Data-centric AI: A new paradigm

Data plays a key role in AI development. Explore the transformations it has undergone in the realm of large-scale models and understand its influence in steering machine learning towards a data-centric AI paradigm.
Watch more:
Watch all of the Women in ML Symposium sessions → goo.gle/WiML23...
Subscribe to the TensorFlow channel → goo.gle/Tensor...
#WIMLsymposium

[@MaxM9000]

This data centric approach has been fundamental to AI from the beginning. We're only going back to these (very important) principles because of the diminishing returns of pretraining with more data than you need. Now the focus is back to more specialized tasks with these large pretrained models as a foundation.

[@RahulKharwadkar]

Very nicely presented how critical the data is for AI models

[@claudiaariasgonzalez6941]

Interesting

[@JorgeOrpinel]

Honest summary: Set up data sweatshops in Kenya.