Research done by the AI expert Jose Lopez-Lira unveiled the potential of ChatGPT AI to predict stock movements. Man-made intelligence models, such as ChatGPT, have amazing abilities to detect and interpret data intended for humans. The research focuses on using this AI chatbot to parse titles of news stories based on their sentiment and predict the following day’s stock profit and losses. Lopez-Lira found that ChatGPT exceeded an “arbitrary baseline” in all scenarios, meaning the results often outperformed the traditional methods of predicting stock movements.
The results have troubling implications for the worldwide financial industry with the potential automation of up to 35% of the existing jobs, per a Goldman Sachs report in March. Using computers to interpret news headlines about stock movements would mean a critical shift in the market landscape. Lopez-Lira believes that with the help of better data sets and computers, AI models can show even more potential.
Despite the results, there are still gaps in the AI technology that hindered the research from reaching its full potential, as it did not incorporate target costs or allow the model to do any mathematics. Similarly, Lopez-Lira mentioned that despite the success of the study, people still need to be careful when relying on AI to interpret data. This is particularly true considering the bots ability to interpret news headlines concerning stock movements, as this could potentially put lucrative industries in danger.
When it comes to ChatGPT, the company has been extremely successful in developing AI-driven solutions for customers. With experts building powerful models that perform complicated tasks at scale, many industries such as financial, healthcare, and media are leveraging its capabilities. Furthermore, ChatGPT is also actively participating in innovative research projects and looking for opportunities to apply AI technologies.
Jose Lopez-Lira is a research scientist specializing in artificial intelligence and machine learning. He completed his undergraduate degrees in Cognitive Science and Computer Science from the University of California, San Diego and his Masters in Computer Science from Stanford University. Lopez-Lira is currently an adjunct professor at University of California, San Diego and is an active member of several AI research groups. He has authored multiple research papers, most recently unveiling the potential of ChatGPT AI to predict stock movements. He is passionate about creating AI solutions for today’s challenges and believes firmly in the potential of ChatGPT AI.
