ChatGPT blows up the internet

January 14, 2023

When I posted Playing with OpenAI on November 30, 2022, it happened to be the same day that ChatGPT was released. At the time, I only heard about it... and well... the rest is history. It has consumed my news feed, social media, personal conversations, etc. over the last several weeks.

In the news

In its first week, ChatGPT gained 1 million users and has been constantly in the news ever since.

Major news outlets have covered it:

Educators are now wondering how they're going to evaluate student essays:

There's also the very legitimately scary and dark side:

My thoughts

Adding my own thoughts to the mix:

  • Generative AI is here to stay. Every technology disruption has come with the good, the bad, and the ugly... and generative AI is no different. Banning it or preventing people from using it isn't the right approach. Instead, how can development progress knowing that there will be bad actors in the system? That's one of the key questions for me.
  • ChatGPT is at its best when used as a tool. I've asked ChatGPT to help me brainstorm ideas, give me summaries of various topics, etc. And that's where it's been personally most valuable. Yes, students will probably cheat on their essays (but they're just cheating themselves from the learning process... they'll learn that the hard way).
  • Handling factual errors. One of the troubling things about ChatGPT is how convincing it sounds, even when it's totally wrong. Aside from actually fixing any inaccuracies, there also needs to be ways for the tools to express their uncertainty, as well as provide support to people so they can discern and verify facts on their own.

Parting words

2022 was a pivotal year for generative AI. Let's see what 2023 has in store. Very exciting times!