ChatGPT, the mass-market artificial intelligence tool which has gained widespread popularity over the last six months, built a stock portfolio that is outperforming the most popular funds in the United Kingdom, a new report has found.
Researchers at Australian financial services company Finder provided ChatGPT with data from the 10 most popular British investment funds and asked the generative AI system to choose stocks for a portfolio. ChatGPT selected 38 stocks which have risen nearly 5% in value over the past eight weeks, thereby outperforming the average returns of the 10 funds, which has declined nearly 1% over the same time horizon, according to an analysis posted by Finder.
ChatGPT selected companies such as Coca-Cola, Visa, Accenture, Amazon, Intel, Mastercard, and Berkshire Hathaway based on criteria such as low debt levels, sustained growth, and assets that generate an advantage over their competitors. The portfolio outcompeted the average earnings from funds created by Vanguard, Fidelity, and HSBC.
Finder released the results of the short-term experiment with a survey about the current use of ChatGPT for investment advice: some 35% of respondents said they would not use the AI system for help with their finances, while 19% said they would receive financial advice from the system and 8% said they had already used the system for such purposes. The most common response, however, was the 38% of respondents who said they had never heard of ChatGPT.
Various official and unofficial studies have demonstrated how randomized stock selection processes, including those administered by a hamster, can outperform trained professionals. Investment management is nevertheless one of several competencies demonstrated by ChatGPT over the past several months.
The tool, which was created by OpenAI with investments from various technology firms, earned a grade between B and B- on a final exam presented to graduate students at the University of Pennsylvania’s Wharton School. Christian Terwiesch, the operations management professor who administered the exam to ChatGPT, concluded in a research paper that the AI system “does an amazing job” with analytical questions but occasionally made “surprising mistakes.”
ChatGPT has also performed near or above the passing threshold for the United States Medical Licensing Exam and earned passing scores on the multiple-choice section of the bar exam.
Even as many knowledge workers stand to gain from ChatGPT and similar tools through increased productivity, white-collar industries are the most likely to see widespread technological unemployment as a result of generative AI developments. One recent forecast from Goldman Sachs estimated that AI could eliminate 7% of positions in the United States, largely in sectors that rely on office work such as administrative support and legal, while positions in sectors such as construction and logistics are forecasted to remain broadly intact.
Some technology experts have warned about the possible effects of AI and contend that government officials should temporarily mandate a pause in development of the technology. Elon Musk, a co-founder of OpenAI who has since resigned his seat on the company’s board, and Steve Wozniak, a co-founder of Apple, made headlines after they endorsed an open letter that called for a six-month moratorium on developing AI solutions. Geoffrey Hinton, a computer scientist who pioneered the neural network method which serves as a basis for AI innovations, resigned from his position at Google over his concerns about the technology.