All posts
Published at Sun Feb 18 2024 in
Rows HQ

2024 W8 - Moderating a Panel on AI

Blog - AI team

Every week I post about one thing that happened at Rows. We're building in public!


This week I was a moderator at an event on AI.

I learned how hard it is to properly moderate a panel.

  • Challenges: You have a clock to run, go through multiple topics, and cut people short.

  • Goal: The only objective is to get the most out of each participant and output a great conversation. If the audience doesn’t take home good ideas and a few laughs as an extended goal, you failed.

  • Upside: Because you get pressure to deliver value fast, your mind will be overstimulated to find bridges and connections and little summaries. At least, this is what happened to me! ☝️ This is also why I get a lot of value from writing these Building in Public posts, forcing me to see the connection of what happened.

This was a panel organized by Microsoft on the topic of AI, as part of the Building The Future conference in Lisbon.

The 4 participants were:

  1. Me for Rows;

  2. Luis U. from Sword Health;

  3. João F. from Pager Duty; and

  4. Rodrigo S.C. from OutSystems.

The plan we came up was:

  • First, explain what products or services or features we have launched with AI, and if we’re charging for it! This makes it clear to the audience of what can be shipped into production, and of course the revenue part is always something that makes people pay attention. Here, I noted that only Rows — being smaller — has specifically charged for some of our AI features, with the others for now integrating the AI within their regular contracts and services.

  • Second, tell the audience where the AI developments came from: was it the engineers hacking away at it on their initiative? Was it a hackathon, a top-down management decision? I found it was a bit of everything for all of us, but clearly the individual teams actions seem to have paid a bigger role than any one particular grand strategy.

  • Third, describe the opportunity we see in AI, and how we mitigate the Risks that come from it. It became quite clear that the approach is to keep humans in the loop for now. Copilots, not Autopilots. Pager Duty is perhaps more bullish on automated actions being routed by decisions augmented by AI data, and given that their core product is already an automation, it makes sense. The biggest risk is to not do AI.

  • Lastly, I prompted each of us to say what fun things might come out of AI. Here I saw a big range of ideas: Better emotions for dealing with patients AI, finding if data in your spreadsheets seems fake (or too good to be true), working with images and visual elements, and better team interactions.

It was fun.

I don’t want to get into this habit, but thanks Microsoft! (For the invite).


See you next week!