Mike Judge raised an interesting point last year about the expected surge of software development driven by AI. He questioned why, despite claims of heightened productivity among developers using AI tools, there wasn’t a notable increase in the quantity of new apps and software. He suggested that if these tools were truly revolutionary, we should see a flood of new creations, akin to a wave of Tetris clones on platforms like Steam. At that time, the anticipated explosion of indie software hadn’t materialized. Recently, Pete Goldsmith observed a significant uptick in “Show HN” posts on Hacker News, noting that the number more than doubled over the past year. He had been building projects with large language models (LLMs) and workflows involving agents, prompting him to investigate whether more people were indeed creating software. His findings confirmed a sharp increase in submissions, resembling a hockey stick graph. This surge might indicate a shift in the development landscape, suggesting that more individuals are beginning to harness AI tools for their projects. The implication here is that we could soon see an influx of weekend projects and amateur software, often labeled as shovelware—products of varying quality churned out rapidly. This trend might signal a significant change in how software is developed and consumed, potentially altering the market dynamics and user expectations.