'Our vision became less clear': Former employee warns OpenAI is risking safety with profit push
Daniel Ziegler, one of 11 current and former OpenAI employees to sign an open letter calling for more transparency about AI risk, told Capital Brief how the startup began to change during his tenure.
It’s been estimated that about 14 million people use ChatGPT every day. When Apple’s next iPhone operating system launches in later in the year, that number will rocket up to over 70 million.
On Monday Apple announced a new partnership that will put ChatGPT onto its upcoming iPhone 16, as well as the iPhone 15 Pro devices released last year. OpenAI’s deal with the former biggest company in the world is the most significant it’s struck since it accepted USD13 billion from Microsoft, the current biggest company in the world.
Yet it comes at a time when current and previous employees of the startup are questioning OpenAI’s commitment to responsible AI development, and calling for a more transparent and open culture.
“We started out with a really clear vision for safety research, but it became less clear as time went on,” said Daniel Ziegler, an engineer who worked on OpenAI’s safety team from 2018 to 2021. He was among 13 current and former employees to last week sign an open letter calling on OpenAI, Google’s DeepMind and other AI labs to establish protections that give employees “a right to warn about advanced artificial intelligence.”
