The artificial intelligence (AI) sector is experiencing an investment boom, yet experts caution that enthusiasm is often outpacing actual advancements in research. As we approach 2026, significant funding is flowing into AI ventures, but many investors are operating based on outdated understandings of the technology. This disconnect is raising concerns about potential overvaluation and unrecognized opportunities within the industry.
Jenny Xiao, a former researcher at OpenAI and now the leader of Leonis Capital, has become a prominent voice in highlighting these issues. In a recent interview, she described a “years-long lag” in the AI hype cycle, where investor excitement frequently does not align with the current capabilities of AI technologies. Xiao, who founded her firm in 2021 after completing her PhD at Columbia University, emphasizes that while leading AI labs innovate in areas such as multimodal models and autonomous agents, the investment community is still focused on concepts that were groundbreaking three to five years ago.
Xiao’s insights stem from her experience at OpenAI, where she contributed to foundational models. She now invests in startups that bridge the gap between cutting-edge research and commercial viability. Leonis Capital emphasizes investments in “frontier AI” ventures, advocating for investors who can critically assess technologies beyond the initial hype.
The Investment Landscape and Its Challenges
The AI investment landscape has seen remarkable growth, with global spending on AI infrastructure projected to exceed $500 billion in 2026. Despite this surge, experts like Xiao point out that much of this funding reflects excitement over technologies that are becoming outdated. For example, large language models (LLMs), which were at the forefront of discussions in 2023 and 2024, are now recognized by researchers as limited tools. As diminishing returns set in, many investors continue to channel resources into LLM-centric startups, often neglecting emerging innovations such as agentic AI systems capable of executing complex tasks autonomously.
Industry observers have noted this trend on social media, with some predicting that 2026 could be a pivotal year for agentic AI, potentially seeing up to 40% of enterprise applications adopting these technologies. Xiao’s call for more technically astute investors resonates as current funding often favors popular pitches over solid technical validation.
A recent report from Capgemini highlights a shift in organizational priorities, moving from hype to practical applications of AI. Companies are increasingly focusing on infrastructure development and workforce upskilling to maximize the long-term value of their AI investments. This trend mirrors patterns seen in previous tech revolutions, from the dot-com boom to blockchain, but the implications in AI are particularly significant due to its potential to disrupt various sectors, including healthcare and finance.
Valuation Mismatches and Market Concerns
The disconnect between investor expectations and technological advancements also manifests in valuation mismatches. Major tech companies, commonly referred to as hyperscalers, such as Microsoft, Google, and Meta, are projected to spend over $500 billion in 2026 on AI-related infrastructure. Analysts indicate that this spending is accelerating faster than profits, creating concerns reminiscent of past market corrections. Some reports have noted that stock prices are rising ahead of earnings, raising alarms about potential bubbles in the market.
Xiao critiques the current state of investment, advocating for a new breed of venture capitalists equipped with technical expertise. She argues that without knowledgeable investors, funding often follows a herd mentality, driven by excitement rather than solid research. This has led to inflated valuations for AI stocks, fostering fears of an impending market correction.
Geopolitical factors further complicate the investment landscape. The Atlantic Council has outlined how AI will influence global affairs in 2026, highlighting competition among nations like the United States and China to dominate AI technology. Investors’ strategies often lag behind these fast-paced developments, raising questions about future investment viability.
To bridge the gap between research and investment, industry leaders like Xiao are advocating for increased education and collaboration. Leonis Capital hosts workshops and publishes insights aimed at demystifying frontier AI for investors, equipping them with the necessary tools to evaluate startups more effectively. This initiative has gained traction, with a noticeable rise in AI-focused venture funds led by individuals with research backgrounds.
As the AI sector moves forward, Xiao’s vision of a more informed funding ecosystem could help mitigate the risk of bubbles and promote sustainable growth. Yet, as 2026 unfolds, the challenges posed by elevated capital expenditures and geopolitical tensions will continue to shape the investment landscape.
Looking ahead, Leonis Capital’s predictions for 2026 aim to correct the missteps of 2025, advocating for diversified portfolios that include local AI growth and embodied systems. These strategies align with forecasts suggesting a significant focus on robotics cycles by 2028. For insiders, this means scrutinizing investments through both technological merit and geopolitical implications.
The AI investment arena stands at a critical juncture. By addressing the lag in understanding between researchers and investors, as highlighted by figures like Jenny Xiao, the industry could shift toward a more balanced and innovative future. This evolution would allow capital to pursue genuine progress rather than merely echoing past trends.
