Below is a partial transcript of this Great Question podcast: 

Sarah Mattalian: I wanted to start off by asking about the top reasons why AI pilots fail. Can you outline three main reasons why this happens?  

Mike Fedorov: It's quite simple. First, I'd say this is because people are overly excited about the technology and look at that as a shiny thing to apply in the business rather than the other way around. something that the business needs and then you find the solution for. Second, it's about the learning curve for people both on the technology and the business. People often tend to start with the most exciting complex problems, and that's not the best way because complex problems are difficult to solve and oftentimes don't bring success. And then third and probably the most interesting piece is how data fits into that. The thing is, the data is the blood of the artificial intelligence and people make either of two mistakes typically, either assuming that the data is already in a great shape or refusing to move forward until it is in a great shape. So the third pitfall is what I think most of the issues are happening around.  

SM: Can you expand on that third pitfall more and that data approach element, and why that can be kind of a creative problem for companies to solve?  

MF: Let me explain. Think about which people are typically bringing AI to companies. They are normally very good, bright engineers who know exactly how the technology works. But they are much less aware of how the operations work because they help so many different operations. So they prepare a solution that works awesomely on simple sample data sets used in the tests. And then guess what? It fails. It just falls apart in the real life. So what happens is what works excellent in the test case meets edge cases, missing data, broken data, user mistakes, you name it.   

Podcast: With outdated and isolated OT, your AI strategy may be built on a blind spot 

The real production data is messy. It's very far from the idealistic model. So when AI engineers come without the experience in the industry, it happens exactly that.   

Think about the elite Olympic athlete goes into special forces and has a mission to swim over a muddy stretch of water. The athlete who is trained to swim really well, really fast in the Olympic pool with the clean water would now meet the mud, the occasional fish, maybe occasional logs. Do you think this mission is going to be a success? I don't think so. So now imagine the same athlete just refusing to swim over the stretch of water because the water doesn't meet Olympic pool standards. So those are two modes of operation. The first mode you dump an Olympic athlete into muddy water with the logs and you get a pure failure. And the second, you don't even start the project because you think that until all the data is cleaned up, you can swim. Neither works to help the mission. So those are two biggest ways how people make a mistake, either requesting that the solution just works, don't mind the data, just make it work. And the second one is how you don't do anything until all the data is cleaned.  

Podcast: Why is manufacturing such a huge target for cyberattacks? 

SM: So to summarize, having messy data is akin to the logs and the algae and having muddy waters, and companies need an AI solution to be able to act like an athlete that is trained to swim through these muddy waters.  

MF: That's exactly the point. And I think this is where you understand it much better than 80% of the companies judging by the results. The real solution is not to do either extreme, the real solution is twofold: first of all, make sure that your special ops person of the mission. knows how to swim in the real water. That is, an AI solution needs to understand and maybe not be as fast and as accurate as the ideal solution in ideal water, but needs to be processing the muddy water, needs to be processing the muddy data, needs to be able to give good enough results in the muddy real world situation. Now the second part of the solution is to set up the governance, the continuous improvement process and the technology to continuously improve the data quality because the second camp, the second school of the data quality first, actually right, the real good results cannot be achieved on the bad data. But that is not the extreme process of don't do anything until you get the water absolutely clean. This is the process where you need to do something, do something that works, and then use that something. to modify people to clean the data to get better results over time.  

See also: As AI adoption matures, data center controversies move to the fore 

SM: So what I'm hearing is those “winning” companies have solutions and AI agents that can really, they're able to swim through these waters. And it sounds like they're also kind of designing these solutions with outcomes in mind. So that being said, can you expand more on what those key outcomes are that companies should be considering?  

MF: The key outcomes on the data side are simple. That needs to be a solution that allows to work with the data that they have right now, understands the imperfection and can treat it with the right measures. And at the same time, the far further reaching outcome is to get a solution that helps improve the data. The guidelines, the tools, the governance, to get the data to a better shape.