🌿 My AI approach
It's hard to know where you stand on a topic when there are strong opinions coming at you from different perspectives. Or when you're being told to follow a direction no matter what, repeatedly, so you don't even get the chance to form your own stance.
This is particularly true when it comes to AI. After being told countless times that it is inevitable (it's giving Thanos) and that it is going to replace jobs, it is difficult to ignore the worry and fear that these messages are pushing.
It took a month of no AI chatter for me to be able to actually stop, consider my questions and concerns to decide how I actually want to use AI beyond making oddly specific search requests.
How I use AI
I use AI as a thinking partner to expand my judgment and agency, not replace it. I've built AI agents before, but primarily use AI as a tool for productivity and refinement.
I want to understand how I'm using AI and why, rather than letting it quietly shape my decisions. Whether it's drafting, learning or evaluating work, AI works best for me when it's a partner in the process, not the decision maker. I'm not interested in convenience for its own sake; I'm interested in using it to think better and stay aware of what I'm doing.
Currently, I'm interested in designing a personal intelligence layer that discovers unexpected connections in my thinking and interests.
Concerns with AI
AI hallucinations has always been a huge concern for me, how the machine can respond to our questions with the tone of confidence without the logical reasoning to back up its claims. This article from The Atlantic pushed this concern further: if primary and secondary sources are no longer reliable sources of proof, how do we know what to trust? Adding a layer of questioning to everything we see or hear expands the distrust not just with the content we're consuming but also between each other. As George Eliot stated, "What loneliness is more lonely than distrust?"