In AI 1 I wrote about the public attention regarding AI. In this post I share my thoughts on the relationship of humanity to AI in the near future.
I suppose because I work in software, AI was a conversation topic brought up by various people I talk to for the past few years. “Do you think coding is no longer a good career? Is AI going to take X Y Z jobs?”
I often answered this way: AI is still at the “tool” level, but one day a breakthrough could change change this tool to something more akin to critical problem solving. A few years go by, and nowadays all of LinkedIn seems to talk “autonomous” and “agentic”. And honestly, I think we might be there now, or very soon. I generally approach tech products skeptically, but some of the brand name AI agents and workers that launched mid 2026 convinced me to acknowledge the level these bots have reached.
In some ways, I’m probably considered behind in my view, especially for someone in tech. In my defense, one of the fundamental perspectives software has taught me is that solutions that get 90% of the problem are trivial, but pushing products beyond that last 10% are often surprisingly far away. But yea. We might have pushed past that 10% recently, the autonomous bots are really that good.
I think that with these AI-first systems woven into decision making layers and gaining more trust, we hope that those in positions of influence are doing a good job of guiding things into the flourishing of all. As 2027 approaches I plan to spend more time tinkering with the latest autonomous software and see where things land in my own life. As of today, my primarily involved usage of AI is in Claude Code. I anticipate this to change significantly next year.
Aside: I think it’s crazy that just several years ago, when LLMs first made their big media break, all the talk was about “prompt engineering” becoming a career path. And now we know that we’ve moved up a few levels of abstraction now that bots can move on their own. Always fun to observe how certain predictions end up.