So yeah, like Alison mentioned, um thank you everyone for joining um this workshop. So this workshop is going to be focused on um how to create summaries and insights using A I um in notably,

So we thought that it would be interesting, helpful, hopefully to start with these principles because it's something that we've been investing a lot of, I would say time resources, energy, you know, cycles trying to navigate this new, you know, what feels new in a lot of ways, this new A I landscape and I'm trying to, I think, find our voice, trying to understand more about the risks and these principles, I think really emerged from our internal process to do that where whereas I think maybe for another type of challenge, we might have come up with a framework or frameworks are very popular here.

So to get into how we use A I and notably, and also where we don't. So we're gonna start talking here about this overall research process and we thought it would be helpful to kind of visualize where in this process AI is playing a role where it isn't and then also talk about the future in ways that we may start, you know, pushing in some directions and where we think it's OK to experiment.

So that's with AI summary templates. And I like to think of or how I like to explain AI summary templates are as your ultra capable research assistant. So more than just a tool that can debrief exactly what was said. But also some something that helps with that initial sort of debrief or that initial analysis of a conversation to find like action items, key takeaways, maybe gaps or opportunities.

So what is powering these templates that they're called prompts and prompts are essentially how humans talk to AI tell it what to do, tell it how to look for things.

So we wanted to show an example of this summary template in action and what it looks like in the app. And we're actually gonna do that on this white paper, which has been pretty influential to us, it's a white paper around the risk of large language models and so we have it added in here.

Thank you for joining us today. I see some familiar names from our session yesterday

I wanted to briefly recap we, for those of you who were here yesterday, we did talk about um like in more in depth of what our principles are around A I and how we think about developing new features and experiences with A I.

we can think of AI insight templates on the other hand, as more of your super creative of like synthesis partner. So somebody who is helping you to pull apart that data in really interesting ways to help you find like this intention and really play with that data in order to generate new ideas.

And I think you made a great point about the quality of data that goes in. So there's certainly something to be said about. You will get out what you put in here in working with AI and these models. So the great thing about insights is that they are happy at the end of that, you know, sort of research process where there is this very highly curated intentional you know, set within a theme or data that's been tagged a certain way. So, I think that that has a lot to do with, you know, the quality of output being really high.

We just thought it'd be fun to like pull back the curtain and share. Never before seen behind the scenes work here. So this shows you how we're actually powering our template.

We are going to actually dive in and start playing with some of these templates together. So we're gonna use this for what's template to generate some new ideas on a data set that we actually collected.

So we don't use your data to train any algorithms. We don't use it for saving any context. We don't work or partner with anyone who does. So your data is not being um like stored anywhere else outside of notably um or used for any training purposes. More specifically one of the things that we do right now, which is something that we've been thinking about giving users more control over.