The episode, in writing
by Richard Lovatt · from the episode above, checked against what was said
Let me show you how simple it is to set up an episode, how it works, and I'll probably find a few bugs along the way. That's what building is like.
This is the Today tab. I've already recorded one episode today, which is why you can see Today Watch Episode 2. I've literally just pushed and built the ability for you to record multiple of these a day. My goal is for founders to be able to just be consistent and keep adding to your directory, so you build up that big library of you building in public and the whole process. I think one a day is pretty good, but at the same time I know as a builder that when you're building something you get these moments of inspiration. It's already becoming addictive for me, because I just thought of something, and instead of recording a snippet for tonight, I thought, why not just record another episode.
So we click on Today as a tab, and I can watch that one back or fill another. I created this Freestyle button, I might actually remove this at some point, and I might call it something else, but the reason I created this voice note button, Freestyle, is because as you're going through your whole day, making pushes, in meetings, emails and Slack, everything going on, you get those moments of inspiration where you think, tonight on the episode I want to talk about this, or I want to remember this. So you click Freestyle and it's just like a voice note, you record as long as you want, and it saves that as context for tonight's episode. When you come back and click create episode, or film one, it'll have that in its memory database and give you suggestions on what the questions could be or what the topic could be. Here's tonight's context.
Context is basically its memory, how it's building tonight's database and questions. There are so many ways it can get context. It connects to your Git, so it can see what you've pushed today, your emails, meetings, recordings, Stripe, so it can see your first customer joining. There are so many ways for it to build this context of what you're doing through the day, so that the interviewer can cleverly ask you about these things as it goes, like a human conversation, so that you don't have to think about that whole process of creating content. You just sit there and answer.
It doesn't only connect with different bits of software through API, it also does screenshots. So if you've just built something, created a screen or a graphic, you can take a screenshot and it will take that as context. You can do voice notes, video capture, there are so many things you can connect to, and the context comes up here. Below, you can see it's reading that context and giving me options for whether something is worth an episode. Here's a topic: does the new render flow hold. You spend hours in Cloud Code turning Underway away from editing, towards silences, cutting out the ums, press play. It might not make much sense to you, but it was an important part, and I could do a whole episode on it. Finishing the walkthrough that cut off. Marketing the app you built with your fiancée. Where the founders come back to record again. Really good topics. I don't have to click any of these, by the way, it's just your option.
If I click and film another, here's what will happen. I'm already doing this, I'm already recording it, which is why you see this, it says set up your shot. Unfortunately I can't even show you, because this is the funny thing about building in public and these kinds of videos, it's not even about perfection, it's not about retakes. I wanted to move away from editing and making it fiddly, it should just be real, genuine content. And you know what, if things don't work and things slip up, it's still there. I think that's the whole point of it.
I set this episode up to record setting up an episode, the whole process, but because I'm actually in the process right now of setting up an episode, I can't show you the screen it would show you before. I can describe it though. You would start to see, you would decide whether it's video or audio, because sometimes you might be on the move, you might not be able to do a video, you might just do a podcast answer via audio, so you can choose that. You can choose the video, portrait or landscape, you can choose whether it's a short or a longer interview. Then you can go into the topic, you can give it a topic because you know what you want to be asked about. You can ask it to think of topics from what you've done today, you can ask it to look at a transcript. There are so many ways for it to come up with what we're going to speak about.
One thing it does ask you when you're setting up is that one of the options is a walkthrough, where you share your screen and walk through what you've been doing. I clicked that and it gives you the topic. I put in setting up an episode for Underway, and once I'd done that, this is the next page. It gave me a bunch of questions it could ask me in this walkthrough. I actually deleted all of them and just left the first one, which is, show me setting up an episode, what is it and where does it stand right now.
So what happens is I share my screen and it asks me, like an actual robot AI, in a really realistic human voice, and I can naturally answer to it. You turn the camera on, it says looks good, start. It will tell you if the lighting's off, whatever. I can actually add in more questions here, I can remove that, and then looks good, start, and it will start recording.
When I finish this, I'll see if I can add screenshots to episodes, or snippets, to an episode that's already been done. That way I can show you what that screen I couldn't show you would look like, because one function I did create is the ability to take an episode you've just recorded and add to it. If you've already done episode 2, you can create an add on to it. I might do that on this one. Let's see.