He helped build some of the world’s most powerful AI systems. Now, he’s unplugging.
A former OpenAI and xAI staffer says the breakneck pace inside frontier AI labs left him burned out, and he’s heading home to Vietnam to recover.
Hieu Pham announced his departure from OpenAI in a post on X on Thursday, describing his time at OpenAI and at xAI as a “once-in-a-lifetime experience.”
“At these companies, I have helped creating extremely intelligent entities that will meaningfully improve our lives. The work makes me proud,” he wrote.
The intensity of the work came at a personal cost, he added.
“I cannot believe I would say this one day, but I am burnt out,” Pham said.
“All the mental health deteriorating that I used to scoff at is real, miserable, scary, and dangerous,” he added.
Pham started at xAI in August 2024, before moving to OpenAI in August last year. Before his announcement on Thursday, he had worked at OpenAI for about 7 months, according to his LinkedIn profile.
Pham said he will return to his home country, Vietnam, with his family.
There, he plans to “try something new, and also search for a cure for my conditions,” he said.
“I hope I will heal. Until then,” he added.
Wave of resignations
Pham’s exit comes as a string of AI researchers step away from frontier labs.
Earlier this month, Mrinank Sharma, who led the Safeguards Research Team at Anthropic, announced his departure from the company.
“We appear to be approaching a threshold where our wisdom must grow in equal measure to our capacity to affect the world, lest we face the consequences,” Sharma wrote in a letter on X.
“Throughout my time here, I’ve repeatedly seen how hard it is to truly let our values govern our actions.”
Business Insider’s Jacob Silverman reported earlier this month that there’s a pattern of top researchers walking away and publicly wrestling with their unease about the technology they helped build.
Dylan Scandinaro, who left Anthropic earlier this month to join OpenAI as head of preparedness, wrote on LinkedIn: “AI is advancing rapidly. The potential benefits are great — and so are the risks of extreme and even irrecoverable harm.”
Some AI researchers also say the industry is starting to resemble China’s infamous “996” schedule — shorthand for working 9 a.m. to 9 p.m., six days a week.
Nathan Lambert, a senior research scientist at the Allen Institute for AI, said in an episode of the “Lex Fridman Podcast” published this month that kind of intensity has become the norm at companies like OpenAI and Anthropic.
“That’s what OpenAI and Anthropic are like,” Lambert said, describing a high-pressure culture where long hours are often self-imposed. Many employees, especially programmers, buy in because they want to do the work, he added.
Read the full article here
