AI innnovation

Using AI To Fix The Innovation Problem: The Three Step Solution

By /

I did a podcast this month on how to use technology to increase innovation. Now, I’m not a fan of innovation for its own sake. If you have something that works, innovation can be a bad thing because it may break what works. But if it doesn’t work, or it works badly, and you’ve looked at past practices by you and others, then innovation plays a role in terms of solving a unique problem that appears unsolvable or sub-optimal.

Innovation is tough, I once got into an argument with a soon to be let go Ford CEO on innovation and I told him that the path he was on was likely to lead to him being an ex-CEO, I’ve sadly done this a lot over my career, and even more sadly, I’ve generally been right. My worst was with Steve Ballmer at Microsoft who really got upset with me before he was let go.

But there are three solid steps that can be used to improve innovation in a company.

First Step: Innovation Isn’t An End Goal

Innovation is a way to solve a problem, but it often isn’t the best way. It’s like looking at surgical triage and thinking that it would be great as normal practice when it is only great when you lack the resources to help everyone and would be a disaster in most any other case. You don’t measure people on innovation, you measure them on meeting goals, and if they innovated to meet those goals, and the result works, you don’t penalize them either. Too often people are punished for doing something in a way that was new but doesn’t otherwise damage the company simply because it wasn’t how it’s always been done.

So, innovation is a tool, and to assure this tool isn’t abused, you need to put it in context with the result. If someone innovated and screwed something up because they didn’t bother to learn the right way they should be penalized, however, if someone facing an unsolvable problem is innovated and still failed, they shouldn’t be punished because, even through they failed, they used innovation properly.

Skunk Works

One of the most successful ways to drive innovation is to create Skunk Works project teams. These are teams of highly qualified people that understand the problem deeply but tend to be rebels. What you do is send them to a location remote from the company, give them a decent budget, provide them with core resources, and then remove the unnecessary shackles that corporations put on employees.

I’ve seen this work a number of times when we were struggling to create a product that met a unique market or customer need and needed to think outside of the box to get there. Some of IBM’s more innovative products were created this way. Having worked at IBM, within the company, I have noticed that there were too often nearly a magnitude more people who would block an innovative approach than were trying to implement it. This was due to a number of issues, such as fear of being overshadowed, fear of looking stupid in the face of a better solution, fear of losing status to a better solution, and fear of change. A skunk works effort, done right, can get around all of that and result in a far faster, better solution than might have been accomplished in the company.

Being Failure Tolerant

Innovation comes with a high potential for failure that is in line with its high potential for surprising success. But if people are afraid to fail, they will be afraid to try anything new and thus won’t be able to innovate out of the problem they find themselves in. This was one of the “discussions” I had with that old Ford CEO, he agreed that you have to be tolerant of mistakes but then said if anyone made a mistake with the F-150 pickup they’d be metaphorically shot. This meant he wasn’t tolerant of mistakes and that the F-150 was at risk.

You do need a carrot and stick approach but if someone takes the risk of innovating and is successful they should be publicly rewarded and part of the reward dialog should be on the mistakes they made that eventually resulted in success which conveys that failure, as long as it is on a path to success, it OK. The stick is for people that innovate when innovation isn’t needed, resulting in avoidable failure. People innovating because they don’t want to do their homework should be penalized because they are harming their company.

Wrapping Up: What AI Brings To The Table

So, you can make Innovation part of the creation process, you can create a Skunk Works to focus on innovating around a hard problem, and you can reward good innovation practices and punish bad ones so what does AI bring to the process?

AI can create concept variants at machine speeds, and offerings like NVIDIA’s Omniverse can allow for simulations of the result in virtual environments so the cost of those mistakes is minimized. With AI tools you can fast fail at an alarming pace which could destroy companies if this failure is in the real world and effect real people. But with strong simulators you can test the offering in the metaverse so that no real person is damaged, and the knowledge of the effort remains contained in the company.

So, AI can perform two roles, one it allows for the really rapid creation of multiple solution concepts, and it provides, in the Metaverse, a way to test the most compelling of the group in extremely safe and private ways significantly increasing the probability of a positive outcome.

In short, while I’ve not been a fan of fast failure, largely due to damage to the firm and its reputation if that failure is made public, with AI you can do fast failure faster and far more safely making it also a far more viable path to success.

Latest posts by Rob Enderle (see all)

Related Posts

Scroll to Top