
Research from the University of Chicago indicates that OpenAI's GPT-4 surpasses human capabilities in financial statement analysis and financial forecasting. The study highlights that GPT-4 can predict future earnings with 60% accuracy and achieve an unlevered alpha of 12% and a Sharpe ratio of 3 using a straightforward Chain of Thought prompt. Despite these findings, skeptics argue that AI lacks the creativity, fantasies, and experience that humans bring to financial analysis.
Interesting paper out of @UChicago: GPT-4 supposedly can predict future earnings with 60% accuracy and an unlevered alpha of 12% and a 3 Sharpe using a straightforward Chain of Thought prompt: https://t.co/YnGsRfJNai Too good to be true?
Just had to switch from GPT-4o to Llama-3 on Groq for a very high volume low-latency classifier You won’t believe it - llama-3 on Groq is almost 10x faster!! 4o looked great but has slowed down a lot over the last few days! Not viable anymore… 🤷♀️
Some talented Stanford students trained a vision language model (Llama 3 8B + Siglip) that performs on par with GPT-4 Vision and Opus. For $500! Open Source blogpost + weights 👇 (note, always take benchmarks with a grain of salt) https://t.co/66k2iiCDlM


