Anthropic CEO Dario Amodei has said that differences over the vision and approach to artificial intelligence were the primary reasons he left OpenAI, the AI company he joined in 2016. Speaking at Nikhil Kamath’s podcast, Amodei said that after joining the company, he and a few other colleagues developed their own idea of how AI should be built. “Eventually, a few other employees and I just kind of had our own revision for how we wanted to make AI and what we wanted the company to stand for. And so we went off and founded Anthropic,” Dario Amodei said. During the podcast, Amodei revealed two major reasons behind his departure from OpenAI. First, he said, was belief on how AI should be developed and vision over scaling AI models.Dario Amodei stated that increasing the size of AI models and the amount of data they are trained on can lead to major improvements in performance. Speaking about his early observations, Amodei said he first noticed this trend in 2019 while working with GPT-2. He explained that as models were made larger and trained on more data with greater computing power, their capabilities improved significantly.According to Amodei, this pattern, often referred to as “scaling laws,” showed that performance gains were closely linked to model size and resources. While some additional techniques such as reinforcement learning were used to fine-tune results, he said the main driver of progress was simply building bigger models with more data and computing capacity.“There were a lot of folks, inside and outside, who didn’t believe it at all. And we really made the case to leadership like, ‘This is important, this is gonna be a big deal.’ And I think they were kind of starting to believe us and ultimately went in that direction,” he added.
Anthropic CEO: AI’s economic implications are gonna be enormous
During the podcast, Dario Amodei further revealed that he doubted OpenAI’s approach over responsible AI development. He believes that if AI models were to become as capable as human brains, they could have huge effects on the economy, politics, and society. “If these models are gonna be kind of general cognitive agents, like general cognitive tools that match the capability of the human brain, we’d better get this right,” he said.“The economic implications are gonna be enormous. The geopolitical implications are gonna be enormous. The safety implications are gonna be enormous. It’s gonna transform how the world works. And so we needed to do it in the right way,” he added.
Leave a Reply