Dr. BotLove or: How I Learned to Stop Worrying and Love the AI

I’ll be blunt: AI is really cool, but also really, really dumb. It can build an app in an afternoon, design a VR videogame overnight, update your systems, and build an entire ETL system. It also can’t remember who you work for, has a habit of broadcasting your secure keys, and clutters your codebase with dozens of debug_duplicate.py files.

AI is the best of the worst tools in our belt. It’s brilliant, it’s broken, and I use it every single day

In this session, we’re taking a frank look at where AI actually is right now, where it shines, and where it consistently falls apart. The critical topics of how to make it useful, where it shines, where it just won't work, and ways to make it less obnoxious. How to accommodate goldfish memory, terrible assumptions and tendencies to go on “adventures”. Strategies the average user to leverage these tools safely and sanely without a dedicated dev team or a complex setup.

There are various ways to integrate it into your work life without losing your mind (or your ability to think). Come learn how to interact with the bots recognizing they are like an overconfident, very smart, but completely chaotic junior dev.

Scope: This discussion will be focused on commercially available AI tools like Claude, ChatGPT, Gemini, and Copilot. I’m sure some of y’all have custom fixes for all these issues, but for this we’re focusing on the out-of-the-box bots were stuck with.

• Beginner
• AI/ML
• Product
• Soft Skills
• Wildcard