ChatGPT has shown capabilities in being able to predict future interest rates through various language models by central bank members.
This was revealed through research at Sheffield Hallam University which ran through a range of language models using ChatGPT to analyse speeches from the Bank of England Monetary Policy Committee members before decisions on interest rates were made.
Researchers identified each speech as dovish, hawkish or neutral based on their vocal tone and content discussed. This was then integrated into an economic model that would predict how each speaker would vote at the next meetings.
Findings revealed by ChatGPT that its analysis on the speeches determined that voters who gave more neutral speeches were most likely to vote for interest rate hikes in the next meetings.
Researchers at Sheffield Hallam University believe this approach could be branched out to other central bank communications, such as guidance. It also demonstrates the rising potential of AI softwares on the wider economic and financial sectors.
Dr. Drew Woodhouse, Senior Lecturer in Economics in Sheffield Business School, said: “Our findings highlight the predictive potential of tools like ChatGPT for processing human beliefs and expectations.
“This has major implications for forecasting policy decisions and modelling economic expectations.”
Whilst AI has seen exponential growth since the turn of the year, calls have come in to safeguard it from its own possibly dangerous potential.
Angel Maldonado, CEO and Founder of privacy enhancing commerce platform Empathy.co, believes that it has come as ‘no surprise’ that OpenAI’s ChatGPT was hit with scrutiny.
He shared to Payment Expert: “It’s no surprise that OpenAI’s ChatGPT has come under scrutiny. The amount of information that AI chatbots can access amounts to an ocean’s worth considering that they have the potential to reach any and all parts of the internet.
“Whilst AI is an incredible power for innovation, we must harness it in a way that is ethical and human-first.”