The global artificial intelligence landscape is undergoing a significant recalibration, marked by pivotal decisions concerning chip exports and a growing investor scrutiny of AI's tangible economic returns. While the past two years have seen AI emerge as a potent catalyst for market gains, a period of reckoning is now upon us, forcing a reassessment of how swiftly this transformative technology translates into dependable earnings growth. This evolution is not solely confined to capital markets; it is also reshaping the job market in ways that defy initial widespread anxieties about mass unemployment.
A critical development this week, as Nvidia prepares to unveil its latest earnings report, is the reported decision to permit the export of advanced H200 AI chips to select Chinese clientele. This move, according to financial advisory firm deVere Group, will profoundly influence the pace and breadth of AI capability proliferation worldwide. Nigel Green, CEO of deVere Group, commented that this decision "matters for investors far beyond the chipmakers themselves," signalling a potential acceleration in the global AI race and a more direct competitive challenge to established AI leaders. For months, Chinese developers have been striving to build sophisticated AI services, often compensating for hardware limitations through algorithmic ingenuity and extensive datasets. The availability of more potent chips could significantly expedite their progress and reduce iteration costs.
This development arrives as investors pivot from a posture of unreserved optimism towards a more pragmatic evaluation of AI's financial implications. The initial exuberance that propelled market indices and tech stocks has begun to wane, replaced by a demand for concrete evidence that the substantial investments pouring into AI infrastructure are yielding measurable and sustainable profits. Divergent financial results from major technology firms have highlighted this disparity, with some struggling to convert AI-related expenditures into robust returns. This impending "reckoning," expected to define investment strategies for 2026, suggests a period where the true economic viability of AI initiatives will be rigorously tested.
Intriguingly, amidst these concerns about corporate profitability, research from Vanguard paints a different picture for the labour market. A survey examining 140 job roles deemed highly susceptible to AI automation has revealed an unexpected trend: these occupations are currently experiencing accelerated employment and wage growth. Between mid-2023 and mid-2025, jobs with high AI exposure saw employment climb by 1.7%, a notable increase from the 1% observed in pre-pandemic periods. Similarly, average real pay in these roles rose by 3.8% post-pandemic, starkly contrasting with a meagre 0.1% growth before the global health crisis.
Adam Sickling, an economist at Vanguard, noted, "At a high level, we don’t see any evidence that occupations exposed to AI are experiencing lower employment." This counterintuitive outcome is attributed, in part, to the current limitations of AI systems. Despite their advancements, AI remains prone to errors, often referred to as "hallucinations," and is not yet sufficiently reliable to fully supplant human workers. The necessity for human oversight, error correction, and strategic deployment means that roles requiring these skills are, for the time being, in higher demand and command greater remuneration. However, the long-term trajectory remains uncertain, with the potential for future AI evolution to disrupt specific job functions as systems become more capable and dependable. The coming weeks will be crucial in setting the tone for the broader economic narrative surrounding AI, determining whether its impact will primarily be one of financial scrutiny or continued job market expansion.