Red Hot ChatGPT, AGI, and Translators- Lessons from my AI talk at USAFA

ChatGPT is hotter than I thought.

And it was already a hot topic.

I was overwhelmed with the excitement and interest in ChatGPT shown by everyone I met at USAFA’s Leadership symposium. Everyone wanted to talk about it, understand it, and determine if it was to be feared or used. And this wasn’t a technical crowd.

This is exciting and creates an opportunity for the AI community.

The excitement is that this technology can have a big impact.

The opportunity is that ChatGPT has expanded interest in AI. We'll need to use this opportunity to clearly explain where ChatGPT fits and let leaders know about other types of AI that may be beneficial.

I thought I would speak to about 30 people (the first day’s audience was even larger). So I tried to up my game!

With that background, here are the lessons from my AI talks at the symposium.

One: if you are talking about AI or analytics, be ready to discuss ChatGPT.

I covered some non-technical material from Stephen Wolfram’s overview. For example, it predicts the next word and does not think like a person. Or that there are an incompressible 175 Billion “knobs” to turn. I wanted to demystify it a bit.

Two: be ready to talk about AGI (Artificial General Intelligence).

ChatGPT seems human when you chat with it. People want to know if this is close to AGI.

I take the side of the skeptics. Marc Andreessen said:

what is AI? It’s math. It’s basically elaborations on linear algebra. I have a hard time getting worked up about linear algebra. It’s math. We’ll be able to keep the math under control.¹

My goal is to educate, so I also point out that smart people take a different position. I made a note to myself to research this more.

Three: point out that there are many different types of AI.

I recently wrote a blog on this, and this message resonated. When people see the different uses of Practical AI, it opens them up to possibilities in their organizations.

Four: discuss the different roles on a data science team.

At the end of the presentation, I covered the fact that there are many roles on a data science team. This message is good for leaders (so they don’t expect magic from their data scientists) and for people new to the field (so they don’t avoid it because they have to know everything).

Finally, I mentioned the importance of the role of a translator. This sparked quite a few conversations after the talk.

1. I pulled this quote from this blog which was quoting this podcast.