OpenAI is launching OpenAI Japan, their first office in Asia.
Takeaway Points
- OpenAI announced on Sunday that it is launching its first office in Asia.
- OpenAI is also releasing a GPT-4 custom model optimized for the Japanese language.
- They are partnering with the Japanese government, local businesses, and research institutions to develop safe AI tools.
A new office in Tokyo
The AI research and deployment company OpenAI announced on Sunday that it is launching its first office in Asia and releasing a GPT-4 custom model optimized for the Japanese language.
OpenAI said that it is expanding into Asia with a new office in Tokyo, Japan, and that they are partnering with the Japanese government, local businesses, and research institutions to develop safe AI tools that serve Japan’s unique needs and to unlock new opportunities.
The firm said that they chose Tokyo as their first Asian office for its global leadership in technology, culture of service, and a community that embraces innovation.
Sam Altman, CEO of OpenAI, said that they are happy to be in Japan because of their history of people and technology coming together to do more.
“We’re excited to be in Japan which has a rich history of people and technology coming together to do more. We believe AI will accelerate work by empowering people to be more creative and productive, while also delivering broad value to current and new industries that have yet to be imagined,” said Sam.
About OpenAI Japan
According to the report, Tadao Nagasaki will be the new president of OpenAI Japan. Mr. Nagasaki will lead the commercial and market engagement efforts and help build the local team that will advance Global Affairs, Go-to-Market, Communications, Operations and other functions in serving Japan.
OpenAI said that they’re providing local businesses with early access to a GPT-4 custom model specifically optimized for the Japanese language. This custom model offers improved performance in translating and summarizing Japanese text, is cost-effective, and operates up to 3x faster than its predecessor. Speak, a top English learning app in Japan, is seeing 2.8x faster tutor explanations in Japanese when users make a mistake with a 47% reduction in token cost, unlocking higher quality tutor feedback in more places and with higher limits per user.
Earlier this month, OpenAI launched new features to give developers more control over fine-tuning with the API and introduced more ways to work with our team of AI experts and researchers to build custom models.
According to the report, the new features include: Epoch-based Checkpoint Creation: Automatically produce one full fine-tuned model checkpoint during each training epoch, which reduces the need for subsequent retraining, especially in the cases of overfitting
Comparative Playground: A new side-by-side Playground UI for comparing model quality and performance, allowing human evaluation of the outputs of multiple models or fine-tune snapshots against a single prompt
Third-party Integration: Support for integrations with third-party platforms (starting with Weights and Biases this week) to let developers share detailed fine-tuning data to the rest of their stack.
Comprehensive Validation Metrics: The ability to compute metrics like loss and accuracy over the entire validation dataset instead of a sampled batch, providing better insight on model quality.
