Nine years ago I decided to work for an Artificial Intelligence company because I believed AI would change the world. My wife had been diagnosed with cancer several years earlier (she’s fully recovered), and at the time, I was convinced that AI could help us find a cure. I went to work for a Silicon Valley AI company that ultimately ended up being sold – largely because generalized AI was difficult to explain and implement, and clients struggled to use our software to deliver clear business value. While the company didn’t IPO as planned, I was even more convinced that AI was a game changer.
Over the past 9 years I’ve had the good fortune to work with over a hundred customers working on AI and automation projects. It’s become increasingly clear that the rate of technical change is growing quickly, and AI is helping people live better, more productive lives. It’s also clear the promise of AI is recognized universally. According to Deloitte, 92% of CEO’s plan to invest in AI in 2023. As I write this post, the excitement around chatGPT continues to dominate the headlines. However, in spite of the hype, most companies continue to struggle to deliver successful AI projects. Many projects take too long, cost too much, deliver too little value, and are difficult to maintain. Achieving true success with AI-fueled automation continues to be elusive for most companies.
Why has AI success been so hard to realize?
This is a complex question, and there are many reasons AI projects under-deliver or fail. Some of the underlying reasons are the result of systemic industry problems. Startup funding patterns, VC expectations, capacity selling models, procurement, and deployment approaches create headwinds to buying AI solutions. However, regardless of the challenges, AI clearly has enormous potential. Deploying AI projects the right way is existential for many businesses, so companies will keep pushing ahead. At this point, for most companies, AI projects are risky. Yet, despite the challenges some AI projects are successful.
Why do some projects succeed while so many fail?
One of the most important aspects of success AI projects starts with management, and the C-suite’s mindset around automation. When management firmly believes that AI and/or automation helps workers be more efficient, or be more productive, the project has an excellent chance to be successful. In successful projects (and programs), the AI does not replace humans – rather it helps someone (such as a department leader) do work better. For lack of a better term, let’s call this Human Centered Automation. One of the major impediments to automation-fueled success occurs when workers believe automation will take their jobs. When this happens, projects slow down or stall out or just don’t work – often leaving management wondering why.
As a business leader, it’s critical to articulate to the people on the team how new technology will impact them in a positive way, and to tie the AI project’s success to the company’s vision. Many companies over-index on technology and the mechanics of doing a project. They often under-index on the human aspect of making a project work, and the importance to employees of having a sense of purpose, worth and meaning. Part of the reason is people making decisions about the project are generally focused on the best features and functions, and the project is tied to metrics – all IQ functions. The idea that technology will help people (and not replace them) is just the beginning, but it’s a first and important step having a successful project. Winning hearts and minds (EQ) includes creating a clear vision for the future, a roadmap for employees to evolve, and establishing credibility with the team on the ground that they will be a meaningful part of the future. All of the most successful AI projects I have seen over index on the human aspect of deploying AI.
AI will change the world, the rate of change will continue to accelerate, and business leaders must drive change for their companies to realize success. While I hope AI will help us find a cure for cancer soon, there is no doubt we will see rapid changes in many industries fueled by existing and emerging AI breakthroughs. Taking a Human Centered approach to AI projects, using AI to help people and teams do meaningful work better, and communicating a clear vision of an AI-empowered future is a great way to create the foundation for AI success.
