Tuya’s Yang Tells Davos the AI Race Lives in Real-World Apps

Tuya Smart co-founder Alex Yang spent three sessions at Summer Davos 2026 in Dalian arguing that the AI competition is shifting from foundation-model benchmarks to real-world applications in smart homes, energy grids and robots. Yang, who also serves as the company’s chief operating officer and chief financial officer, framed the next phase of the AI race as a contest over who ships AI into physical devices, not who tops the model leaderboards. His pitch landed at the 17th Annual Meeting of the New Champions, where the World Economic Forum said more than 1,800 leaders from over 90 countries and regions gathered under the theme “Innovating at Scale.” Tuya reported over 1,970,000 registered AI developers on its AI Developer Platform as of March 31, 2026, and that figure sat underneath every point Yang made.

In the session “China’s Lobster Fever,” Yang laid out three reinforcing advantages he said drive China’s rapid AI adoption: a vast developer base, broad openness to new technology, and a society-wide aspiration for development. Tuya’s own platform reflects that scale: the same release shows the company pulls in real-time electricity pricing from more than 800 utility companies across Europe to optimize energy loads for its customers. Yang said Tuya is dedicated to bringing AI into real-world industry scenarios, with the smart home and energy sectors as the first two deployment fronts. The 17th Annual Meeting of the New Champions ran from June 23 to 25 at the Dalian International Conference Center. Yang also gave interviews to CNBC, Bloomberg and CGTN during the forum, sharing views on AI commercialization, developer community development and the large-scale deployment of AI hardware.

Yang’s Real-World AI Argument in Dalian

The substance of Yang’s Dalian argument was that foundation models have stopped being the bottleneck. “Breakthroughs in foundation models and AI agents are accelerating the rise of artificial intelligence as a new frontier of global technological competition,” the Tuya release quoted him as saying. What Yang proposed next was that the next winners will be the companies wiring those capabilities into devices, meters and machines that touch daily life.

Tuya’s open-source TuyaOpen development framework and its AI Agent development platform sit at the center of that bet. The company positions itself as a neutral AIoT platform that lets device makers, commercial integrators and industry developers plug into multimodal AI without rebuilding their own stack. Yang used “China’s Lobster Fever,” “AI and Future Growth” and “Owning the Interface: Strategy in the Age of AI Agents” as his three stages, with the session titles themselves tracking the pivot from a domestic adoption story to a broader question of where AI agents are headed. AI researchers betting on physical world models over chatbots are running parallel arguments in academic and startup circles, and Yang was effectively bringing that line of reasoning to a corporate-stage audience.

Why Is China’s AI Adoption Moving So Fast?

Yang’s account of why China’s AI adoption is moving so quickly boils down to three reinforcing forces, and the order is deliberate. He puts developers first, openness second and cultural aspiration third. The list is a Tuya-flavored reading of a story other Davos speakers told in more cautious terms. The framing also lets Tuya position itself as an enabler rather than a model vendor.

• “The country has a vast number of developers who are actively exploring and building AI applications.”
• “Both individuals and businesses have demonstrated a remarkable openness for exploring and testing new technologies.”
• “More fundamentally, the strong aspiration across society for development and progress has become a powerful intrinsic driver of AI innovation.”

The World Economic Forum’s closing press release for the meeting put over a quarter of global real GDP growth in 2026 in China’s column, and framed the meeting around five questions, including how to “harness technology for outcomes in the real economy.” Bain & Company chair Orit Gadiesh told delegates at the same meeting that the companies pulling ahead are “putting their best people and resources on just two or three big bets,” giving Yang a structure to answer inside. How China scaled generative AI to 602 million users is the macro story underneath that bet, with a $98 billion compute and backbone build behind it.

Tuya is also not the only platform playing the developer-community game; it shares the stage with chip vendors, hyperscalers and other AIoT clouds. The company’s bet is that its head start in standardized AI modules and a pre-existing installed base of smart-home devices gives it separation against any of those layers encroaching. TuyaOpen, the open-source framework Yang highlighted in Dalian, sits at the center of that head-start argument. The pricing shift Yang described would also apply across the AI industry if his framing holds.

Tuya’s Three AI Bets Get Specific Names

Yang used the forum to put a label on each of Tuya’s three product pillars: AI Home, AI Energy, and AI Robot. AI Home covers the smart-home business, where Tuya says it lowers the barriers to building AI-native products through standardized AI modules and an active developer community. AI Energy spans the full energy lifecycle from generation and transmission to storage and consumption, with the platform ingesting real-time pricing data to optimize energy loads and reduce costs. AI Robot draws on TuyaOpen and the AI Agent engines to integrate multimodal AI capabilities for robotics developers.

The energy pillar is the one Yang talked about most concretely, because the data linkage is measurable. The home and robot pillars are still mostly framed as capability claims rather than shipped products at scale. The pivot from “AIoT cloud” to “AI cloud platform service provider” in the company’s own description is the connective tissue between the three, and the Dalian pitch is the public version of a strategy that has been quietly shifting for two years.

Pricing Shifts From Usage to Outcome

Yang’s pricing argument at Davos was the clearest forward-looking claim he made. He argued the AI industry is shifting toward pricing models that charge for delivered outcomes rather than for raw usage. For an AIoT platform like Tuya, outcome pricing fits naturally: the company sells energy-cost reductions and smart-home automation rather than raw model calls. For foundation-model vendors, the same shift would mean customers paying only when AI produces a verified result. Yang described the pricing shift as one the broader AI industry will move through alongside platform players like Tuya.

from traditional usage-based pricing models toward an outcome-driven value-based pricing model

Outcome pricing is not a new idea in enterprise software, but it is rare in AI, where most vendors still meter by tokens, seats or API calls. Proponents say it ties revenue to customer value and removes the buyer’s risk that they will pay for an AI they cannot use. Critics argue it makes revenue harder to forecast and gives sellers an incentive to cherry-pick customers whose outcomes are easiest to guarantee. Yang’s framing came in a session that also featured fellow panelists including Cathy Li of the World Economic Forum.

The Developer Base Does the Heavy Lifting

Yang’s prepared remarks repeatedly pointed to the same figure: Tuya’s developer count. The company’s AI Developer Platform had over 1,970,000 registered AI developers as of March 31, 2026, drawn from more than 200 countries and regions. Tuya has cited the same number across its 2026 announcements.

• Over 1,970,000 registered AI developers on the Tuya AI Developer Platform (as of Mar 31, 2026)
• More than 200 countries and regions represented on the developer platform
• More than 800 utility companies feeding real-time pricing data into the energy platform
• More than 1,800 leaders at Summer Davos 2026 from over 90 countries and regions
• 17th Annual Meeting of the New Champions ran from June 23 to 25, 2026, in Dalian

The developer count is a leading indicator; it does not translate directly into recurring revenue. Yang’s pricing argument, if adopted by the broader industry, would be the mechanism for that change. The contradiction at the heart of the pitch is that developers want predictable usage pricing while customers want outcome pricing, and Tuya is trying to sell both at once. The company is incorporated as Tuya Inc. and trades on NYSE under TUYA and on Hong Kong Stock Exchange under 2391.

A Forum Tailor-Made for Scale Stories

The venue was not incidental to Yang’s pitch. Summer Davos 2026 convened under the theme “Innovating at Scale,” with the World Economic Forum’s closing release framing the meeting around five questions, including how to “harness technology for outcomes in the real economy.” That wording aligns almost line for line with what Yang has been arguing Tuya is built to do. The forum drew more than 1,800 leaders from over 90 countries and regions, including Chinese Premier Li Qiang, six heads of government, 90 government leaders and 1,000 senior business executives.

Other co-chairs of the meeting included Robin Zeng, founder and CEO of Contemporary Amperex Technology (CATL); Orit Gadiesh of Bain & Company; Eric Tse of SBP Group; and Margery Kraus of APCO. The WEF’s own write-up of the meeting leaned on the “from experimentation to deployment” language, which mirrors Tuya’s positioning. The closing statement on the 2026 Dalian meeting lays out the full set of priorities the forum set, and the Annual Meeting of the New Champions 2026 program page carries the full session list.

The Parts Yang Didn’t Address

Yang’s prepared remarks in Dalian focused on three threads: developer count, energy data, and a pricing argument. The developer count came with a concrete figure (over 1,970,000 registered AI developers as of March 31, 2026). The energy thread came with a concrete dataset (real-time pricing from more than 800 European utility companies across Europe). The pricing argument was forward-looking, with no shipped product or disclosed revenue tied to it in the Tuya release.

The risks Yang did not name in his prepared remarks are the obvious ones: foundation-model providers shipping their own device-side agents, hyperscalers offering competing AIoT platforms, and a global smart-home market that has spent most of 2025 in a slowdown. Each of those headwinds would compress the room Tuya has to convert developer count into shipped devices. The WEF’s closing release noted that the meeting “examined how advances in AI, energy systems and natural capital are reshaping competitiveness.”

The WEF’s statement and Tuya’s pitch sit in the same frame: a competition over who ships AI into physical infrastructure. Yang’s main argument at Davos is that pricing for outcomes is replacing pricing for usage, and Tuya’s monetization is moving with it. Yang’s prepared remarks at the three Dalian sessions are now public. The next set of disclosures from Tuya will show whether the company’s numbers follow them. The summer 2026 earnings cycle will carry a public checkpoint on that question.