AI Credits, 4 Improvements, and Podcaster Kevin Rose  🔗 **Source:** [AI Credits, 4 Improvements, and Podcaster Kevin Rose](https://share.snipd.com/episode/9c05a392-d97a-449a-8dd4-0da57de20025) ## Highlights - **Buy Extra AI Minutes In App** - Do buy additional AI minutes in-app when you run out instead of contacting support. - Snipd now lets power users purchase extra minutes directly on any non-processed podcast if their premium allocation is exhausted. Transcript: Kevin Smith The big improvement announcement is we've finally, finally implemented the ability to buy more AI minutes. So if you run out of the AI minutes that are included in your premium subscription, you don't have to reach out to us anymore and beg us for more. You can just acquire more directly in the app. We slowly rolled this out last month. So you might have already seen it or might have already bought some. If you haven't seen it yet, it's available to everyone now and you can find it on any non-processed podcast if you do not have enough AI minutes left to process it. Yeah, this was one of the most often requested features by our power users. It's for all of the people who are uploading a lot of audiobooks to Snipt, listening to long-tail podcasts or have private feeds where you have to process a lot of the content yourself. ([Time 0:00:17](https://share.snipd.com/snip/62f5ecc4-281c-4ded-b731-e0fb06a42b63)) - **San Francisco Meetup Sparked More Events** - Kevin Smith recapped a lively Snipd meetup in San Francisco hosted by SWIX from Latent Space. - He found the energy and attendees so positive that Snipd plans more meetups and asks locals to help organize. Transcript: Kevin Smith Next up, I actually want to briefly look back at the Snipped user meetup that we did in San Francisco not too long ago. First of all, big shout out to SWIX, the podcast host of the Latent Space podcast, who generously hosted us. The event was incredible, thanks to everyone who came. The energy was great, the people who joined were interesting, and it was just a lot of fun. So the biggest takeaway for us is we should really do these meetups more often and in more cities. So the next time we're in a big city, we'll reach out to you. And if you happen to live in a big city, like, you know, New York City, London, Berlin, and you can help us organize something for around 20 to 30 people, do reach out to us. ([Time 0:01:07](https://share.snipd.com/snip/a4906fc9-5f3a-4f23-8185-d482826d510a)) - **Snipd UI And Snipping Model Improvements** - Snipd improved several UX elements: AI snips on episode screen, autoplay toggle, cross-device progress sync, and a better snipping AI model. - The snipping model change includes a new model and better prompts, especially honoring longer custom prompts in snipping settings. Transcript: Kevin Smith So first of all, we have improved the AI snips on the episode screen. This was actually running as an A-B test for a while. So for half of you, this is actually not new. But to the other half, hey, you have improved AI snips on the episode screen. So enjoy. Second up, you can now turn off autoplay. You can find it in the player settings and toggle it on or off. Third up, we are now syncing the player progress across devices. So this is for everyone who's using Snipt on multiple phones. We're finally syncing the progress correctly. So I hope this resolves this issue that a small but very vocal group of users had. I know this was a problem. We fixed it. And finally, we have a better snipping model. So the AI model behind it, we've improved various things about it. For one, we used a different AI model. We improved the prompt. One area where you will notice this the most is if you have a custom prompt. For everyone who doesn't know that yet, you can actually create custom prompts that you can use in your snipping. You can find it in the snipping settings. And especially now, you can use longer prompts, and the AI model will adhere to these instructions much, much better. ([Time 0:02:01](https://share.snipd.com/snip/cc0566a8-b51f-493f-bf28-5f9ef6803864)) - **Side Project With Top Podcaster Influenced Roadmap** - Kevin Smith shared that DJ 2.0 development paused briefly because the team began a side project with a top podcaster. - The side project will roll some features back into Snipd and was inspired by testing and real podcaster needs. Transcript: Kevin Smith Okay, also a quick update on the developments of the Snipped DJ 2.0. In my last update, I told you that we're happy to see the positive reactions to the SNP DJ 1.0, and we're excited to develop this further. I think in the meantime, we found the direction that we want to go into with this second iteration. So thanks for everyone who tested the DJ 2.0 beta. Your feedback was incredibly valuable. And I hope that we'll have something to release soon-ish. However, also to be very transparent, we did get a little bit distracted recently. In a good way, however, we've started a little side project with one of the most popular podcasters on Snipped where we're building something for them that will ultimately also make Its way into the Snipped app. ([Time 0:03:33](https://share.snipd.com/snip/7a894178-256d-4dca-b43d-bc89120cac83)) - **Kevin Rose Kept His Feed While Refocusing Content** - Kevin Rose rebooted The Kevin Rose Show but kept the original feed to avoid forcing listeners to find a new RSS. - The show focuses on ahead-of-the-curve tech topics and guests who spotlight future trends. Transcript: Kevin Smith So for today's episode, I have an interview with Kevin Rose. He is most known as the founder of Dig, which he founded more than 20 years ago. And since then, he's hosted several podcasts over the years and has recently revamped his show called The Kevin Rose Show. And he just happens to also be a longtime Snip user. ([Time 0:04:53](https://share.snipd.com/snip/db3f7cde-9f40-47ec-89fc-6e275cf847fb)) - **Podcast Focus On Emerging Tech Trends** - Kevin Rose focuses on finding emerging tech trends and introducing listeners to things they'll talk about in 2–3 years. - He frames the show as a way for listeners to stay ahead amid fast AI-driven change. Transcript: Kevin Rose Yeah, I mean, historically, one of the things that I've loved to do is just be on that forefront of all things technology. So I've been a technologist for my entire career, and I have some gray hairs now, so there's been quite some time. I've been through all the different bubbles and bursts and booms and all of that, and one of the things I love is finding that next thing before it pops off. So finding that thing that we'll all be talking about two, three years from now and exposing my audience to it as early as possible. And thankfully, through the years, I've had a pretty good track record of having guests on that are talking about their next big thing. And that, to me, is a way to kind of stay ahead in the game of life. I mean, things are evolving so quickly with AI and everything that's going on. ([Time 0:05:42](https://share.snipd.com/snip/0608ea8e-2345-4ffe-a2ec-7f40cf0fa924)) - **Start Building Now With Low-Cost Experiments** - Do build now because modern tools make it cheap and fast to experiment; treat many projects as low-cost bets. - Kevin Rose says cloud and AI let you test ideas for pennies versus past tens-of-thousands barriers. Transcript: Kevin Rose Yeah, I mean, it is such a magical time right now where, and part of the reason, so I was a venture capitalist for the last like decade or so. And so I started my investing career at Google Ventures. We had a $300 million fund every year that we had to go out and deploy. And then I went to True Ventures. We were managing a little over $3 billion in capital, investing in great companies like Fitbit and Peloton and all that. And I finally realized this is the moment to be building and anyone can do it. And that is, it's democratizing the idea that we can all be builders in the technology space. And so, you know, five years ago, you'd have a buddy come up to you and say, okay, I got an idea for an app. Who's going to do this for me? And you're like, oh, here we go again. It's going to cost you 75 grand and go talk to this person. I don't think it's a good idea. And now that can literally be done in less than a half a day just by knowing how to work with the technology in the right way. So for me, I get really excited about this moment in time and when it's going to unlock for everyone. And I realized in my podcast on The Kevin Rose Show, I've had Elon Musk on the show. I've had Jack Dorsey, the founder of Twitter. I've had, you know, many, a handful of fortune 500 CEOs on the show at this point. And one of the things that I realized is that they were all tinkers. They all played. They, they didn't know what they were onto something until actually much later. So they just had a random idea that they wanted to try. And they all failed so many times. And failure is just truly admitting that you've learned something new. And I think the more that we can show people, these are the tools to use. Here's how to fail. Here's how not to beat yourself up about it and get back on the horse and ride again and try again. And you can have multiple shots on goal for pennies now when it used to cost you tens of thousands of dollars to get anything off the ground. Like that is the show that I want to create that really kind of brings on the creatives and the people like yourself. I just had you on an episode that talk about their journey, how they got here, how they're building, how they're like in your podcast that we just recorded. It's a wonderful thing that you're only three people. That's amazing. And now I would say that you're probably punching at like a 15 or 20 person team because of AI, right? And that probably wasn't the case just a couple of years ago. And so that's just so amazing of a story to tell and an ongoing story that needs to be told. And so that's the reason for the podcast. I also started an incubator called Basic Intelligence that is all about, you know, kind of eating our own dog food and that we're going up and building all these little apps and ideas. They're all wildly different and crazy and creative and fun. You know, 75% of them will fail and that's okay. ([Time 0:08:14](https://share.snipd.com/snip/96ecb3c3-8d5c-4302-b19c-953b806b4352)) - **Using AI To Learn Hands-On Skills At Home** - Kevin Rose used AI video guidance to rewire his house and learn electrical work himself. - He noted it saved him potentially paying a pro and also admitted he got shocked once from not turning off a circuit. Transcript: Kevin Rose So, you know, a couple examples of this for me personally is I actually have used AI and its video capabilities to turn me into an electrician. And so I went and rewired my entire house and showed it and it was pointing out correctly which wires to touch and not touch and how to turn things off. And I would have paid someone, you know, $1,000 to go do that. And I was able to solve that by myself. And it also explains. And I survived. I did get hit one time by... But I don't want to talk about that. But I did get hit one time by the current that I did not turn off. That's not fun. But it can also bring you up to speed on things and teach you new crafts as well. So, you know, I don't know about you, but when I was a kid, one of the hardest things for me to do in class was admit that I didn't know something and raise my hand and looks silly. Because you'd be sitting there and you'd be in a class and be like, I don't quite understand this piece of the math or I don't understand the subject matter. And I don't want to raise my hand because then I'm the one that looks like I'm. And I can go in now and have these conversations with AI. And it can be something as serious as explain quantum computing to me like I'm a ninth grader. Or it can be something as fun as, hey, I'm looking to get into pottery for the first time. How can I get started? ([Time 0:12:24](https://share.snipd.com/snip/44324bae-0e7f-4f03-8aab-7593438f9a68)) - **First Principles Unlock Transformative Design** - Reasoning from first principles yields transformative solutions rather than incremental fixes. - Kevin Rose recalled Elon Musk describing discarding assumptions to rebuild problems from basic goals like 'hold liquid' for a better coffee cup design. Transcript: Kevin Rose Up with three for you. Kevin Smith Only one. Kevin Rose Only one? Really? I've got three short ones that are amazing. Let me think of one. Well, can I give you two? Please let me have a little hall pass. I've got two really good ones. They're very wildly different. Constraints. Okay. Constraints are good. Okay. I agree with that. As a product person, constraints are good. Kevin Smith If the people listening want more, they'll listen to your show. Kevin Rose That's fair. Okay. So the one I won't mention is a quote from a Zen master that I have on my show. But the one I will mention is actually something that Elon Musk said to me many, many years ago when I had him on the podcast. And he was just getting SpaceX off the ground. And he was just, Tesla was starting to ramp up production. And he started talking about designing from first principles. And it was the first time I had heard him kind of talking about it publicly. And essentially how he boiled this down was that we have all these biases about how things should function in our everyday life, how the coffee cup that I'm holding in my hand should be Structured, how the lid should snap on the way that it does. And there's all this pre-existing information that if we just continue to iterate on top of that, it won't be transformative. It'll just be a sanding down of the rough edges of what already exists. And so his whole thing about kind of reasoning from first principles is forget that you even know about the coffee cup to begin with and say, I want to hold liquid and go back to that level Of just the kind of like beginner's mind. I don't know anything about this problem and I want to solve it with fresh eyes. And if you can truly do that, you can break molds all over the place and come up with new and exciting ways to change the world because you're not working on iteration. ([Time 0:14:00](https://share.snipd.com/snip/71c565c5-00d1-4b55-9796-b7473b7fd887))