“You have to be careful where you position it, because it’s not for everybody,” Compucom VP Mike Monahan said.

Although the artificial intelligence PC revolution is still booting up, solution providers are already planning how to work with customers to determine the right time to buy, with early selling points for customers focusing on the devices’ improved performance and improved security and the old-fashioned fear of missing out if a competitor adopts AI PCs first.

“There’s going to be plenty of FOMO,” Dawn Sizer, CEO of Mechanicsburg, Pa.-based 3rd Element Consulting – a member of CRN’s 2024 MSP 500 – told CRN in an interview.

But in some cases the solution provider will have to help determine whether a customer can buy an AI PC depending on industry regulation or whether certain types of employees are better off having AI PCs than others to keep costs down and maintain better control over the data AI accesses.

“There’s going to be plenty of the good cop, bad cop factor,” Sizer said. “We want to make sure that we get the right tools in front of the right people at the time that they need to use them. You need a tool for something. I don’t want to give you a wrench when you really need a hammer.”

Keeping customers’ AI expectations in line, making sure they follow regulatory requirements, setting up controls to allow or block AI access, and deciding who within the customer organization gets an AI PC are all areas of opportunity for partners even after AI PCs hit peak bandwidth, multiple solution providers told CRN.

[RELATED: Solution Providers Are Helping Customers Navigate Complex AI PC Landscape]

AI PCs

The AI PC revolution is coming fast. Mike Monahan, vice president and chief technology officer of advanced technology at Fort Mill, S.C.-based Compucom, No. 59 on CRN’s 2024 Solution Provider 500, said he has seen data showing that more than 8 million AI PCs shipped in the second quarter, with 40 million projected for all of 2024.

That number should reach 100 million next year, representing 28 percent of the PC market, and hit 60 percent of the market come 2027, Monahan said.

Monahan said content creators and video collaboration application users are two job types that will make sense for early AI PC adoption.

“Not everybody’s going to have one, but you’ll see a lot of people starting to shift that way,” Monahan said. “Like anything in IT, if you build it, they’ll come. … You have to be careful where you position it, because it’s not for everybody, and trying to use it for the right use cases is what matters.”

Data Governance A Focus

Wayne Roye, CEO of New York-based solution provider Troinet, told CRN that in the AI PC era, one of his focuses when talking to customers will be data governance – not just AI PC security benefits.

For Roye, data governance reflects a more proactive approach, addressing “what piece of information can AI possibly touch and see that I [the customer] may not be aware of,” he said.

Data governance gets at the root of where AI can go wrong and readies a customer for putting controls around decades old data that AI suddenly might have access to. Plus, it excites customers for AI capability.

“If you come in and talk security, there’s a fear factor,” he said. “But you are talking AI, it is an operational thing. I need to improve something. I need to create content. My salespeople can enable quicker. … My people that are doing analytics can actually get data quicker. … I see the value of it compared to seeing a security play – do these things because you have to.”

Roye said he foresees a lot of companies wanting to create and train their own language models. “Some of the hardware you see going out is going to allow people to build their own AI for themselves,” he said. “And that’s where you’re going to see the drive for AI PCs.”

Dennis Perpetua, global chief technology officer for digital workplace services and experience officer, vice president and distinguished engineer at New York-based Kyndryl, No. 9 on CRN’s 2024 Solution Provider 500, told CRN that he will talk with customers about the “radical benefits” that come with moving GenAI to the edge with AI PCs.

AI PCs will speed up adoption for regulated customers concerned about AI access to sensitive personal information (SPI) and customers concerned about the effect of AI on energy costs. AI-enabling small language models have been run off computers costing less than $100, as an example of their flexibility.

“We’re now just going to get this capability and it’s really not going to cost me anything more from a home-user perspective,” he said. “This is going to put AI in everybody’s home. … This is going to be fundamentally the next chapter in the GenAI story.”

Perpetua predicts a growth adoption curve for AI PCs that he has seen with AI applications – customers start with a small piloting group before ramping up to more users.

Customers who have already been dazzled by AI applications will be good targets for AI PCs, he said. “If OpenAI was your first foray into AI … you’re still blown away” by AI PCs, he said. “You’re probably even more blown away than folks that hadn’t been using it before.”

Perpetua will talk to customers about how personalized he believes AI PCs will get for users based on the work they are doing, whether that work is best handled by the machine or through a cloud-based service, for example, ultimately improving digital experiences for workers and customers.

“There is going to be more of a personality, if you will, associated with how the AI is constructed on each individual’s devices over time,” he said.

A key element of winning customers over will be mixing in the harder to quantify “soft” benefits of AI as well as the more easily measurable “hard” benefits, he said. Perpetua described the soft benefits he’s already seen leveraging AI in meetings.

“It’s almost like having an audience listen to things and allow these ideas to bubble up and rise to the top and become the dominant ideas in the meeting,” he said. “That, to me, is a soft ROI. It’s a productivity enhancement. It’s a quality enhancement. Those are the types of things that I’m super excited about seeing because it just allows you to actually focus on the task.”

Kyndryl’s approach to customer conversations around AI has focused on improved experience for workers and customers and productivity recommendations to employees using digital employee experience (DEX) tools, for example, he said.

“When I call the service desk, not only do they know who I am, what device I have … but they can also proactively with automation before I even speak with somebody, push some scripts to my device,” he said.

AI PCs will improve on-device diagnostics and device management as well, he said. “That can drive the refresh in a way that articulates business value associated with it.”

‘A Perfect Storm’ Of Factors

Some customers will simply be won over by an AI PC’s improvements over older devices, even without factoring in some GenAI applications, multiple solution providers told CRN.

When it comes to security, their neural processing units (NPUs) can reduce latency and cost and take threat detection, anti-phishing and deep fake detection to another level, Monahan said. The devices will also handle continuous tasks in the background with less power draw, making for “tremendous” battery life.

“They’re utilizing the power of the PC to do it locally,” Monahan said. “Better performance, better battery life, lower latency and somewhat better cost because we don’t have to do that on a more expensive machine.”

Solution providers have “a perfect storm” of improved hardware, the age-out of machines bought during or even before the height of COVID-19 and end-of-support for Microsoft’s widely used Windows 10 operating system planned for October 2025.

Those factors might be enough to motivate customers to pull the trigger on AI PCs, Monahan said.

3rd Element’s Sizer said that pharmaceuticals, finance, and research and development (R&D) are fields likely to adopt AI PCs early. Users of data analytics and customer relationship management software such as Microsoft PowerBI and Dynamics might also prove good candidates for early AI PC adoption.

Important to remember is the role of solution providers to sometimes play that bad cop and be the voice of reason, she said – not every employee will need the power and expense of an AI PC.

“There are plenty of people out there saying, if I just had a flip phone–cause all I need to do is make telephone calls,” Sizer said. “There are still going to be people that are on that. I just need a PC to do a specific thing and it doesn’t need to be AI powered.”