An excerpt from The Atlantic that you may find interesting.
Too Much Is Happening Too Fast
The AI boom is meant to overwhelm you.
By Charlie Warzel
"Elsewhere online, programmers are beginning to describe an affinity for coding agents that is veering into unhealthy territory. “I’m up at 2AM on a Tuesday,” Anita Kirkovska, the head of growth at an AI company, wrote recently, “not because I have a deadline, but because Claude Code made it so easy to keep going that I forgot to stop.” She describes a “competence addiction” caused by the tools making her so productive: “You hit a prompt, the agent succeeds, you get a dopamine hit. The agent fails spectacularly, you get adrenaline. Both are reinforcing. Both keep you at the terminal.” Kirkovska argues that she sees this among all kinds of AI power users—an unsustainable flow state in which decision making begins to falter and people become sloppy as they grind away."
Thank you Steven for writing about these questions. My feeling is we have been moving down the road towards cognitive reorganization for a while with social media, and AI use. I agree we have to remain curious and question how we use these tools and if we are guiding problem solving or being led by AI. When it is not possible to understand how and why AI comes to conclusions or solutions, the level of risk can increase as dependance on AI for critical systems control is used.
My AI projects are significantly less complex than what you are executing, which makes them much easier to chunk into relatively small phases. That facilitates the pattern you mentioned: code, document, generate a new prompt, and continue in a new thread. It’s a pattern I have been following for quite some time.
I’m definitely addicted to AI dopamine. Every morning, before I even get out of bed, I scan summaries from various tech news aggregation sites to get the latest scoop on AI. Yesterday, I used Claude Code to build my own AI news aggregator.
An excerpt from The Atlantic that you may find interesting.
Too Much Is Happening Too Fast
The AI boom is meant to overwhelm you.
By Charlie Warzel
"Elsewhere online, programmers are beginning to describe an affinity for coding agents that is veering into unhealthy territory. “I’m up at 2AM on a Tuesday,” Anita Kirkovska, the head of growth at an AI company, wrote recently, “not because I have a deadline, but because Claude Code made it so easy to keep going that I forgot to stop.” She describes a “competence addiction” caused by the tools making her so productive: “You hit a prompt, the agent succeeds, you get a dopamine hit. The agent fails spectacularly, you get adrenaline. Both are reinforcing. Both keep you at the terminal.” Kirkovska argues that she sees this among all kinds of AI power users—an unsustainable flow state in which decision making begins to falter and people become sloppy as they grind away."
https://www.theatlantic.com/technology/2026/05/too-much-happening-too-fast/687177/?utm_source=feed
Great forward!
Thank you Steven for writing about these questions. My feeling is we have been moving down the road towards cognitive reorganization for a while with social media, and AI use. I agree we have to remain curious and question how we use these tools and if we are guiding problem solving or being led by AI. When it is not possible to understand how and why AI comes to conclusions or solutions, the level of risk can increase as dependance on AI for critical systems control is used.
My AI projects are significantly less complex than what you are executing, which makes them much easier to chunk into relatively small phases. That facilitates the pattern you mentioned: code, document, generate a new prompt, and continue in a new thread. It’s a pattern I have been following for quite some time.
I’m definitely addicted to AI dopamine. Every morning, before I even get out of bed, I scan summaries from various tech news aggregation sites to get the latest scoop on AI. Yesterday, I used Claude Code to build my own AI news aggregator.
Now it’s time to get my fix.