Build An Audio AI App Transcripts
Chapter: Feature 4: Chat Q and A
Lecture: Exploring the Podcasts with AI Q and A

Login or purchase this course to watch this video and the rest of the course contents.
0:00 I'm so impressed with this feature. Let's just play with it for a minute. If this this particular
0:06 video doesn't inspire you, feel free to just skip ahead. But we'll wrap things up with this chapter.
0:11 But let's just play with it a little bit more and see where it goes. Okay. Over here. Let's talk.
0:16 Let's do this one first. Okay. Let's talk to the Python bytes. So in this episode,
0:20 you can see in the summary that Brian is going to create a Pytest suite timeout
0:28 plugin for Pytest. The title is that episode where we just ship open source. So he releases
0:34 that to PyPI. And then it turns out I created something for you mommy and release that to
0:39 PyPI as well as something for list monk hosted email service also released that to PyPI. So the
0:47 three of the four things we discussed that week are actually open source projects that we created
0:52 that week. It's kind of insane. So let's ask some questions about that. The transcript exists,
0:57 so we can chat with it already. What open source projects did the hosts create and release?
1:05 Awesome. The host, Michael and Brian. So already kind of cool that I figured out
1:14 discuss the following open source projects they created and released Brian. Oh my gosh, look at
1:19 this. Brian mentioned that he created Pytest plugin called Pytest suite timeout adds a timeout
1:25 for the entire test suite. The transcript states, remember we asked it for show me your work. Brian
1:31 mentions the new Pytest plugin he created called this and adds a time. Okay, that's kind of
1:36 repetitive. It's okay. Michael shares that he created a Python client for a self-hosted analytics
1:43 tool, umami, that allows sending custom events from Python code. The transcript says, Michael
1:49 shares this thing. Yeah. Okay. Perfect. They also discuss some open source alternatives. So in
1:54 summary, the two, there's actually three, but I think it missed it. open source projects,
1:59 the host created are these two. Please let me know if you need clarification. Awesome. Let's ask one
2:04 more question. How can I collect web analytics on my users without violating their privacy?
2:16 That's kind of an open-ended question that wasn't exactly mentioned in the transcript,
2:20 but the umami platform being self-hosted means the data doesn't leave or get shared. Or there's
2:25 none of that retargeted garbage that seems to so many people on the web think is one
2:32 reasonable and two is the only way. Let's, let's see what that says.
2:39 Oh, let's be a little more specific. How can it without sending their data to a third party?
2:47 That made all the difference. Just be a little more specific, I guess. Here are the two key
2:53 points on collecting web analytics without sending data to a third party. Michael talks about setting
2:58 up this analytics tool, umami on his own infrastructure, using Docker to gain insights
3:02 without sharing. Exactly. And here's the transcript source. The host agrees that running the services
3:09 like analytics and mailing lists on one's own infrastructure gives you more control. In summary,
3:15 to collect web analytics without sending data to a third party, one option discussed in the podcast
3:20 is to set up open source analytics tools like umami on your own infrastructure using Docker.
3:25 That is pretty darn good. That is pretty good because I didn't ask a question of what did
3:34 they say or summarize this. I kind of asked for advice in the, in the context of this episode.
3:42 Okay. I'm happy with that one. That's really, really cool. Let's go over here and let's chat
3:48 with this one. Remember this is about Emma Stone. How did Emma convince her parents to move to LA?
4:00 See if it'll know that. Yeah. So Emma Stone convinced her parents to move to LA so she could
4:06 focus on acting. Emma discusses experiencing anxiety from a young age, convinced her to
4:12 homeschool and later move to LA. Okay. So according to transcript. Yeah. Yeah. Okay.
4:17 One more. Let's go over here to this one. Remember this one is about Apple watch.
4:23 What is the legal issues? What are the legal issues with the Apple watch? Based on the transcript
4:29 summary provided, the main legal issues discussed were related to ongoing patent dispute between
4:35 Apple and an Italian company called Massimo sounds Italian. I'm thinking of it's American.
4:39 Here's some of the details. Oh my gosh. How awesome is that? So in summary and such and such and such,
4:45 I think for some reason it can't process the entire two and a half hours of text. So it's
4:52 missing some, some of the detail here, but that's okay. It's still quite excellent. What is getting
4:59 us here? I'm impressed. I think I might be most impressed with this one because it talks about
5:04 so many different topics. That's kind of the point of that show is to bounce around a bunch of things,
5:09 which I think might throw it off a little, but it seems to do really well. All right, there we have
5:14 it. We've, we've asked a bunch of questions about our podcasts. I think we're getting there. I think
5:19 it's about, about living up to its name, right? We have the x-ray podcast app that doesn't just
5:26 load up the podcast and let you play them transcripts, summaries, key moments, search,
5:32 and now even the ask AI incredible stuff. So hopefully you really enjoy this app that we built.
5:40 I'm really happy with it. I think this is quite a neat demonstration of what can be done with AI
5:48 and assembly AI and all these things. And there's, there's plenty of places you all could take this
5:52 or apply it to your particular area. And honestly, there's a bunch of other things that we haven't
5:58 even leveraged out of assembly AI. Still very fun. Be sure to check out the source so that you have the final version.

Talk Python's Mastodon Michael Kennedy's Mastodon