How AI is Eating Finance - with Mike Conover of Brightwave

TL;DR

Mike Conover, Founder of Briwve, discusses the potential of language models in the financial services industry and the importance of domain expertise in developing AI systems.

Transcript

hey everyone welcome to the lateen space podcast this is alesio partner and CTO and residence and deible partners and I have no co-hosts today as you can see um this works is in Vienna at iclr having fun in in Europe and uh we're in the brand new studio um as you might see if you're on YouTube there's still no sound panels on the wall Mike tried re... Read More

Key Insights

• 🐕‍🦺 Language models can provide valuable insights in the financial services industry, expanding individuals' reasoning capabilities and enhancing decision-making processes.
• 🆘 Fine-tuning and instruction tuning of models can help create specific behavioral regimes that are useful for different financial tasks.
• 🤗 Open-source language models have a diminishing advantage as companies move towards creating their own unique datasets to train models specific to their needs.
• 🖐️ The development of knowledge graphs and structured data extraction plays a crucial role in enabling advanced quantitative reasoning in financial analysis.
• 🪛 Human-in-the-loop evaluations and feedback are essential for ensuring the accuracy and performance of AI-driven financial models.
• 💁 The potential for AI hedge funds exists, but the precise form and extent of their capabilities are still uncertain. It is likely that a combination of automated analysis and human decision-making will prevail in the financial industry.
• 🤗 The future of open-source language models lies in the development of new training data that elicits specific behaviors, leading to differentiated and valuable AI systems.

Questions & Answers

Q: How do language models pick up on polarization and biases in the data they are trained on?

Language models are a compression of the world as it existed at the time of their training, and they encode the biases and semantics present in the data. This includes gender biases and other societal beliefs, which can manifest in the models' behavior.

Q: How does Briwve differentiate itself in the financial services industry?

Briwve provides a partner-in-thought approach, leveraging language models to analyze vast bodies of content and help users reason about investment opportunities. Its focus on qualitative reasoning and deep insights sets it apart from other AI-driven financial models.

Q: How does Briwve ensure accuracy and factuality in its outputs?

Briwve employs multiple evaluation methods, including human annotators and domain experts, to assess the quality and factual accuracy of its generative outputs. It also compares the outputs against trusted sources and other models to ensure consistency and reliability.

Q: How does Briwve handle privacy and confidentiality concerns in the financial services industry?

Briwve has built-in controls to manage and protect sensitive data throughout its lifecycle. With its rigorous approach to data security and access management, the company instills confidence in its customers when sharing sensitive financial information.

Summary & Key Takeaways

• Mike Conover explains how his background in machine learning and finance led him to create Briwve, a financial services company that utilizes language models to provide deep insights.

• Briwve's focus is on expanding individuals' ability to reason about the structure of the economy and the markets, leveraging the power of language models to perceive more than humans can.

• The company's product acts as a partner in thought, helping finance professionals make informed investment decisions and quickly develop views on various investment themes.