Recently, Jim Buckley, an elderly businessman, received a shock when his 14-year-old grandson showed him a ChatGPT AI chatbot which mis-identified him as the IRA terrorist responsible for bombing the Baltic Exchange in 1992. chatGPT is an AI chatbot trained with Reinforcement Learning from Human Feedback (RLHF) and is capable of producing answers like a human. Mr. Buckley is recognized for his role as CEO of the Exchange and had to experience consequences of the attack firsthand. It's clear that AI chatbots should not be relied on heavily and should be cross-checked for accuracy.
Jim Buckley, a 79-year-old former chief executive of the maritime organization Baltic Exchange, was recently left shocked when an AI chatbot wrongly identified him as the perpetrator behind the fatal 1992 IRA bombing attack on the Exchange, killing three and injuring 91. ChatGPT, an AI language model, made an incorrect statement which could have serious implications. This incident shows the dangers of relying too much on AI without proper caution.
Explore the evolution of tech policy from Obama's optimism to Harris's vision at the Democratic National Convention. What's next for Democrats in tech?