Iain Dunning: The exponential pace of AI is reshaping market predictions, current dynamics resemble gambling, and the complexity of models challenges traders’ interpretability | Odd Lots

AI advancements are transforming trading strategies, raising questions about sustainability and profitability in market predictions.

Listen on Odd Lots

Share

Add us on Google by Editorial Team Jun. 5, 2026

Key takeaways

• The frontier of AI technology is challenging to reach, and there is skepticism about maintaining pace with rapid advancements.
• AI progress is accelerating exponentially, impacting multiple markets simultaneously.
• Continuous investment in AI for market predictions may yield diminishing returns over time.
• AI models are evaluated based on their ability to reduce errors in quantitative research.
• New AI model releases have led to fluctuating expectations about their capabilities.
• Current market dynamics are perceived as resembling gambling rather than traditional investing.
• Complex models can provide insights into market behaviors, but their complexity makes full understanding difficult.
• Models can identify connections between meme stocks and other assets, despite lacking fundamental justification.
• Risk management strategies from high-frequency trading are difficult to apply to long-term discretionary trading.
• The world is repricing money and debt, indicating a shift back to real assets.
• AI advancements are reshaping the landscape of trading strategies and market predictions.
• The evolution of AI models has led to a reassessment of their role in augmenting human research capabilities.
• There is a critical view of the sustainability of AI investments in trading and their impact on profitability.
• The complexity of AI models poses challenges for traders in interpreting their outputs effectively.
• The repricing of assets and monetary policy changes are influencing economic fundamentals.

Guest intro

Iain Dunning is Head of AI at Hudson River Trading, where he leads the firm’s work on deploying AI across its trading operations. He has discussed how one of the world’s largest market makers is using AI, including the firm’s growing compute and token usage.

The challenge of reaching the AI frontier

• Reaching the frontier of AI technology is seen as a daunting task.

• I guess reaching the frontier is clearly a very daunting task so maybe it’s with some effort deep seek but beyond that is not a claim I’d be willing to make…

  — Iain Dunning

• Many believe it’s possible to keep up with AI advancements, but there is skepticism.

• I don’t know what the future of open models is if they’re all kind of a serious step back and the frontier is progressing so fast I don’t know how you keep up with that but many people believe that it’s possible I’m not so sure I I’m one of those people there.

  — Iain Dunning

• The competitive landscape of AI model development is intense.
• Challenges exist in maintaining pace with rapid AI advancements.
• The future of open AI models remains uncertain.
• There is a nuanced perspective on the challenges of AI development.

The exponential pace of AI progress

• The pace of AI progress is exponential, impacting various markets.

• The amount of compute I’ll have next year versus this year and the amount of compute I have this year versus last year looks exponentially and we’re doing things today that I didn’t really…

  — Iain Dunning

• AI advancements have broad applicability across markets.
• The rapid advancements in AI are reshaping market predictions.
• Continuous investment in AI may lead to diminishing returns.

• Surely you can’t just like keep getting better at predicting markets forever it’s gotta be some sort of forcing function where you know your margins go to zero…

  — Iain Dunning

• The sustainability of AI investments in trading is questioned.
• There are economic implications of AI advancements in trading.

Evaluating AI models in quantitative research

• AI models are evaluated based on their ability to reduce errors.

• …we see…an ever reducing set of errors they make…we spent some time in the past couple of weeks trying to come up with objective measures to index them against humans in the act of quant research ideating signals…

  — Iain Dunning

• Error reduction is a key focus in AI model assessment.
• The release of new AI models has led to fluctuating expectations.

• …we had a sort of a false start with ai…we left feeling empty because we realized that it was not there and was not able to meaningfully augment human researchers…

  — Iain Dunning

• Changing expectations surround AI model releases.
• The evolution of AI models affects their perceived capabilities.
• AI models’ role in augmenting human research is reassessed.

Market dynamics and the gambling analogy

• Current market dynamics resemble gambling rather than investing.

• I feel like we’re in that world today… the cynical thing is everything is gambling and so everything is some sort of like gambling market including public markets.

  — Iain Dunning

• There is a shift in market behavior and valuation methods.
• The evolving nature of market investments is critically viewed.
• Traditional valuation methods are being departed from.
• The analogy highlights the speculative nature of contemporary markets.
• The gambling analogy reflects a critical perspective on market dynamics.
• Market behavior is increasingly speculative.

Complexity and interpretability of AI models

• Models provide insights into complex market behaviors.

• I do think there are diagnostics we’ve done where we can see things that we can understand… it might be illusory because it’s a very very complex object…

  — Iain Dunning

• Understanding complex models fully is challenging.
• The complexity of models poses interpretability challenges.
• Models identify connections between meme stocks and other assets.

• …it felt like the model kind of understood meme stocks from first principles… from a fundamentals perspective this has no right…

  — Iain Dunning

• Correlations identified by models challenge conventional wisdom.
• The effectiveness of models is crucial for traders.

Risk management in trading strategies

• Applying risk management from high-frequency to long-term trading is complex.

• …I don’t know how you generalize this logic to long term discretionary trading where the idea of like risk checking… it’s not so obvious to me how you apply that…

  — Iain Dunning

• Differences exist between high-frequency and long-term trading strategies.
• Risk management complexities are emphasized across trading strategies.
• The application of risk management strategies is not straightforward.
• Traders face challenges in adapting risk management to different strategies.
• The complexities of risk management are vital for traders and investors.
• Understanding these differences is crucial for effective trading.

Repricing money and debt in the economic climate

• The world is repricing money and debt, focusing on real assets.

• …gold’s run may tell you something about how the world is repricing money and debt all of those point back to real assets…

  — Iain Dunning

• Asset prices are influenced by monetary policy changes.
• Economic fundamentals are impacted by the repricing of assets.
• The current economic climate is shaped by these changes.
• There is a strong viewpoint on the implications of asset prices.
• The relationship between asset prices and monetary policy is crucial.
• Understanding these dynamics is important for economic analysis.

Disclosure: This article was edited by Editorial Team. For more information on how we create and review content, see our Editorial Policy.