The search for value in the AI era: Insights from Evan Zehnal- July 29, 2026

AI is being tested by real-world economics

Much of the current debate around AI centres on economics rather than technology. According to Evan, the most important question is whether the enormous investment flowing into AI infrastructure will generate sustainable returns. The technology itself has demonstrated impressive capabilities, but the long-term success of the build-out will depend on whether businesses can earn attractive economics from it. There are reasonable arguments on both sides. The optimistic case is that AI is beginning to unlock valuable new applications that organizations are willing to pay for. The more cautious case is that the benefits may not ultimately accrue to the companies financing today's infrastructure expansion.

Coding has emerged as a powerful early use case

Conversations with both private and public companies helped build conviction that AI-powered coding tools were becoming a practical and increasingly valuable application. He pointed to coding as one of the clearest examples of AI moving beyond experimentation and into real-world use. Historically, software development has often been constrained by a shortage of developers rather than a lack of demand for software. AI coding tools may help address that constraint by making developers more productive and allowing more work to be completed in less time. He also noted that coding agents can operate continuously, potentially increasing productivity and expanding what organizations can accomplish with existing resources. Given the scale of global spending on software and services, Evan described coding as a potentially monetizable use case that could have meaningful economic implications.

The biggest winners have yet to emerge

One of the central questions surrounding AI is where the economic value will ultimately accrue. AI has the potential to create substantial value, but it remains unclear which participants in the ecosystem will capture the greatest share of the benefits. Questions remain around how value will be distributed among model developers, cloud providers, semiconductor manufacturers, software companies and the organizations deploying AI solutions. The optimistic case is that AI is beginning to unlock valuable applications that can drive meaningful productivity gains and support additional spending. The more cautious case is that the economics may not accrue to the organizations funding the build-out, particularly if competition intensifies or alternative solutions become sufficiently effective. For now, the jury is still out. AI-related revenues have only recently begun to accelerate, making it difficult to draw firm conclusions about which business models will ultimately prove most successful.

China is emerging as a serious AI competitor

Competition is also intensifying globally. Recent advances from Chinese AI developers suggest that China may be closer to the leading edge of AI development than many had previously assumed. In his view, recent developments suggest the gap between leading Chinese models and U.S. frontier models may be measured in weeks rather than months. While discussion often centres on copying or model distillation, he argued that those explanations alone do not fully account for the progress being made. China has significant engineering talent, access to technical resources and the ability to continue advancing AI capabilities. The result is a more competitive global landscape, with Chinese firms increasingly participating in AI innovation rather than simply following developments elsewhere.

AI may transform tasks more than jobs

Questions about employment remain a major part of the AI conversation. Rather than focusing solely on whether jobs will disappear, he believes it is important to distinguish between jobs and the individual tasks that make up those roles. AI may help reduce repetitive work, allowing people to spend more time on higher-value activities. He pointed to radiology as an example. In his view, advances in AI-supported tools have helped improve productivity rather than eliminating the need for radiologists. The technology assists with specific tasks while allowing professionals to focus more of their time on the aspects of their work that require expertise and judgment. While some organizations have cited AI-related efficiencies when reducing headcount, Evan also suggested that some workforce reductions may reflect earlier periods of over-hiring. As a result, he does not view current trends as clear evidence that AI is broadly replacing workers across industries.

Conclusion: What comes next for AI