The smartphone-centered world is one Qualcomm helped build. Cristiano Amon is now saying that era is starting to wind down.
In a May 10 interview with Fortune, Qualcomm's CEO said the smartphone's role as the primary device could begin to give way around 2027 to 2028. The statement matters because Qualcomm is not a bystander: roughly 5 billion devices run on its chips, and phones still accounted for more than 70% of its revenue in the company's early-year financials.
Amon has put numbers behind the shift. By 2029, Qualcomm wants its non-phone businesses to be roughly as large as its handset business, with non-phone revenue reaching $22 billion. It reads less like a prediction than a transition plan written by the company that profited most from the last computing cycle.
From phone-first to an "ecosystem of you"
Amon calls the next device pattern an "ecosystem of you": glasses that see what you see, earbuds that hear what you hear, and an AI agent that links those signals across devices and acts on them.
In that model, the phone does not vanish. It loses rank. Instead of being the center of computing, it becomes one sensor among others, alongside glasses, earbuds, rings and other interfaces to the physical world.
Amon's hardware bet is clear: glasses win. They sit closest to the eyes and ears, the two senses that matter most for this kind of ambient computing. Jewelry, pins and pendants may exist, but he treats them as supporting roles.
That logic lines up with Meta's Ray-Ban effort, Google's renewed interest in glasses and Apple's rumored next Vision hardware. The difference is Qualcomm's position. Consumer brands want the mindshare; Qualcomm wants the silicon layer underneath whoever wins.
Amon expects wearables to grow from tens of millions of units today to hundreds of millions, possibly even 1 billion units, within five years. If that curve materializes, Qualcomm can try to repeat its smartphone-era playbook: avoid owning the consumer's face while owning a critical part of the supply chain.
2026 as the year of agents
Amon has repeatedly described 2026 as the year agents go mainstream, including in earlier remarks to Semafor and again in Fortune. His definition is not simply a chatbot that answers questions. He is talking about software that can perform small, low-friction tasks in physical life.
His examples are practical: an agent that sees a restaurant bill and pays it, compares prices while you look at a product, or calls a doctor to reschedule when a calendar conflict appears.
The hardware premise is important. The camera has to see what the user is seeing. A phone can do that only after being pulled out, aimed and tapped, which breaks the low-friction promise. Glasses are built for passive, continuous perception.
That is why Amon links agents and glasses so tightly. Agents need a new hardware surface. Phones were designed for apps that users actively open; glasses are designed for context that is sensed continuously.
6G is the hidden assumption
Amon also spent time on 6G, calling it the largest generational shift in wireless history. The reason is straightforward: cameras on glasses would stream visual context to cloud agents, which would process it and send instructions back. That loop needs bandwidth and latency characteristics that 5G may not support well enough.
There is another implication he did not spell out directly. Continuous visual uploads could help create a constantly updated digital twin of the physical world in the cloud. City-scale spatial maps built from live video streams would matter far beyond consumer apps, including autonomous driving, robotics and urban management.
Qualcomm happens to operate across wireless chips, AI chips and automotive chips. For Amon's story to work, all three lines have to mature at the same time. That helps explain why the company has expanded into cars, PCs, industrial robotics and data-center chips since he took over five years ago.
The China detail
One detail in the interview stands out: Amon pointed to ByteDance having released a smartphone with an OpenClaw-style agent.
That example is worth noting because, when describing the future of "agent plus device," he instinctively reached for a Chinese vendor rather than a U.S. one. It suggests he sees AI-agent adoption in China's smartphone market moving faster than in North America.
Chinese phone makers have more operating-system control, faster hardware iteration and fewer platform barriers for deeply integrated agent features than the iOS and Android camps face elsewhere. Amon's mention effectively acknowledges that observation in public.
What Amon is betting on
Qualcomm's strategy can be reduced to one sentence: make AI agents and the hardware they run on into the next operating-system layer, and sell the silicon beneath that layer.
If the bet works, Qualcomm moves from smartphone-chip leader to a broader foundation for on-device and edge AI computing. If it fails, because agents do not take off, glasses do not scale or 6G slips, the company remains tied to a slowing smartphone market while Apple and Google keep pushing their own chips.
Amon has three decades inside Qualcomm behind him, but his five-year tenure as CEO is already halfway through its test. He has set the 2027-2028 window himself. By then, the "ecosystem of you" will either look like the next platform shift, or the end of the smartphone era will be slower and more expensive than he expects.
Sources: CocoLoop; Qualcomm CEO Cristiano Amon says 2026 is the year AI agents go mainstream—and the smartphone's reign as your primary device is ending (Fortune); Qualcomm CEO is working with 'pretty much all' major AI players on top-secret devices (Fortune)