From AI Ambition to Execution: Michael Dell on How Leaders Turn Strategy Into Competitive Advantage

Almost every enterprise has an AI strategy but very few have the execution to match it. The companies pulling ahead aren’t necessarily the ones with the boldest roadmaps; they’re the ones making deliberate infrastructure decisions, prioritizing effectively, and building with long-term flexibility and scalability in mind.

Recorded live at Dell Technologies World 2026 in Las Vegas, Patrick Moorhead and Daniel Newman sat down with Michael Dell, Founder, Chairman & CEO of Dell Technologies, to discuss what separates organizations successfully operationalizing AI from those still stuck in the pilot phase. Michael shares his perspective on how enterprise leaders should think about balancing urgency with execution, navigating competing priorities, and building infrastructure foundations that can support AI across cloud, data center, edge, and endpoint environments.

The conversation also looks ahead at what characteristics will define enterprise leaders in the AI era over the next 12 to 24 months and what executives need to be doing right now to position their organizations for that future.

Key Takeaways:

🔹Prioritization separates AI leaders from AI followers. With no shortage of AI vision in the market, the organizations driving outcomes are the ones that have identified which initiatives connect directly to business results and have the discipline to focus there rather than chase every opportunity.

🔹AI leaders balance speed with discipline. The biggest risk comes from moving too slowly to compete or moving too fast without the governance, infrastructure, and operational alignment to sustain progress.

🔹 Execution, not technology, is the differentiator. Many organizations have access to the same models, the same cloud platforms, and the same tooling. The ones that win do so because of organizational focus, change management, and the ability to move AI from proof of concept into production at scale.

🔹 Where AI runs matters as much as what it's running. The conversation has shifted from whether to use cloud to determining the right environment for each workload. Flexibility, cost, control, and governance must all be weighed as AI expands across cloud, on-premises, edge, and endpoint infrastructure.

🔹 The next 12 to 24 months will define enterprise competitive position for years. The organizations that emerge as AI era leaders will be those that have already built the operational foundation, infrastructure strategy, and leadership alignment to compound their advantage as the technology continues to accelerate.

Watch the full video at sixfivemedia.com, and be sure to subscribe to our YouTube channel so you never miss an episode.

Disclaimer: Six Five Media is for information and entertainment purposes only. Over the course of this webcast, we may talk about companies that are publicly traded, and we may even reference that fact and their equity share price, but please do not take anything that we say as a recommendation about what you should do with your investment dollars. We are not investment advisors, and we ask that you do not treat us as such.

MICHAEL DELL:
When you're changing a lot, it's good to overcommunicate. Explain why are you doing this? Why is it important? What happens if you don't do it, right? And sort of everybody gets on board and understands that this is the future.

PATRICK MOORHEAD: 

The Six Five is on the road here in Las Vegas Nevada. We're at Dell Technologies World. We just got through the keynote. It was awesome. In fact my Dell XPS laptop literally almost vibrated off the table at the beginning with all of the base. But that was it was this was all about reality in enterprise agentic workflows.

DANIEL NEWMAN: 

Yeah, bringing AI to life. The keynote ended with a great back and forth with Michael and Jensen where they were talking about bringing AI to life and that we've really just hit this inflection where AI is just now really becoming real. And it was kind of fun to do that little bit of a playbook because you remember three years ago what we were hearing about what generative AI could do. And we were so excited about the possibilities. And then another year came and we started hearing about the next thing it could do. And I still kind of go back to like the December 2025 model inflection. And so much of what happened there is really building up to this moment right here, Pat. And it was just a great keynote overall.

PATRICK MOORHEAD: 

Yeah, it was a great reality check. And it's closer than I think anybody thinks at this point. So let's bring in our special guest. Michael, great to see you.

MICHAEL DELL: 

Great to be with you guys. Thanks for being here at Dell Tech World 2026. Always appreciate the great work you all do.

PATRICK MOORHEAD: 

No it's good. And you know the most important question Michael were you actually driving that rack up on stage. Was that you like driving with a steering wheel as it was making its way onto the stage.

MICHAEL DELL: 

No I was not. I was not doing that. We got we got it. We got a rack guy.

PATRICK MOORHEAD: 

I wasn't an agent or something like this.

DANIEL NEWMAN: 

Yeah you know I don't really know actually all right kind of happens it looked cool though it did look it was kind of cool coming out of smoke and there it was all the sudden big and beautiful and you didn't even have to sign it because your name was already on it it's already it's it's it's like lots of time on it so it's like all over that it was all over that will listen to people here at Dell tech world are. hear from what we were just talking about this inflection of powerful A.I. to drive the enterprise. Michael and I think we probably spend the last two three years talking about this experimentation and P.O.C. and what I think we want to hear now about is really what it takes to make it real at scale. And so much of what you announced today I think are the are the foundations of that. But what are you sort of advising to the enterprises to the companies right now that are looking to take A.I. and bring it to scale beyond just those P.O.C.

MICHAEL DELL: 

So there has been a lot of progress. And now we have 5,000 customers with Dell AI factory. So it's no longer kind of a test or an experiment. Definitely moving from proof of concept into production with lots of organizations. You saw Eli Lilly as an example, 1,000 plus GPUs. I think what's happened in the last couple of years is as we've moved to the reasoning models and the recursive self-improvement, The possibility of reimagining the workflows. Really drives a different kind of speed inside organizations and so I think the organizations that have figured out that they figured out that they've got to organize all their data. You know the siloed data that's really not going to cut it right and so when you understand the trajectory of. These models and what's possible and then you reimagine the workflows then you get these kind of exponential and parabolic improvements in productivity and you know. We all know sort of small examples of that. But now it's starting to happen at in in the largest companies in the world. And in fact they're figuring out that if they don't do that they've got themselves a big problem. And so there is a tremendous amount of appetite. They want to learn how to do it. How do they take all these coup de X libraries and all these models and you know our AI data platform pull other day together and actually make it happen we've got great now blueprints reference platforms tons of partnerships and also you see the frontier models are coming on Prem so lots of changes over the last couple of years from where we started.

PATRICK MOORHEAD: 

Yeah there's a lot of. I mean if nothing else if you are if you haven't started this already you're already years behind and you may not be in business years from now. But there still is a a dynamic tension out there. between, you know, hurry up, do fast, fail versus, you know, taking your time to do it in a very measured way. And I want to say a safe way. How are leaders balancing these two things of, you know, get it done yesterday, where is my ROI versus not?

MICHAEL DELL: 

You know I think one of the mistakes that we often see with this is there's sort of an instinct to just go do it right. And you want to you want to sort of grab onto the shiny object of AI. And actually, it does take some real work to do the re-architecture and understand what the workflows will really look like. And the other important thing here is, in a very large company, you want to identify the biggest processes that actually move the needle and have the most number of people, the most amount of data, and are actually central to what the company's purpose and real business is. And so once you do that and you put the right effort behind it and does take some time then you could really make a significant change. If you're just running off and trying to install something very quickly. It's not really going to have much of an impact. And so people are moving beyond that to sort of a more thoughtful understanding of how you do this. And we help them. Our partners help them. We've seen a lot of the mistakes over the last few years. And so now we have a playbook that we can share with our customers. And the good news is we're just being inundated with customers that want to learn how to do it. And we're helping them.

PATRICK MOORHEAD: 

And it's a playbook that you you're actually running internally at Delia too as opposed to hey I read about it or a I help me create this.

MICHAEL DELL: 

We are we are customer zero embracing it in every possible way. And everybody knows about it. I mean we've been talking about it for years inside this company. And I think also when you're changing a lot. It's good to over communicate. Explain why are you doing this. Why is it important? You know what happens if you don't do it. Right. And so everybody gets on board and understands that this is the future.

DANIEL NEWMAN: 

You know it's interesting. You've built this franchise that is Dell on a very broad ecosystem. You've you know from a Silicon diversity standpoint you know whether it's been laptops or infrastructure you know of course you have a great partnership with NVIDIA. But you work with AMD and you work with arm and you work with different technologies. And It's really interesting because we've seen with model diversity like one day it's anthropic. Then all of a sudden next week codex comes out and it's right. And you're seeing that basically technology itself is certainly exponential and it's certainly changing what's. What's possible but at the same time operationalizing a I you can do it with any of any very you could use a Dell server with AMD you could use a Dell server within video you could use codex you could use andropic you could use Gemini. You could do some of it on Prem you could do it off Prem correct so like as you're thinking about what's going to sort of. determine the winners of AI. It's not just the tech itself. It's something else. What is that sort of intangible that you're seeing with enterprises. Maybe Lily is an example. Maybe Honeywell these companies that you brought out on stage. What's what's in their DNA that's really you know making these companies the winners of putting AI to work.

MICHAEL DELL: 

So I think there's two points. The first, you were saying about the ecosystem. This is incredibly important because if you fast forward a couple of years, who are the leaders going to be in all the different areas you mentioned? Well, nobody's really quite sure, right? So you want to build an architecture where you can essentially use anything and have it work well. On your point about what's you know, the companies that are succeeding with this. What is different about them? I would say there is a leadership courage where they understand that this is incredibly important and it comes from the top. You know I believe organizations are not going to spontaneously prove themselves. And certainly the old silos that existed inside companies are not really the way to think about this new agent technology. So you really have to reimagine a business. And so the that requires leadership and courage. And you know some companies have that in spades and some companies don't have that. And I think ultimately you know speed is is the is the thing that matters here because the benefits are accruing to the companies that are executing. You can kind of see that in this. There's a separation in the results in companies that are figuring this out and ones that are not.

PATRICK MOORHEAD: 

So Michael what I was most what I like most about the keynote was that you brought the reality of agentic enterprise workloads on prem and the message wasn't hey don't use the public cloud right. But what it was and this was the big aha for me. I mean it was one thing when you did this with Google. Right. And you've knocked out some. I mean just it. You've enabled some amazing things even at the federal government level with them. But now we have open AI GPT 5.5.

MICHAEL DELL: 

We have reflection. We have you know obviously all the open source models and much more to come. We have Palantir. That's right. SpaceX AI. And so pretty much what's happening is all the frontier models are coming on prem.

PATRICK MOORHEAD: 

And that is a big I mean people don't stand back and say this is a huge change folks.

MICHAEL DELL: 

And actually most of the data the valuable data, the hard one you know data that differentiates a business. You would never give up to anybody else. That's that's the data that you want applied to your eye.

PATRICK MOORHEAD: 

How should the executive staff in your enterprise look at balancing flexibility, security cost and latency. Because you know we're optionality and Daniel talked about it whether it's you know whether it's on prime whether it's on edge whether it's in a colo sovereign cloud or public cloud. How should how are you recommending the balance ?

MICHAEL DELL: 

The answer will vary depending on the industry and regulations and security and how sensitive or IP is. But the reason why we have five thousand of these now versus let's say 50 a couple of years ago is because. Customers have figured out that it's really expensive to generate these tokens externally. In addition to all the regulatory security challenges. So if you can have unmetered intelligence particularly against these open weight models. Because let's face it, not all problems require the most powerful model. And you're going to have this concept of token routing. How do I take the query, the API request, and I process it with the best model for that given token? Jeff is going to actually talk a little bit about this tomorrow. And that will require a new set of capabilities, But at the level of the individual user, maybe 85%, 90% of the tokens can actually be generated right on your PC. OK. That's a big number. And I think, by the way, that number has been true. There's a Stanford study that actually shows that the number is 89%. OK. It's an even bigger number. I didn't just make that up.

PATRICK MOORHEAD: 

You don't often make up numbers, Michael. So I trust you on that.

DANIEL NEWMAN: 

But I think that part of the conversation be elevated in a big way in the next couple of years because the industry is largely subsidized token use. So so many enterprises that have signed up and this is nothing against the great work that these big frontier model labs have done. But there's so many people are using you know exponentially more tokens than they're paying for. And then when all of a sudden you deploy these agentic workloads enterprise wide on a global scale across your the business rules rails regulations governance sovereignty that you need. Plus you're doing this for thousands of employees or tens of thousands on millions potential agents. that doing it on your infrastructure on Prem with your model is going to be a very different economic situation than trying to run that all through the cloud.

MICHAEL DELL: 

And again we don't. We think it's an all the above approach. Right. Sure. Workloads right that belong in. All the places that we've talked about. Yeah.

DANIEL NEWMAN: 

I think the answer is and I think you're going to use it all with tech.

PATRICK MOORHEAD: 

It's always and just percentages.

DANIEL NEWMAN: 

I think we're going to spend a lot of effort over the next couple of years trying to figure out you know how do we do this more efficiently. Gain that 20 plus 30 trillion dollars of 20 to 30 trillion that we expect.

MICHAEL DELL: 

And then again you know you think about. the operating room in a hospital or the advanced manufacturing site where Honeywell is embedding these systems or Eli Lilly is producing all these drugs and medications. You want the intelligence at those locations where the data is being generated. So the reality is if you're. You know if you have data you're not using AI you're doing it wrong. We've talked about this in the past and you want to bring the eye is close to where the data is being generated as possible. That's where the most cost effective, most secure token is going to be generated.

DANIEL NEWMAN: 

Well Michael the future looks very bright and very exciting. I think on behalf of both Patrick and I we can say thanks so much for joining us here on the six five. And congratulations on all the success. And we look forward to the rest of Dell Technologies World 2026. Thank you guys so much. Thank you. And thank you everybody for being part of this Six Five. We are on the road here at Dell Technologies World 2026 in Las Vegas Nevada. Great event. Stick with us for more coverage. We'll be back really soon.

‍