Exclusive Interview with Lenovo’s Intelligent Devices Group President Luca Rossi: The Greatest AI Application Has Yet to Be Invented

Over the past year, large model technology has been advancing at a breakneck pace, evolving almost quarter by quarter. From text-to-image generation to multimodal capabilities, and now AI Agents, functionality has become increasingly sophisticated while model parameters continue to scale upward. Yet, the faster the technology soars, the more apparent the lag in end-user hardware becomes — especially in the PC sector, a category once considered the bedrock of technological infrastructure, now facing the urgent need for reinvention.

The term AI PC has emerged as both an industry buzzword and a complex challenge. For users, it sounds like the promise of the next-generation productivity tool. For manufacturers, it signals a fundamental overhaul of system architecture, a rethinking of compute bottlenecks, and a transformation that starts at the chip level.

At the IFA trade show, Luca Rossi, Executive Vice President of Lenovo Group and President of its Intelligent Devices Group, sat down with ifanr and other media outlets for an in-depth discussion. For him, the more important question is not “What can AI do?” but rather “What should the PC become?”

AI PC: A Timely but Not Singular Solution

“Last year, we predicted the PC market would return to growth in 2024 — and that forecast has largely been proven right,” Rossi began, opening the interview with data that validates his earlier outlook.

In recent quarters, the global PC market has swung back into positive territory, posting year-over-year growth between 4% and 7%. Lenovo has consistently outpaced the market by 4 to 5 percentage points, with the most recent quarter seeing shipments climb by over 10%.

While AI PCs have contributed to this surge, Rossi is quick to clarify they are not the sole driver. The ongoing shift from Windows 10 to Windows 11 is another major factor, especially with Microsoft’s support for Windows 10 ending in October 2025. Only about half of enterprise systems have upgraded so far, meaning the refresh wave will likely continue for another two to three quarters.

Currently, roughly 30% of Lenovo’s new-generation devices feature AI PC capabilities — a figure that is still climbing. Rossi believes hardware readiness is already well in place; the next big leap will come from software and ecosystem innovation. “The real explosion will happen next year,” he predicted.

Understanding AI Isn’t Always Necessary for Users

When it comes to the consumer market, Rossi offers a pragmatic view: most users today don’t fully understand what value AI PCs bring — but that doesn’t stop them from buying them.

Early adopters who deliberately choose AI PCs for AI functionality remain a niche of professional users. For the majority, purchasing decisions are still driven by more familiar priorities — sleek, lightweight designs, long battery life, and future-proof expandability.

“Our AI PCs are compelling even without the AI label,” Rossi noted. “With 12-hour battery life, slim profiles, and premium design, the appeal is already there. AI reveals its value gradually through use, not a single presentation.”

He envisions AI PC adoption unfolding in three stages: first, strong foundational hardware experiences; second, the rollout of early AI-driven scenarios; and finally, an explosion of ecosystem applications. Globally, AI PCs now account for 20–25% of the market, with Lenovo already at 30%. He expects that figure to hit 50% within 18 months, and 70–80% within three years. In four to five years, virtually every PC will be an AI PC.

Hybrid AI: The Future of Computing

The real debate around AI PCs isn’t about hardware — it’s about their perceived value. One long-standing question is whether on-device AI truly matters.

Rossi’s position is unequivocal: “We firmly believe the future will be hybrid AI.”

He describes a three-tiered computing landscape: cloud, edge, and device-level AI — each with its own critical role.

Currently, 80 to 100 independent software vendors (ISVs) are migrating their core applications to run on CPUs with NPUs, not as a technical showcase but to address real needs: lower latency, enhanced privacy, and reduced cloud compute costs.

For Lenovo, this is why going “all-in” makes sense. AI doesn’t have to mean cloud inference — in scenarios where generative models are tightly tied to user context, local processing can be the superior choice.

In China, Lenovo’s “Xiaotian” assistant has recently evolved into a Super Agent — poised to appear across laptops, smartphones, tablets, and eventually even smartwatches. This Super Agent is not just an interface experiment but a pillar of Lenovo’s hybrid AI vision, enabling seamless context-sharing across devices — something cloud alone cannot achieve.

Form Factor Innovation: AI as a Driver, Not a Gimmick

When asked whether AI PCs will spark a revolution in device design, Rossi is optimistic: “It’s not a barrier — it’s an opportunity.”

He points out that laptops have looked and functioned largely the same for two decades. But AI — particularly advances in natural language processing, large models, and multimodal reasoning — opens the door to breaking those conventions.

Voice-only devices without keyboards are no longer science fiction, though Rossi stresses they won’t arrive in the next quarter or even next year. The real challenge is not feasibility but reliability, and AI is now bringing that goal into reach.

Lenovo has been experimenting with new form factors for years — foldable screens, modular PCs, gaming handhelds like the Legion Go — with some concepts going commercial and others laying the groundwork for future breakthroughs. “Failed experiments pave the way for successful ones,” Rossi remarked, citing the company’s 2017 foldable PC concept, which eventually informed the Moto Razr’s hit relaunch.

Will AI Create Entirely New Device Categories?

Rossi believes the answer is yes. Within the next five years, AI’s influence on how devices are used could lead to entirely new product categories across laptops, tablets, and smartphones.

One area he’s particularly intrigued by is smart glasses: “I see glasses as one of the promising future platforms, with varying degrees of success possible within the next 2–5 years.”

Adoption, not just technology, will be the challenge. Glasses already enjoy mass familiarity, unlike many other novel form factors. Still, three technical hurdles remain: compute power, battery life, and lens design.

He praises Meta and Ray-Ban’s smart glasses as a notable step forward: “They’re not a runaway hit, but more successful than most examples I know.” More competition and innovation, he argues, will only benefit the market — and while Lenovo isn’t in the space yet, Rossi hints: “We’ll join when the timing is right.”

The Best AI Application Has Yet to Be Invented

As the conversation wrapped up, Rossi returned to the big question: what will be the “killer app” for AI PCs?

“The best application hasn’t been invented yet,” he said.

In the commercial sector, AI’s surest value will be in boosting efficiency; in the consumer market, he sees Super Agents as having the potential to redefine interaction and unlock new value.

Just as it took years for the App Store to build its ecosystem, AI PCs are still in their infancy. Two years ago, almost no developer knew about NPUs; today, hundreds of ISVs are engaged.

“We never believed AI PCs could replace cloud-based ChatGPT,” Rossi concluded. “But we believe they can offer a lighter, more private, and more instantaneous intelligent entry point in each user’s unique context.”

For Lenovo, this is not about extending the legacy of PCs as yesterday’s productivity workhorses — but about positioning them as the forward base of the AI era.