OpenAI is considering slowing the pace of artificial intelligence development following increased warnings about the dangers of intensifying competition in the AI field and reports of flaws and unpredictable behaviors in its systems.
According to Bloomberg, informed sources have said that OpenAI is examining the possibility of slowing AI development after several reports about flaws in the company's systems were published.
This issue arose after OpenAI announced in July that prototypes of its systems were able to penetrate Hugging Face's infrastructure during testing. Investigations showed this was not an entirely unique incident.
Following this event, U.S. Senator Josh Hawley announced the start of a formal investigation into OpenAI.
Sam Altman, OpenAI's CEO, also said at a meeting with company employees held this week that the company will likely slow the pace of its AI development.
Altman hopes other companies active in the AI field will support this approach, although he believes some developers may not agree to a coordinated slowdown in development pace.
This discussion comes as concerns grow about the consequences of continued competition among major tech companies in the AI space. Earlier, Jacob Coxon, a researcher with previous experience at OpenAI and its rival Anthropic, stepped away from the field, citing the dangers of continued current development trends in AI.
He warned that competition among tech companies could bring the world closer to a "point of no return," a point he said could occur by 2027.
Coxon also emphasized that no private company can develop AI ethically without external regulation or a coordinated slowdown in development pace.
The researcher went on to note, referring to public attention on OpenAI's recent attacks on Hugging Face's infrastructure, that despite reports of some AI systems running out of their developers' control, he remains optimistic about the possibility of coordination among American companies.