Why the enterprise desktop still matters in 2024. (Hint: it’s not about apps or AI PCs.)
It’s 2024, and somehow the enterprise Windows desktop isn’t dead! In fact, it will be around for a long time.
While most people agree with that premise, they think it’s because (1) enterprises still have critical Windows apps, or (2) the recent announcements about the “AI PC” or “Copilot+ PC” have reinvigorated the desktop.
Both of these are true, but neither is the reason enterprises still hang on to their Windows desktops. The real reason is the same boring reality that enterprises have understood for decades: Even in 2024, the Windows desktop is the easiest way to secure, aggregate, and deliver everything knowledge workers need to get their jobs done.
But enterprises still need Windows apps!
Despite all my amazingly brilliant and memorable quotes from 30 years of writing and speaking about enterprise tech, the thing people most often quote back to me is a throwaway line I first said in 2010:
“In the far distant post-apocalyptic dystopian future, we’ll still have cockroaches, Twinkies, and Windows apps.”
While I still believe that quote is true, (with renewed hope for the “far distant” part), in 2024, you don’t need a Windows desktop to use Windows apps. VDI & DaaS “just work”, and today you can cheaply, securely, and seamlessly deliver legacy Windows apps to any device, anywhere, with just a browser. You can buy this directly from Microsoft, Citrix, VMware (Omnissa), or any one of several next-gen VDI/DaaS providers.
In today’s world, the management and delivery of Windows apps has been wholly decoupled from the need to manage or deliver a Windows desktop. So, while the two are strongly correlated, a Windows desktop is not actually needed just because some important apps are Windows.
But what about AI & Copilot+ PCs? ?? The future!!!!
People of LinkedIn have been losing their collective minds over the recent announcements from PC makers (AI PCs) and Microsoft (Copilot+ PC). Evan Ben Thompson of Stratechery wrote Windows Returns yesterday.
My reaction: ??
My full analysis of this latest hysteria would require its own article (which maybe I’ll write, we’ll see), but the TLDR is PC makers have pounced at a thing they hope can sell new PCs which is “AI washing” to the extreme. Microsoft’s response is a similar but trademarkable version, “Copilot washing”.
The average consumer does not care about TOPS or Copilot+ laptop branding, as they do not perceive that a new laptop is what's preventing them from having an amazing AI future. I already have Copilot in my GitHub and Office apps, Google regularly real-time translates between French and English on my 2-year-old iPhone, and my 3-year-old M1 MacBook Air with 8GB RAM is able to blur my camera background in every app which uses the camera.
Frankly the only AI thing I really want is the fast-talking GPT-4o app from last week's OpenAI demos, which also seemed to run fine on existing laptops and phones.
But what about Recall? How amazing is that?? I agree. But Recall is just an app which is strikingly similar to other existing apps, like Rewind.ai, which also run fine on my existing laptop. I understand the desires from Microsoft and PC makers to create demand for new hardware, but I struggle to see how anything announced yesterday is the killer app that consumers and employees will desire. (Most of what I've been reading suggests people are not looking forward to AI-accelerated workplaces as they worry about being asked to do more with fewer people.)
Finally, I don't believe the “AI runs on your laptop” model is right for the future anyway. The AI / Copilot+ PC marketing hype focuses on the importance of privacy and keeping things local, but again, these are not concerns that average people have, as Google, TikTok, and countless other apps have repeatedly shown that people will happily and continuously trade their privacy for a steady dopamine drip.
This is just another, “When all you have is a hammer, everything looks like a nail.”
So why do people care about the enterprise desktop in 2024?
If neither Windows apps nor the AI/Copilot+ PC are the real drivers of the enterprise desktop, then what is?
Put simply, for the past 30 years, the Windows desktop has been the aggregation point for everything an IT department needs to provide to the user, including:
Sure, that was true 30 years ago, but is it still true today?
Yes!
There are simply no alternatives today. The most common "desktop replacement" device, an 12” iPad Pro with keyboard, trackpad, and pen, can’t get you there. (Have you actually tried to replace your laptop with an iPad? I don’t mean leaving your laptop at home for a long weekend—I mean for real for real. It’s “inefficiency by a thousand cuts”!)
Real money is being poured into the enterprise desktop. (And I specifically mean the "enterprise desktop", not the AI desktop.) Look at KKR’s $3.8B purchase of the old VMware EUC business, Palo Alto recently acquiring Talon Security for $635M, or enterprise browser company called Island closing a $175M D round at a $3 BILLION valuation. (This is on top of Microsoft's obvious investments in Windows Enterprise and the whole ecosystem there.) Serious money believes the Windows desktops is not going anywhere anytime soon. It's simply too important, and too ingrained in the ways companies manage how people work.
What happened to the disaggregation of the desktop?
Over the years, I've talked about how what we call the “desktop” is not a monolithic brick, but rather a collection of the individual components mentioned above which could be broken up and delivered via different technologies and systems:
We addressed this quite a bit in our 2012 book, The VDI Delusion, especially in the final chapter called The Future of the Desktop.
Much of this has come true over time. In 2024, we use a mix of devices, each with their own unique hardware, UI, app launchers, and methods for app integration. We use lots of different apps—web, desktop, mobile—each with their own runtime and delivery frameworks. And many of the “user” settings are truly targeted at the human instead of the desktop (security containers, provisioning targets, and configuration).
But the desktop has not actually gone away. Even though it's also no longer a monolithic brick, it’s still the central hub of what what organizations provide to their employees.
Even the EUC consulting we do at ILKI, and what we discuss in the digital workplace / future of work masterclass, the focus is the desktop. Sure, much of that talk is how we deliver it, modern VDI & DaaS, modern UEM-based management, modern provisioning, Entra ID replacing on-prem AD, and configuration profiles replacing GPOs. The desktop remains a core part of the end-user device ecosystem (including mobile, tablets, watches, and smart devices) which work together to deliver the right app, in the right way, secured via conditional access.
Yes, AI will have a big impact—even on the desktop—but it will be in the context of the larger ecosystem of digital workplace technologies, not chained to the desktop itself. So while 2024 might again be christened “The Year of the Desktop”, it's not because of AI, and not because of Windows apps, but because the desktop's 30-year history of being the desktop will continue.
Program note: We discuss and analyze all of this, and more, in my digital workplace master class which I'm teaching in several cities in Europe this year. The first one was in The Netherlands last month. Upcoming dates include:
Technology Executive | Veteran | Champion in integrating technology and business initiatives while transforming IT solutions to solve real-world business problems
6 个月I was in the room when you first presented the "cockroaches, Twinkies, and Windows app" line. I still repeat that, with surprising regularity, today.
Sales executive | IT Performance Optimization Strategist | Driving Enterprise Digital Excellence
6 个月I agree that AI everything is over-hyped. At this stage, it feels like another thing that needs managing. However, I'm interested in seeing whether this approach can help run agents at the Enterprise level and optimize AI spending. I think the larger implementation right now is more of an AI device with, hopefully, a better outlook than some well-known but unsuccessful ones like Rabbit and the Pin. I also see it becoming valuable to collect data within silos. While models are getting cheaper, they’re far from cheap. Understanding when and how to use these could be very valuable; like a Turbo button from a few years ago. Additionally, back to the collecting data part, there's a significant opportunity to optimize processes and narrow the bell curve.
Architecture Leadership| Strategy & Roadmap| Technology Evangelist |Digital & Technology Services Transformation Leader| #Future of Work Advisor| Talks about # Digital Architecture, # Strategy, #IT Transformation
6 个月I agree with you Brian Madden.