US-based artificial intelligence startup Anthropic has opened a Tokyo office to expand its business in the Asia-Pacific.
The company announced the opening at a news conference last week, where it also said it plans to establish a research and development center in Japan.
Anthropic was founded by former employees of OpenAI, the developer of ChatGPT. The company says its focus is on safe development of AI, not just commercialization.
Anthropic CEO Dario Amodei said in an interview that the strategy in Japan is to work with local companies.
“We focus more on working with businesses and less on the consumer side,” said Amodei. “And so our view is that there is really an opportunity to create economic value with Japanese businesses.”
Amodei says Anthropic’s research focus is on responsible and reliable AI development.
“Myself and some colleagues had just developed a number of new AI models GPT-2, GPT-3. And you know we really took step back,” he said. “If we scale it up, it’s going to transform everything very quickly. We really need to do this right. We really need to do this in a responsible manner.”
Amodei says Anthropic doesn’t develop models that generate images and videos. That’s because of the risk of excessive use by consumers and the development of addictions and other social problems.
He also touched on the massive investment in AI companies and the risk of “AI bubbles” developing in stock markets.
“My view is that this will be a boom if progress in the underlying capability of the models continues,” he said.
“If that ends up not being correct, then you know, it would certainly be a bubble. I will bet pretty heavily based on the decade more of history that I’ve seen that the progress is going to continue.”
Amodei concluded that certain restrictions or regulations on AI are necessary for the long-term development of the technology.
