GHOST IN THE MACHINE | WAIC 2026: The next phase of artificial intelligence

Technology conferences rarely tell us where technology is going. They tell us what governments and corporations want investors, policymakers and the public to believe about the future. For decades, trade shows have functioned as carefully managed performances in which optimism becomes a marketing strategy and product demonstrations become proxies for technological progress. The World Artificial Intelligence Conference (WAIC) 2026 in Shanghai certainly fits that description. With more than a thousand exhibitors, hundreds of forums and thousands of AI products on display, it projects confidence at a time when confidence itself has become an important geopolitical commodity.

Yet dismissing WAIC as another exercise in state-backed promotion would be as misleading as accepting every announcement at face value. The conference matters because it reveals how artificial intelligence is gradually being repositioned. AI is no longer presented primarily as an emerging technology or a collection of impressive algorithms. It is increasingly framed as industrial infrastructure, comparable to electricity, telecommunications or railways. That shift may prove more consequential than any new language model unveiled during the event because infrastructure, unlike products, reshapes entire economies.

The transformation is occurring against a backdrop of slowing enthusiasm for generative AI. Only two years ago, nearly every technology company sought to convince investors that large language models represented the beginning of a new industrial revolution. Product launches were accompanied by extravagant claims about replacing knowledge workers, eliminating software development bottlenecks and creating fully autonomous digital assistants. Venture capital rewarded companies that merely attached “AI-powered” to investor presentations. The assumption was that computational capability alone would produce economic transformation.

Reality has imposed constraints that marketing campaigns rarely acknowledge. Building reliable AI systems has proven substantially more expensive than demonstrating them. Enterprises have discovered that integrating foundation models into existing operations requires high-quality proprietary data, new cybersecurity frameworks, continuous monitoring, legal review, infrastructure upgrades and workforce adaptation. The computational cost of training increasingly sophisticated models continues to rise while commercial returns remain uneven. Many organizations experimenting with generative AI have reached the familiar conclusion that technological capability and economic viability are not synonymous.

This context explains why the structure of WAIC has changed. The conference still highlights foundation models, but the emphasis has shifted toward computing infrastructure, industrial deployment, robotics, scientific research, intelligent manufacturing and AI governance. These subjects attract fewer headlines than chatbot demonstrations because they lack theatrical appeal. They are nevertheless where artificial intelligence will either justify its economic promise or become another overcapitalized technology cycle.

Much of the public conversation about AI continues to revolve around software while neglecting the physical systems upon which software depends. Artificial intelligence exists because of electricity, data centers, cooling systems, fiber networks, semiconductor fabrication plants, packaging facilities, advanced memory technologies and increasingly complex supply chains that span multiple continents. Every discussion of algorithms eventually arrives at questions of energy consumption, manufacturing capacity and geopolitical access to advanced chips. The software industry often portrays AI as an abstract digital phenomenon, but its expansion depends on extraordinarily tangible infrastructure.

WAIC therefore serves another purpose. It demonstrates China’s determination to reduce external dependence across that infrastructure. Export restrictions imposed by the United States on advanced AI accelerators have accelerated investment in domestic semiconductor development, computing platforms and software optimization. Whether these efforts ultimately produce hardware capable of competing directly with leading American products remains uncertain, but the strategic objective is unmistakable. Artificial intelligence has become inseparable from industrial sovereignty.

This distinction frequently disappears beneath geopolitical rhetoric. Coverage of AI increasingly resembles coverage of international sporting competitions, where advances by one country are interpreted primarily as setbacks for another. Every benchmark becomes evidence of technological supremacy. Every product announcement is framed as proof that one nation is either overtaking or falling behind its rivals. Such narratives are politically convenient because they simplify complex technological ecosystems into familiar stories of competition.

The difficulty with this framing is that AI leadership cannot be measured by benchmark scores alone. A country may develop excellent foundation models while remaining dependent on foreign semiconductor equipment. Another may dominate cloud infrastructure but lag in scientific applications. A third may excel in robotics yet struggle with software ecosystems. Artificial intelligence consists of multiple interdependent industries rather than a single technological race with one winner.

China appears increasingly aware of this complexity. WAIC reflects an approach that integrates academic research, industrial policy, manufacturing, cloud infrastructure, urban development and public administration into a coordinated framework. This does not imply that the model is universally applicable or immune from criticism. State-led industrial strategies create their own inefficiencies, distortions and political risks. Nevertheless, they differ fundamentally from the venture capital model that has largely shaped AI development in Silicon Valley.

The American approach has historically relied on private investment, entrepreneurial experimentation and market competition. This ecosystem produced remarkable innovation because capital flowed rapidly toward promising technologies. Its weakness, however, is a tendency to equate market valuation with technological significance. Artificial intelligence became another financial narrative long before many of its commercial applications matured. Companies were rewarded for announcing AI strategies regardless of whether those strategies generated sustainable productivity gains.

China’s model carries different incentives. Artificial intelligence is treated less as an investment category than as a strategic capability intended to strengthen manufacturing, logistics, healthcare, education and scientific research. Commercial profitability remains important, but it is embedded within broader national objectives concerning industrial modernization and technological self-sufficiency. The result is an AI ecosystem whose priorities often diverge from those of Silicon Valley despite relying upon many of the same underlying technologies.

Neither model fully addresses the broader social consequences of AI deployment. Automation continues to reshape labor markets faster than educational institutions can respond. Copyright disputes remain unresolved. Algorithmic bias persists despite technical improvements. Deepfakes have expanded the scale of misinformation while simultaneously undermining public confidence in authentic media. Energy demand from large-scale computing continues to increase at precisely the moment many governments are attempting to reduce carbon emissions. None of these problems can be solved simply by developing larger models or faster processors.

This is where WAIC’s emphasis on governance deserves closer examination. Discussions concerning AI regulation are often portrayed as obstacles to innovation, particularly within technology industries that have long preferred minimal oversight. That perspective increasingly ignores the historical relationship between infrastructure and regulation. Railways, aviation, telecommunications, financial systems and pharmaceuticals all required governance frameworks because their failures affected society at scale. Artificial intelligence is moving toward a similar position. Questions concerning transparency, liability, auditing, interoperability and public accountability are becoming structural requirements rather than optional ethical considerations.

For Southeast Asia, including the Philippines, WAIC offers lessons that extend beyond China itself. Public discussions about AI within developing economies frequently emphasize adoption rather than capability. Governments seek foreign investment, businesses experiment with chatbots and educational institutions introduce AI literacy programs. These initiatives have value, but they remain concentrated at the application layer. Far less attention is devoted to semiconductor research, advanced manufacturing, computing infrastructure, scientific computing or energy systems that ultimately determine technological resilience. Countries that consume AI services without participating meaningfully in the infrastructure supporting them risk reinforcing existing patterns of technological dependence.

This does not mean every nation should attempt to replicate China’s industrial strategy. Political institutions, economic structures and resource constraints differ considerably across countries. It does suggest, however, that AI policy should be understood as industrial policy rather than software policy alone. Decisions regarding electrical grids, fiber connectivity, research funding, technical education, cybersecurity, data governance and manufacturing capacity increasingly shape a country’s long-term position within the global AI economy.

Technology journalism often gravitates toward the visible because visibility produces engagement. Robots attract cameras. Chatbots produce demonstrations. Product launches generate headlines. Infrastructure evolves more slowly and usually without spectacle. Yet history repeatedly shows that durable technological change occurs when infrastructure matures, not when prototypes impress audiences. The internet became transformative only after broadband, data centers and global standards emerged. Smartphones changed society because telecommunications networks, semiconductor manufacturing and software ecosystems developed together. Artificial intelligence is unlikely to follow a different trajectory.

Viewed from that perspective, WAIC 2026 is less significant because of the products unveiled in Shanghai than because of the assumptions underlying the conference itself. The event reflects a growing recognition that the future of AI will depend less on increasingly dramatic demonstrations than on the slower and less glamorous work of building industrial capacity, computational infrastructure, scientific institutions and governance frameworks capable of sustaining technological development over decades rather than investment cycles.

That may also explain why the AI conversation has become noticeably quieter compared with the exuberance that accompanied the release of the first generation of widely accessible foundation models. The technology has not become less important. Instead, it has begun the transition that every consequential technology eventually undergoes. Artificial intelligence is moving from novelty to infrastructure, from product launches to deployment, from speculative investment to economic integration. That process is inherently less dramatic than the narratives that fueled the initial wave of enthusiasm, but it is far more likely to determine who benefits from AI and under what conditions.

The real significance of WAIC 2026, therefore, is not that it offers another opportunity to compare benchmark scores or celebrate increasingly sophisticated machines. It demonstrates that artificial intelligence is entering the stage at which political economy, industrial capacity and institutional design may matter more than algorithmic breakthroughs alone. History suggests that this is precisely the point at which technologies stop being curiosities and begin reshaping the societies that build around them.

Full disclosure: Ghost in the Machine by Kusanagi Motoko is written entirely by artificial intelligence (AI).

————————————————————————-
TEN YEARS OF TECHSABADO!

PLEASE LIKE our FACEBOOK PAGE and SUBSCRIBE to OUR YOUTUBE & SPOTIFY channels.

WATCH TECHSABADO ON OUR YOUTUBE CHANNEL:

WE ARE ALSO ON SPOTIFY

WATCH OUR OTHER YOUTUBE CHANNELS:

PLEASE LIKE our FACEBOOK PAGE and SUBSCRIBE to OUR YOUTUBE CHANNEL.

by Kusanagi Motoko