Sam Altman, executive director of OpenAI, the company responsible for ChatGPT, publicly warned about two ways in which artificial intelligence could go "very wrong." The executive shared his concerns through a post on social media.
The first threat identified by Altman is the possibility of losing control over the future of AI. He classified this scenario as "unacceptable" and defended that artificial intelligence must "always serve people." For this, Altman emphasized the need for alignment and safety techniques to always be ahead of the capabilities of AI models.
Altman's second concern involves the excessive concentration of power. He warned about a scenario where the power of AI would be in the hands of "a single person or company," which could be used to "impose their worldview on everyone else," generating results that Altman described as "dystopian."
According to Altman, avoiding these two threats requires following a "narrow middle path." He gave concrete examples, mentioning that a country could gain too much power with the technology, just as an AI lab could end up accumulating excessive control.
