AI Automation Is Here to Stay (Shaun Dawson)

Full transcript

Welcome back to the podcast, guys. Today we are joined by Sean Dawson, a longtime automation leader and entrepreneur, most recently at UiPath and one of the early people building in RPA back when the category was still new. Sean, welcome to the podcast. >> Thank you so much. I really appreciate being here. >> To start, could you share a little bit about yourself and your background? >> Yeah, absolutely. I kind of lucked into the RPA space about 15 years ago now. I was running my own little software business and I got approached by a couple of guys who wanted to start an RPA company. This is 2012 before anybody even knew how to spell RPA and we launched this little company called Virtual Operations and we just happened to be in the right place at the right time. We developed relationships with the three main RPA players, Automation Anywhere, UiPath and Blue Prism at the time, and we just caught the wave just as it started happening. It was a really, really great experience.

Ended up helping to build a a 2000 plus automation team at Cognizant, went on my own for a while, ended up kind of finishing up my RPA career so far at UiPath. >> So, as I mentioned, you were involved in RPA very early. What did people misunderstood about automation back then and what turned out to be true that most people didn't see coming? I think the thing that turned out to be true and this is really applicable to the AI agent space. All of this hype that we have about AI is exactly it mirrors exactly the hype that we saw in the early days of RPA. And so the thing that was true that turned out to be true was that there are huge benefits associated with getting the technology right. It can really transform. I can tell you an example of we came in and we did an implementation at an insurance company where they were looking at a manual process that had to scale up by a factor of five because of some regulatory requirements. They were spending $5 million a year on the process.

They didn't have the $25 million to spend on a scaled up process. And so we came in to help them automate it and we were able to automate that process and actually handle the scaled up load and reduce their cost. They started spending $3 million instead of $5 million on the process and they were able to do five times as much work. Now, that's the true thing. Here's the thing that people didn't understand and didn't appreciate is the amount of effort, design, and intentionality that it took to actually do that successfully in a way that was able to scale and in a way that wasn't particularly brittle and in a way that would be able to stand the test of time. It took a tremendous amount of what I call engineering discipline. That is a bunch of frontend effort that you had to put in in order to succeed. And that's what I see happening in the AI space. Now, the fact is it's the same thing.

It's an amazing demo. The initial 10 minutes with the product is just really amazing what you can do and what you can demonstrate. But to actually get and capture the business value of the technology takes a lot more effort. The technology implies that it can do a lot more than it actually can. So it really takes some real engineering discipline in order to obtain and capture the business value of the technology. >> Do you think AI automation is here to stay? >> Yeah, no question. It is absolutely going to pervade our work and lifestyle. I don't know if you've ever sat down with AI coding tools, for example. I come from a coding background and so that's one of the places where it has just absolutely started creating lots and lots of value with all the attendant problems with that as well and it's my opinion that AI cla code for example is so tremendously powerful and it can really provide leverage for somebody who's already a really good software engineer to just develop so much more so much more quickly. But it also provides that exact same leverage to somebody who doesn't know what they're doing. They can build all sorts of unmaintainable and horrible spaghetti code just as quickly or even more quickly than the person who can generate really really really good code with a really really good architecture.

I think it takes somebody who's a very good developer to be really effective with that stuff. But Anthropic who had released Cloud Code, they were watching their developers and realized that their developers ended up using cloud code for a bunch of stuff that wasn't related to coding. And now they've released what is essentially a version of cloud code for regular business people. And it's the exact same agentic interface, but now you're using it to to handle your emails or your calendar and things like that. And it's just tremendously powerful. But again, you need to recognize that this power comes with some responsibility, which mean you have to be a little bit careful. But then, man, once you get the racehorse pointed the right direction and in the right lane, it can really create a lot of value. And so, that's what I see happening. And it's probably going to take longer than everybody expects.

And it's not going to exactly come to fruition in exactly the format that everyone thinks it will, but it's definitely here to stay and it's definitely going to happen. >> You really explained my situation. I tried called code for my part of business that I'm trying to automate and it's exactly what you've described. Like I don't come from coding background. So I have no clue what it does. I just write things and something happens and yeah it's like a magic box. >> Yeah. The nature of that problem is that anything cla does it gets somewhere between 80 95% of it right. And if you can go in there and to, you know, imagine you were weaving a tapestry and it was a machine weaving the tapestry, but it only got it 90% right and it kept making mistakes and mistakes and mistakes. Well, those mistakes over time compound. And that's what you end up seeing with novice developers try and do AI coding. Those mistakes compound, compound, compound, and everything kind of works until it doesn't.

and then it all of a sudden blows up and then you have no way of fixing it. Whereas if you're in there like kind of monitoring it, you see the mistake happen and you're able to fix it and you're oh, you're able to make fix that thing or you send cloud code off in a little rabbit hole and it comes back and it just blows everything up, but you know how to back out from that and go back to a previous stable version in your source code repository and stuff like that or just take the bits that you're happy with or rearchitect as necessary. Those are the things that AI is really really good at. So what I would recommend to folks that are in your situation is there is no better learning tool than AI. AI can help you learn a new language or learn new technologies or any of that stuff, a new framework.

Oh, you don't know how to use library or something? AI is tremendous at that. So use it to give you the skills and you the engineering skills to be able to build stuff, but you need to be able to build the thing on your own without AI and then you can use AI to speed it up. >> It amplifies your skill. So if you are already good at coding, then it just makes you 10 times more powerful. >> That's exactly right. And on the negative side, if you're not good at coding, it amplifies your mistakes. Yeah, and they are 10 times worse. That's right. Where do you see these agents, co-pilots, and Gen AI changing the next chapter of automation? Do you think it's all hype? There's a lot of hype. I was around during the very first internet bubble. I graduated UT Engineering School in 1999. This was a time when folks who were in Stanford were getting $200,000 a year offers to quit computer science school and go to work for living.com or whatever the startup of the day was.

You know, 95 98% of those internet companies that were around in 1999 were gone by 2002. It feels like that now. That time was a time of just absolutely amazing upheaval. It was a time of a lot of optimism and it was a time of hype. Literally, you could just put.com on the end of your domain name and all of a sudden your valuation would shoot up by 30 40%. The exact same thing is true today except it's not.com anymore. It'sai, right? So, you need to take that, you need to recognize that and you need to recognize that there's a ton of hype and that means that you have to take everything you read and everything you see and all the claims that are being made with a grain of salt and then that doesn't mean that the technology is not suitable for purpose. It doesn't mean you take the bare attitude.

The fact is that 5 years from now, the way that we work, the way that we live, the capabilities that we have are going to be so different than they are today. Today will be unrecognizable to the way the world will be in five or 10 years. Exactly the same way it was. Imagine the way the world seemed before smartphones, right, in 2005. And then by 2010, everybody was carrying a supercomputer in their pocket. Think about how the world is now so much different because we all have a supercomputer with a great camera on it. That is chicken scratch compared to what AI has the capability of doing and it's going to happen. You would never have been able to predict the kind of businesses that could have existed in the smartphone era versus before it, but they're there. And the same is true for AI. You recently marked your last day at UiPath. What are you excited to build or explore next? Man, it is a really, really interesting time to be without a job in this space.

It's scary in the sense that our industry is getting disrupted. A lot of people are getting laid off. There's a lot of labor out there looking for something to do. But the opportunities are enormous. It is now possible to do things that was just completely impossible 3 or 4 years ago. It's definitely scary, but it's also really exciting. And I I have a couple of projects in the works that I'm pretty excited about and some stuff some passion projects that I've been working on that are really interesting. So you definitely watch the space and and I look forward to talking about what the next thing is. >> Is there anything that you could share about the projects? Maybe the field. >> Yeah, I'm currently uh consulting for a number of AI startups and I'm in the investor space doing a bunch of that stuff. But I also own an aviation accessories company and we are uh looking at what we can do to kind of really make an impact in the aviation market.

That's that's one of the things that I'm really kind of passion I've been passionate about flying airplanes and being on all sides of the aviation business. And there's some ways to apply technology in that business that are pretty interesting. That's definitely one of the things that um that I'm evaluating. And there's also some stuff in the AI space that I'm currently evaluating and and some opportunities with some businesses that I'm consulting. >> I'm curious, how do you get these pilots? >> You know what? Aviation space is old school. It is word of mouth. It is making friends. It is hanging out at the airport and talking to people who have airplanes. I mean, literally today, as soon as we get done recording, I'm going to be heading into the airport. I've got some meetings with some instructors to talk about a few things and I've got to pick up some hardware from somebody. You go out to the airport, make friends is really how it work. >> Okay. Well, Sean, do you have any message to put out to the people in the same industry as you?

>> Yeah, I have a business mentor. I talk about this a lot. There's probably two thing that I would mention. Both of them are things that I like to talk about a lot which is first thing is I used to have a business mentor that one of the things that he would say and it really rang true to me and I found this to be true in my whole career is that your biggest problem is also your biggest opportunity and vice versa. You have to recognize both sides of that coin. There's a tension and you have to recognize that tension and you have to respect it. And so that means if you're focused on what a big problem it is, you also need to recognize the opportunity side and you need to cultivate that. And if you're focused on this great opportunity that you have, you also have to recognize the problems that come with it and you have to address those. The second thing I have a um another lesson that a business mentor of mine, he really drove it home for me in a real practical way was that you need to be doing something you love with people that you like, trust, admire, and respect.

And if you ever look around the room and it's been a while since you've done something you love and you're not really surrounded with people you like, trust, admire, and respect, then you're in the wrong room. And it's really important to be in the right room. Even if it's a little risky and even if it's a little scary, go find the right room to be in and I promise it'll work out. >> Awesome, Sean. Thanks again. I will add links so people can check you out in the description. And thank you for listening. We will see you in the next one. Thanks a ton.