all at orders of magnitude lower computational cost.
I presume disregarding the 'more than one million hours' of training in the order of magnitude comparison?
4 Matching Annotations
- Nov 2025
-
www.microsoft.com www.microsoft.com
-
-
Aurora outperforms operational forecasts in predicting air quality, ocean waves, tropical cyclone tracks and high-resolution weather
Aurora said to extrapolate for air quality, bathymetry, cyclone tracking and 'high resolution' weather (I suspect they mean the opposite, check in paper).
-
Aurora is a machine learning model that can predict atmospheric variables, such as temperature. It is a foundation model, which means that it was first generally trained on a lot of data and then can be adapted to specialized atmospheric forecasting tasks with relatively little data. We provide four such specialized versions: one for medium-resolution weather prediction, one for high-resolution weather prediction, one for air pollution prediction, and one for ocean wave prediction.
MS created foundation model Aurora, trained on over '1 million hours of diverse geophysical data' (they mean 1 million compute hours??), to use to predict atmospheric variables (temp) in August 2024.
-
- Apr 2024
-
www.rabbit.tech www.rabbit.tech
-
LAM is a new type of foundation model that understands human intentions on computers. with LAM, rabbit OS understands what you say and gets things done.
The Rabbit people say their LAM is a new type of foundation model, to be able to deduce user intention and decided on actions. Sounds like the cli tool I tried, but cutting human out of the loop to approve certain steps. Need to see their research what they mean by 'new foundation model'
Tags
Annotators
URL
-