, selfish, and totally inconsiderate of the needs and feelings of others. But people who actually act THAT narcissistic, are just assholes. They’re also hypocrites, because in being so inconsierate of others, they are violating others' inherent boundaries of basic respect. When healthy boundaries, as opposed to toxic “boundaries,” are an underdeveloped skill, it’s easy to forget how important it is for reciprocity, in boundaries going BOTH ways.

Anyways, onto some good Mark Manson content on this topic: markmanson.net/boundarie…

10/10 would recommend. Check out the corollary articles too, such as on love and vulnerability: markmanson.net/love, markmanson.net/unconditi… and markmanson.net/vulnerabi… Good stuff!