On this week's episode of Inside Outside Innovation, we talk about the red pixel in the snow, why MVPs should be delightful, and the robot AI deployment gap. Let's get started.

Inside Outside Innovation is the podcast to help innovation leaders navigate what's next. Each week we'll give you a front row seat into what it takes to grow and thrive in a world of hyper uncertainty and accelerating change. Join me, Brian Ardinger and Miles Zero's, Robyn Bolton. As we discuss the latest tools, tactics, and trends for creating innovations with impact, let's get started.

Podcast Transcript with Brian Ardinger and Robyn Bolton

[00:00:00] Brian Ardinger: Welcome to another episode of Inside Outside Innovation. I'm your host, Brian Ardinger. And with me I have Robyn Bolton. Hello, Robyn. How are you? 

[00:00:48] Robyn Bolton: I am great. How are you, Brian? 

[00:00:50] Brian Ardinger: We are surviving the cold.

[00:00:52] Robyn Bolton: The sub-freezing temperatures. Yes, I know it's January, but that doesn't mean it has to be as bitterly cold as it is. 

[00:01:01] Brian Ardinger: Absolutely. Well, hopefully this conversation will warm people's souls and hearts. As we talk about innovation in its various forms, we'll get right into it. We've gathered a couple of different articles that resonated with us over the last couple weeks.

How AI and Drones Are Transforming Search and Rescue Innovation

So, the first article we want to discuss is titled A Red Pixel In the Snow: How AI Solved the Mystery of A Missing Mountaineer. And this came from the BBC. It's very fascinating article for a couple different reasons, but the basic premise, it's a story about a missing mountaineer. This person was hiking and went missing a 66-year-old hiker and they sent out all the helicopters and that to try to find him. They were unsuccessful, but closer to the spring when some of the snow was melting, they decided to go back out and see if they could actually find the body.

And they used drones and AI, as a way to map the area. And what they found was they could put all that AI pictures into the system and they were able to find a red pixel in the snow that was effectively his helmet, that they were then able to find the person and go and retrieve the body and such.

What I found fascinating about this is, again, in this particular instance, it wasn't successful in finding him and saving him, but just the ability for new technologies like drones, just taking random pictures and then putting that in through the AI and having the AI look for anomalies. They were able to identify something that they couldn't have done in the past, and obviously at a much faster speed than they could have done in the past as well.

[00:02:26] Robyn Bolton: This was such a great story, tragic ending for this hiker, but a phenomenal story of when AI is good, it can be great. And you know, it's an instance of AI doing something that humans are not good at. We're not good at finding a pixel in the snow. We have bias when we see things, and so we're more likely to overlook something red. Because we just don't see it.

So, it was just a great story of how AI is augmenting what humans do. It is taking things that need to get done that we're not good at, and that it's equipped to do better than us. And you know, even though this story didn't have a happy outcome for the hiker, I bet the family is still happy to have him recovered and not be wondering. And as AI gets better, there's probably more people who will be rescued because of it. So, I thought it was just a wonderful story.

Augmenting Human Judgment with AI and Drone Technology

[00:03:25] Brian Ardinger: And it was interesting just to read through actually how the AI worked. The software managed to detect a kind of a red color, even though the helmet was in shade. So again, a human might not have been able to detect it, and it was very good at identifying anomaly.

So, it didn't necessarily say this is exactly where the hiker is, but it was able to go through the mounds of image data and say, here's some possible places. Humans still had to go through and actually find it, but it again, sped up the process.

And then I guess the other interesting point about this is the other technology, if you stack that on top of AI, the drones themselves, being able to get into crevices and places where traditional helicopters couldn't get into.

What's interesting is again all these particular technologies that we're talking about are hitting all at once, and when you start looking at the cumulative effect of how these things can add value or create interesting solutions and that, that's what's accelerating innovation. It's this ability to add on, and it's not just one thing that can make a difference. It's this combination of things. 

[00:04:20] Robyn Bolton: And it's the combination of the technology and the humans versus trying to use the technology to replace humans. I mean even the drones, as you mentioned, the drone operators had to go to the sites and train on how to fly the drones so that the drones could see into the crevices and into the shaded areas.

And so. It's and not or when it comes to technology, it's not, okay. AI has replaced the humans, or AI can't do this at all. It's only humans like, no, put 'em together and let everyone do what they're best at.

MVPs, Product Sameness, and the Push for Delightful Experiences

[00:04:53] Brian Ardinger: All right. The second article is titled Why MVPs Should Be Delightful,and it's from the UX Collective.

And this was a great article. MVPs are near and dear to my heart. We do a lot when we're, you know, launching new products and working with startups, and we always talk a lot about the MVP. This particular article by James Skinner. It really talks about the fact that as we're living in a world that AI is now omnipresent. Workflows, you can spin up in a dime at a low cost. It's creating this kind of sea of sameness. And you know, lately products have begun to look the same and feel homogenous. And how do you create new products, new services that delight the user, not just meet the bare minimum of the functionality.

His call to action basically is, you know, stop looking for the good enough or just the functional aspect of your product or service, but how can you inject delight into it?

[00:05:43] Robyn Bolton: I am going to roll out my soapbox on this one. And it comes back to, the reason I have a soapbox is what is an MVP? It started off as a term, a minimum viable product. Literally, minimum viable. A true MVP should just function. We shouldn't be worrying about delight. We shouldn't be worrying about, you know, how does it make the customer feel like should it function?

Solve the problem that we need it to solve. And then there's version two and version three and version four. And then when you get to kind of the quote unquote final version that you are shipping, like, yes, it should delight people. Yes, it should be differentiated. But if we're going to be super strict about language, which I believe is very important because it avoids confusion, a true MVP actually shouldn't be delightful. It should just work. And then what you ultimately launch should absolutely be delightful. But that's not an MVP. I will get down off my soapbox now.

Minimum Lovable Products and Differentiation in a Competitive Market

[00:06:53] Brian Ardinger: I agree. I think where this author was talking about is I think a lot of people even miss the viable part and they launch and don't actually do anything for it. So, if you at least have the mindset of it's got to be more than just solving a puzzle, it's got to create an environment that gets the person to the value proposition faster, better than before. 

I think that's one of the important things he really talks about. He actually has an addition saying, you're looking for a minimal lovable product, so MLP and the ability to create delight is a core component ultimately, if you're going to be successful. Maybe not as the first time you get something out in front of a customer, but at the end of the day, things that delight, especially in in a world where everything's the same, may give you the chance to have a leg up in the world that's so competitive with everything looking the same, feeling the same functioning, the same.

[00:07:42] Robyn Bolton: I know we've talked about on past podcasts about how it feels like things are converging towards sameness and the importance of being different and being delightful and bringing the human element in and thinking about the humans who use your product. So, a hundred percent. Don't call it an MVP. Call it a minimum lovable product. Call it a minimum delightful product. Give it a different acronym, and then I'm on board.

The Physical AI Deployment Gap in Robotics and Automation

[00:08:08] Brian Ardinger: Let's pivot. We talked about a couple weeks ago an article around the Robot Olympics. So the third article is kind of following in on that trend and maybe a little bit deeper dive, but it's from a16z, and it's called the Physical AI Deployment Gap.

And this article talks about the gap between the frontier of robot research and actually the deployment and the gap that's going to exist on, it may be very different than other products. It's unlike the iPhone has everything solved in one nice package. The breakthrough in robots may be a continuation of creating a platform or an operating system that enables these devices and tools and applications to work together more akin to like an Android type of an experience because the ability for a robot to exist in the world and deal with the uncertainty and complexity of the world is much different than just creating a software product or something along those lines.

But found it interesting and fascinating how the difference and how people think about the deployment of robots as a different kind of discipline than other types of technologies.

[00:09:13] Robyn Bolton: One of the things that in addition to everything you said that I thought this article called out really well is the enormous gap that exists between the lab and the shop floor. And so, there's been just an explosion of robot demos that we've seen. I was watching a video just earlier today about an adorable robot that I'm like, I want one, it's so cute. But it's a demo and it's in a lab environment and it's in a very controlled environment.

And so, you're seeing lots of demos, but then when you look at what's going on in the factory floor, the production floor, you're seeing a very different type of robot that has a very specific, specialized task and looks much more like we're used to. That's because one, the physical environment of a production floor versus a lab, totally different in terms of stuff in the air mess, all of that stuff.

But then also like if you have a robot that's 95% successful. That is a huge problem when you go to production because that means you're making like 50 manual interventions a day when you think about how often a robot is doing something. And so, the reliability of what a manufacturer or producer needs just isn't there yet. So, and that was an interesting call out as to why we're seeing a lot of stuff in the demo space and yet our production story is still kinda look and feel the same.

Why Scaling Robotics Is Harder Than Scaling Software

[00:10:41] Brian Ardinger: And I think we often overlook the scale portion of this. So like, like you said, if you can get 95% certainty in a lab environment and you put it out into the real world and it becomes an 80% reliable type of solution, that sounds great on paper until you're realizing that, you know, the whole reason why you put a robot system into place is to do something at scale.

And 80% of scale is very different than 99.999% at scale. And the amount of failures that would occur hundreds of times a day requiring, you know, constant human intervention and things like that are, that's where the struggle and the gap actually exist.

And closing that gap is much, much harder than it would be in traditional software development or things along those lines. And it's kind of like a compounding. And so, you have 80% right on this particular task. And the second task it provides, is 80%. So, then you have a doubling, a compounding effect across the board. So that's why it's so hard, and that's why we don't have robot armies storming the capitol at this point. 

[00:11:36] Robyn Bolton: For anyone who is concerned about that happening. The author also makes an argument that there's not going to be kind of an iPhone moment as he calls it, where like suddenly everything changes. Feels like that's what happened when the iPhone came out, that this is going to be a much more gradual evolution because there are so many factors to consider in getting a robot to that 99.9 9 9 9 9% that it needs to be at scale.

[00:12:01] Brian Ardinger: And that the gap won't be closed through pure research, it's going to take actually implementation and getting it in concert with infrastructure and tooling and operational capabilities that don't currently exist. It's not going to be something that is spun up and finished in a lab. 

[00:12:17] Robyn Bolton: So, we have some time before the robot Army takes over.

[00:12:19] Brian Ardinger: At least a week. 

[00:12:21] Robyn Bolton: At least a week. Excellent. Go live your best lives, folks.

Digital Detox Trends, Appstinence, and Bricking Your Phone

[00:12:24] Brian Ardinger: That's right. Well, that brings us to our tactics to try. My tactic to try this week, it was based on another article that I read called Bricking Your Phone is the New Dry Januarywas in Business Insider. The article talks about a new form of kind of dry January or abstinence, and this is spelled “appstinence” with APP and the focus on how younger ones are especially focused on how do they make rather than dry January bricking their phones. So that they make it a dry January from the media and from the tools that they use, et cetera, cutting back on screen time. And so that has become the new flex. So not just drinking in January, but, but bricking your phone and a new dry January opportunity. 

[00:13:08] Robyn Bolton: My tip to try is the exact opposite, and mostly in response to, you know, the news that Clawd Bot has come out. And just so much focus on how you can create apps or agents or things using, you know, ChatGPT or Claude or Gemini to do things for you. 

So we were talking about before we hit record, some of the things people are talking about doing, and I was sharing a story of one person on social media talking about how he set up his little Clawd bot agents to do things, and a few days later, suddenly it gave itself an avatar as an owl. And then a couple days later it started talking to him because it had given itself a voice. That's creepy. I'm not going that far. But I am going to start dipping my tone to just seeing what Clawd Bot can do and running a little, a few experiments here and there. But if it suddenly comes to life, we're all done.

[00:14:00] Brian Ardinger: Yes. 

[00:14:00] Robyn Bolton: Done with that.

Experimenting with AI Agents Without Letting Them Take Over

[00:14:01] Brian Ardinger: The sales of Mac Minis have increased this week, I think, from people trying to create their own little infrastructure system of AI agent. Let it run and, and see what happens. So report back. If you're listening to this and you've tried this, please let us know how it's going. Well, that concludes another episode of Inside Outside Innovation. Thanks for coming out. We'll see you next time. 

[00:14:18] Robyn Bolton: See you next time. 

[00:14:22] Brian Ardinger: That's it for another episode of Inside Outside Innovation. Today's episode was produced and engineered by Susan Stibal. If you want to learn more about our teams, our content, our services, check out insideoutside.io or if you want to connect with Robyn Bolton, go to MileZero.io, and until next time, go out and innovate.


Articles Discussed