The Cloudcast

Bringing Open Source to AI

May 11, 2024 Massive Studios
Bringing Open Source to AI
The Cloudcast
More Info
The Cloudcast
Bringing Open Source to AI
May 11, 2024
Massive Studios

Happy Mothers Day! And let’s explore some new happenings at the intersection of open source and AI. 

SHOW: 820

SHOW TRANSCRIPT: The Cloudcast #820

SHOW VIDEO: https://youtube.com/@TheCloudcastNET 

CLOUD NEWS OF THE WEEK - http://bit.ly/cloudcast-cnotw

CHECK OUT OUR NEW PODCAST - "CLOUDCAST BASICS"


SHOW NOTES:


HAPPY MOTHERS DAY TO ALL THE MOMS

  • Call you mom, be good to your wife, and be a role model for the young girls in your life


OPEN SOURCE MEETS AI

  • The big AI question: 1 model, or many models?
  • 2023 was the year of model diversity, and the beginnings of “open” models (Llama 2, Mistral, etc.)
  • Can we make it easier to contribute to an open model?
  • Can we create communities around models?
  • Can we reduce the costs of expensive parts of AI? 
  • How can businesses bring their knowledge to a model, without security concerns about their data? (options like RAG (retrieval augmented generation) exist)


FEEDBACK?

Show Notes

Happy Mothers Day! And let’s explore some new happenings at the intersection of open source and AI. 

SHOW: 820

SHOW TRANSCRIPT: The Cloudcast #820

SHOW VIDEO: https://youtube.com/@TheCloudcastNET 

CLOUD NEWS OF THE WEEK - http://bit.ly/cloudcast-cnotw

CHECK OUT OUR NEW PODCAST - "CLOUDCAST BASICS"


SHOW NOTES:


HAPPY MOTHERS DAY TO ALL THE MOMS

  • Call you mom, be good to your wife, and be a role model for the young girls in your life


OPEN SOURCE MEETS AI

  • The big AI question: 1 model, or many models?
  • 2023 was the year of model diversity, and the beginnings of “open” models (Llama 2, Mistral, etc.)
  • Can we make it easier to contribute to an open model?
  • Can we create communities around models?
  • Can we reduce the costs of expensive parts of AI? 
  • How can businesses bring their knowledge to a model, without security concerns about their data? (options like RAG (retrieval augmented generation) exist)


FEEDBACK?