The solution might be cancelling my AI subscription
Canceling an AI subscription may be the solution to regaining focus after building 50+ abandoned projects.
The author built dozens of projects with Claude and other AI tools—speech recognition systems, email tools, games, news sites, SaaS products—but maintains almost none and finds little value in them. The core problem: AI removes friction so completely that it becomes a attention hazard, functioning like a thermonuclear ADHD amplifier. Each session devolves from solving a specific problem into sprawling feature work. The author quotes Cal Newport's digital productivity paradox: tools that reduce friction often increase shallow task volume and context switching, weakening deep work. Quality writing requires commitment and focus; the author's voice-to-blog pipeline produced garbage because zero effort meant zero commitment. The author argues that friction enables focus, and focus enables product. Without friction, generative AI becomes a liability despite its genuine technical capability.
What HN community is saying
The top-ranked comment reframes the author's pile-up not as waste but as missed learning opportunity, arguing that building things with AI—when done with rigor—teaches you how to use the tools well, contrary to the author's framing. A lower-ranked counter-reply disputes this, claiming the author did not build those projects; AI did, and there is little skill or learning involved. The thread's center of gravity supports the author's concern: multiple commenters note that AI's ease of use encourages project-switching and shallow work, with one comparing it to GPS navigation eroding navigational knowledge. A high-ranked comment adds that AI makes easy work easier while shortening attention span, so there is productivity gain but little real progress. One commenter notes using AI as a rapid prototyping tool before scrapping and rewriting by hand once complexity grows—a bounded use case—differs from unfocused generation.