OpenAI has pressed pause on its initiative to allow explicit conversations within ChatGPT. Announced on Thursday, the move comes amid worries about the potential harmful effects and the reputational risks it could pose for the generative AI frontrunner. The decision, brought to light by the Financial Times, arrives as the San Francisco-based tech giant shifts its focus towards enhancing professional tools in a bid to ramp up revenue.
Back in October, OpenAI CEO Sam Altman made headlines by suggesting a relaxation of ChatGPT’s restrictions, aiming for a December rollout with the promise to "treat adult users like adults." Fast forward five months, OpenAI informed the FT of its desire to delve into long-term research on the implications of sexual interactions and emotional connections. The company acknowledges a lack of "empirical evidence" at present, urging caution before making further decisions.
“We're committed to understanding the true impact before moving forward,” an OpenAI representative commented.
Meanwhile, OpenAI has remained mum regarding further comments but internally, some employees have voiced concerns about the so-called "lemon mode." Questions linger over its alignment with OpenAI's overarching mission to ensure technology benefits all of humanity. Investors, too, are wary of the reputational risks despite potential commercial benefits, according to the British daily.
The rollout isn’t just fraught with ethical concerns. Technically, the age verification system developed by OpenAI reportedly suffers from an error rate exceeding 10%, a detail raising alarms about potentially exposing minors to explicit content. With the company already embroiled in several legal proceedings in the U.S., this remains a sensitive hurdle to overcome.