Uploaded November 2024 | Updated September 2026, 2 weeks ago
Filmed just three weeks before the launch of ChatGPT, OpenAI Co-Founder and President, Greg Brockman, sat down with Anyscale Co-Founder Robert Nishihara to discuss OpenAI’s mission of advancing AI in a way that’s beneficial to humanity.
In this chat, Brockman shares the infrastructure challenges OpenAI faced when they began to train larger and larger models, and how Ray helped them train at scale. With the infrastructure problem resolved, Brockman then turns to why large language models will be the immediate next step, and dives into how training on broad datasets helps build better models.
Finally, Brockman and Nishihara end their conversation discussing the future of AI. Brockman closes by sharing that every year at OpenAI, they aim to achieve something that was previously considered impossible. Of course, mere weeks after this conversation, OpenAI launched ChatGPT and revolutionized our relationship with AI.
--
🔗 Connect with us:
Subscribe to our YouTube channel: youtube.com/@anyscale
Twitter: https://x.com/anyscalecompute
LinkedIn: linkedin.com/company/joinanyscale
Website: anyscale.com
Filmed just three weeks before the launch of ChatGPT, OpenAI Co-Founder and President, Greg Brockman, sat down with Anyscale Co-Founder Robert Nishihara to discuss OpenAI’s mission of advancing AI in a way that’s beneficial to humanity.
In this chat, Brockman shares the infrastructure challenges OpenAI faced when they began to train larger and larger models, and how Ray helped them train at scale. With the infrastructure problem resolved, Brockman then turns to why large language models will be the immediate next step, and dives into how training on broad datasets helps build better models.
Finally, Brockman and Nishihara end their conversation discussing the future of AI. Brockman closes by sharing that every year at OpenAI, they aim to achieve something that was previously considered impossible. Of course, mere weeks after this conversation, OpenAI launched ChatGPT and revolutionized our relationship with AI.
--
🔗 Connect with us:
Subscribe to our YouTube channel: youtube.com/@anyscale
Twitter: https://x.com/anyscalecompute
LinkedIn: linkedin.com/company/joinanyscale
Website: anyscale.com
![[Ray Meetup] Ray + vLLM in Action: Lessons from Pinterest and Large Scale Distributed Inference](https://i.ytimg.com/vi/HDSy09hrm2I/mqdefault.jpg "[Ray Meetup] Ray + vLLM in Action: Lessons from Pinterest and Large Scale Distributed Inference
Listen in to our Ray Meetup where we explored batch inference at scale with Ray and vLLM! Learn how Pinterest scales batch inference using Ray, and get a first look at Anyscale’s latest tools—Ray Serve and Data LLM—for orchestrating large-scale LLM inference. We’ll cover topics like batch inference, prefill-decode disaggregation, DP/EP parallelism, and custom request routing.
Speakers:
Chia-Wei Chen, Software Engineer, ML Training Infra, Pinterest
Kourosh Hakhamaneshi, AI Lead, Anyscale")
