AI system ‘building blocks’ – a deep dive on value, IP, open v proprietary and legal risks

In this episode, Vik Khurana explains the ‘building blocks’ of an AI system and various legal and commercial trends, risks and issues. In particular:

• What are the building blocks?  A deep dive on training data, weights and parameters, model code, inference code and computational resources.  
• Where does value and IP reside within each block?  What is typically accessible and what is not?  What is proprietary and what is open sourced?
• The different licensing models and regulatory exceptions around the development and use of open source AI.

Here is a link to the ‘building blocks’ diagram Vik mentions in the episode that may be useful for listeners to view during the first half of the episode.

Send us a text

Thanks for listening!
If you have any feedback, questions or comments, please email us at theroadmap@bristows.com

Find all the episodes as we release them here, and don’t forget to subscribe! Follow us on X and LinkedIn using #TheRoadmapPod