Upon its release in November 2022, OpenAI’s flagship platform – ChatGPT – created major waves across the internet and the legacy media. Google’s (NASDAQ: GOOGL) offering – Bard – while not as famous as the other artificial intelligence (AI), has also garnered a significant fanbase.
For about a year now, Finbold has been testing the two AI’s ability to analyze both the stock market and the cryptocurrency industry, as well as how capable they are when it comes to predicting future prices in a long and ongoing series of articles.
With many predictions made and with the New Year coming in just a few days, it is time to assess just how accurate the two platforms have been in their educated guesses on how the markets might move.
Picks for you
Google Bard’s predictions for late 2023
In late May 2023, Google Bard was put in a position that could be described as a conflict of interest by some as it was asked where Google’s stock might stand at the year’s end.
At the time, the AI opined that the tech giant is likely to be priced close to $150 and, given that it isn’t quite the last day of 2023 yet and that GOOGL is priced around $140 at the time of writing, Bard proved fairly accurate.
Early in June, Bard was tasked with assessing the fortunes of Bud Light’s owner, Anheuser-Busch InBev (NYSE: BUD), and, given the difficult and uncertain situation at the time, the AI proved rather accurate.
It estimated that BUD would end the year near $67, and, indeed, its price at the time of writing stands at $63.63 – not perfect, but given how many months have passed since the question was asked, rather good.
Mere days later, the AI had to put its analysis skills to the test yet again – this time to make an educated guess about Amazon (NASDAQ: AMZN). At the time when AMZN was priced just above $120, Bard predicted that it would stand at $120. Given that, by the time of writing, Amazon rose above $150, the artificial intelligence certainly failed to see where the wind was blowing.
In the final case under consideration, Google Bard was asked to predict the fortunes of the Chinese electric vehicle (EV) company Nio (NYSE: NIO). In June, the AI predicted the company’s stock would stand between $10 and $12. Given that NIO managed to briefly rise above $14 during the summer before falling and finding itself close to $8 in late December, Bard wasn’t particularly accurate in this assessment.
ChatGPT’s predictions for late 2023
This year, ChatGPT had significantly more opportunities to make predictions about the start of 2024, but given that it is not yet January 1, it is still impossible to assess how accurate these are.
In October, however, OpenAI’s flagship platform was given a task that could be verified already in late December. When asked how the S&P 500 index might fare at the end of the year, the AI opined that it might reach new highs and go up to 4,500 and beyond. Interestingly, ChatGPT mentioned the ongoing artificial intelligence boom as one of the reasons for its seeming bullishness.
Even more interestingly, the AI proved too cautious in its assessment as the S&P 500 is very close to 4,750 at the time of writing.
Head-to-head: predicting Tesla (NASDAQ: TSLA)
There is one stock that has, throughout 2023, drawn significant attention from investors and that has been repeatedly inquired about – Elon Musk’s EV maker, Tesla (NASDAQ: TSLA). The first question was directed toward the company’s stock price upon the launch of the cybertuck.
In reality, Tesla’s price experienced significant fluctuation around the time of the launch, which occurred on November 30, 2023. In the immediate timeframe of the delivery event, the stock actually peaked on November 28, closing at $246.72. On launch day, it closed at $240.08.
It continued to decline until December 4, when it closed at $235.58 and then entered an uptrend it maintained for about a week.
Interestingly, when ChatGPT was first asked about the event, it simply noted that there are many factors at play, such as heightened media coverage in the short term, and added that, depending on the quality of the actual vehicle, it could ultimately drive the price in either direction. It didn’t offer any figures.
The second attempt, made on November 21, wasn’t particularly fruitful either. ChatGPT offered a precise range – between $200 and $250 – which, while technically correct, is definitely too wide to conclusively assess the quality of the AI’s analysis.
Google Bard was asked to make the same prediction on October 25. This artificial intelligence proved more precise and more explicitly wrong as it set TSLA’s price at cybertruck’s launch at $218. Some grace can, however, be extended to the AI as Tesla’s stock just ended a significant downturn and hovered around $205 at the time of the inquiry.
Both AIs were also asked to predict Tesla’s price at the end of 2023.
Given that the price of TSLA is near $255 at the time of writing and that Bard predicted in late May that it would, by the end of the year, be somewhere between $200 and $300, it can be said that it was technically correct.
ChatGPT’s answer, provided in early June, is actually unfalsifiable as the AI refused to provide a prediction. It merely analyzed the factors likely to affect the price of Tesla stock at the end of 2023.
Conclusion
Ultimately, the current versions of the two prominent large language models (LLMs) appear significantly better at summarizing broader market trends and relevant factors, as well as figuring out the general direction, than at making stock price predictions.
While it can’t be decisively said that one had better estimates than the other, ChatGPT appears to have been programmed to speculate significantly less than Google Bard.
Buy stocks now with Interactive Brokers – the most advanced investment platform
Disclaimer: The content on this site should not be considered investment advice. Investing is speculative. When investing, your capital is at risk.
