What unique techniques does Keith Newman use to scale AI beyond GPUs with guests like Ted Shelton?

In the episode with Ted Shelton, COO of Inflection AI, I explored various techniques for scaling AI beyond GPUs. Ted shared how their team pivoted from consumers to enterprise after their Microsoft deal and why seamless infrastructure is key. They moved their entire platform to be able to run on heterogeneous architecture, meaning they can deploy on Intel, AMD, and NVIDIA. This approach allows for the deployment across multiple clouds, chip architectures, and models, running it all seamlessly.

Ted also emphasized the importance of "getting to the no" for founders and how enterprises can embrace sovereign AI. He highlighted how Inflection’s approach to model customization unlocks massive business value.

Would you like to know more about any specific aspect of scaling AI beyond GPUs?