AI coding tools allow experienced programmers to automate repetitive tasks, freeing them to focus on more challenging aspects of software development. This automation significantly increases output and makes many previously infeasible small projects now achievable. Millions of small, useful programs that were too costly to develop before are now within reach due to the reduced effort required. AI tools democratize programming, enabling individuals with no prior experience to create simple tools. Junior programmers can take on more complex tasks and learn faster, while experienced developers can complete long-delayed projects more efficiently. Even those who don't directly write code will leverage AI to perform tasks, with the system generating programs in the background. This shift creates a new market for software addressing narrow, specific use cases that were once uneconomical to build. Previously, the high cost of software development meant companies optimized for large audiences and products. Now, the declining cost of building software means even niche, single-purpose programs are becoming financially viable. A key challenge emerges in distinguishing between temporary tools and essential programs that become critical to operations. The author advises asking if a program is depended upon, if it handles important data, or if work would halt if it disappeared. Programs answering no to these questions should be treated as disposable experiments. If the answer is yes, the program requires proper management, ownership, and a plan for maintenance or replacement. The level of engineering rigor should correspond to the consequences of the program's failure, with disposable tools remaining that way. Ultimately, AI makes it economical to solve many small, repetitive problems that were previously too time-consuming to automate.