Science: Using AI Makes Workers 19 Percent More Likely to Make Mistakes

VideoFlow /
VideoFlow /

Artificial intelligence (AI) is being hailed as a revolutionary new technology that will allow big corporations to fire tons of people and replace them with sentient robots. But is it? As more and more companies are turning to AI and relying more heavily on it, they’re finding that it’s cutting into their bottom line. That’s according to new research carried out by Harvard University and Boston Consulting Group (BCG).

AI does, in fact, increase efficiency in some areas. Harvard researchers had BCG workers perform tasks using Chat GPT-4 for the study. The workers were able to carry out 12.2% more pointless, paper-pushing HR tasks using Chat GPT. They also carried out soul-crushing, redundant office tasks 25% faster.

So, there is some upside to using AI in the corporate workspace (HR departments love it). There are also downsides that they found, however. When workers relied too heavily on AI, they were 19% more likely to make mistakes.

These types of mistakes in the corporate world not only waste a lot of time, but also cost money to correct. The wrong mistake in the wrong area could even lead to a company being sued. The lesson, as always, is that human workers cannot be fully replaced by the robots.

It’s also an open question whether this technology is even going to be viable for much longer. It costs $700,000 per day for the parent company OpenAI to operate ChatGPT. The company lost $540 million in 2022. Industry analysts are openly speculating that OpenAI will be bankrupt by this time next year if they don’t turn things around quickly.

That’s not as expensive of a mistake as Bud Light made earlier this year, but the rush to AI is still going to be expensive to investors if the whole thing backfires.