Like almost everyone in my field of software engineering, I’ve seen my company, team and projects pivot toward developing AI-powered products in the past year-and-a-half or so. Everyone’s doing it. I want to set aside my personal feelings on whether this is a good thing generally, since that’s such a big and messy question of which I only have fair-to-middlin’ expertise, and instead focus on something I have much more experience in: Ensuring AI assistants’ user experiences aren’t terrible. I wrote about making my own proto-LLM way back in 2016, then about the first major LLM coding agent in 2021, and then about interviewing software engineers who use AI coding agents last year, so I know a thing or two about the evolution of AI in the software space. I also designed and led development of an LLM-powered chatbot for Elastic’s support portal, which recently won an industry award for best use of AI for customer support. And in that time, I’ve seen and purposefully declined tons of…
No comments yet. Log in to reply on the Fediverse. Comments will appear here.