I’ve had a chance now to catch my breath after my second annual trip to the World Economic Forum in Davos, Switzerland. And I keep coming back to something that I discussed with my business and broadcasting partner Daniel Newman in our wrap-up video: In 2025, participants talked about what would happen or could happen with AI, but the actual applications discussed were mostly limited to pilot programs and proofs of concept. To a striking degree, this year’s WEF — formal presentations and smaller conversations alike — reflected a real maturation of the dialogue, with some of the smartest and most practical-minded executives you’ll ever meet describing very grounded operational approaches to implementing AI in the present tense.

Real-World Use Of AI In The Enterprise, Industry And Beyond

HPE CEO Antonio Neri has been talking about real-world enterprise use cases with me for a while; in Davos, he told us that his company has 70 of them deployed in production for customers. Jeetu Patel, who’s president and chief product officer at Cisco, talked about getting beyond “hyperbolic” AI narratives in the media and society at large — where AI must be either so effective it will put everyone out of a job or else so bad that it can’t accomplish anything — and instead focus on the real-world challenges where it’s making an impact.

Nakul Duggal, who heads the automotive and industrial businesses for Qualcomm, talked about the great speed of change in the industrial space, and the specific challenges of implementing AI at the edge. (I took part in some great conversations about edge AI at this year’s WEF — including a one-on-one onstage at USA House with Duggal’s boss, Qualcomm CEO Cristiano Amon.) And Dr. Eric Xing, the computer scientist who’s the president of MBZUAI in Abu Dhabi, put forward a very commonsense and humane view about using AI to free people from drudgery — to give them more options in life. (I’ve written about MBZUAI’s work in reasoning models and world models.) He made the comparison to a century ago, when 90% of people in the U.S. worked on farms; that number is in the range of 3% now. Would you rather be stuck on the farm with no viable alternatives? Or have your choice to pursue farming or countless other ways of life? I agree with Xing that AI will likely have a similarly profound impact on how we live.

Choosing The Most Practical Starting Points For Enterprise AI

Rob Thomas, chief commercial officer at IBM, is another leader who is immensely practical. He doesn’t concern himself about using AI for moonshots that could completely transform a business in one go. Instead, he’s spent the better part of the past decade — since way before the current generative/agentic AI boom — advising IBM’s customers to use AI on the boring, repetitive tasks down in the trenches. Control what you can control, move briskly from pilots to production and reap the benefits.

I don’t know how much Thomas has ever talked with Anand Swaminathan, who’s the global leader of the Communications, Media & Technology business at Infosys, but I think they would find a lot to agree on. Swaminathan has an outlook that is similarly tied to on-the-ground business reality — which makes sense, considering the degree to which Infosys is in the implementation business. He wants companies looking at areas where they can spend less money thanks to AI, alongside those areas where new investment can make a difference for acquiring customers, expanding offerings, or the like. Fundamental business questions, not speculative what-if daydreams.

These are just a few excerpts from the conversations we had in Davos. For more detail from each of the executives mentioned, and to see interviews with other leaders from Snowflake, Check Point, Celonis and more that are actively engaged with AI implementations, check out “The View from Davos” video library.

WEF 2026 Showcases A More Mature View On AI

There are other things that will stick in my memory about this year’s WEF, like the airtight security throughout the town ahead of President Trump’s speech, including heavily armed security forces and snipers on rooftops. And there were a million business and technical details to digest about everything from digital sovereignty to data structures supporting agentic AI.

But above all, I think I’ll remember this as the year enterprise AI grew up.