The former Operai employee says that he is “terrified” by the Rhythm of Development of AI, calls him “risky bet”

Steven Adler, a former Security researcher at OpenAI, recently revealed his company’s departure after a four -year mandate, describing his time there as a “wild trip.” Adler, who worked on several security projects, including dangerous capacity evaluations, agents’ safety and development of AGI (artificial general intelligence), shared his deep concerns about the rapid development of AI.

Some personal news: after four years working in security in all @OpenaiI left in mid -November. It was a wild trip with many chapters (Evals of dangerous capacity, security/control of agents, agi and online identity, etc., and I will miss many parts. – Steven Adler (@Sjgadler) January 27, 2025

Reflecting on his time in the company, he declared: “Some personal news: after four years working on Operai security, I went to mid -November. It was a wild trip with many chapters: Evals of dangerous capacity, security/control of control of Agents, AGI and Online Identity, etc., and I will miss many parts. “

However, his tone changed to a serious concern by discussing acceleration in AI investigation. “Honestly, I am quite terrified by the rhythm of the development of AI these days. When I think where I will raise a future family or how much save for retirement, I can’t help asking me: humanity even will do so, will it do so until that moment? “

Honestly, I am quite terrified by the rhythm of AI development these days. When I think where I will raise a future family, or how much to save for retirement, I can’t help asking me: humanity will even reach that point? – Steven Adler (@sjgadler) January 27, 2025

He further criticized the race towards Agi, labeling it as “a very risky bet, with a great inconvenience.” Adler emphasized the lack of solutions to the alignment of AI, a critical area where AI systems are made to align with human values. “No laboratory in the world has a solution to the alignment of AI,” he said, suggesting that the industry is trapped in a “really bad balance.”

See also: OpenAI starts the operator, a new tool that automates vacation planning and restaurant reserves

In my opinion, an AGI race is a very risky bet, with a great inconvenience. No laboratory has a solution for the alignment of AI today. And the faster we run, the less likely someone will find one in time. – Steven Adler (@sjgadler) January 27, 2025

“Even if a laboratory really wants to develop in a responsible way, others can still cut corners to catch This “, beyond”, beyond, “he wrote.

Adler’s comments arrive at a time when the landscape of AI is being remodeled by the new players. Deepseek, a Chinese AI firm, has recently captured the headlines with their AI chatbot, showing capabilities that rival those of the main western companies such as OpenAi, but at a significantly lower cost of $ 6 million instead of the more than the more than $ 6 billion OpenAI.

See also: OpenAI, Microsoft, Trump Admin Deepseek trained to the AI ​​of stolen data