Back in April, AI Hub launched as a beta but now the search giant is giving the hosted AI repository a major overhaul which includes a new home page, machine learning taxonomy and additional content as well as the ability to favorite notebooks, models and other assets.
While these new features will improve AI Hub, the sharing features will probably have a much larger impact on data science workflows. G Suite (opens in new tab) users will likely notice some similarities between the sharing, permission and collaboration features that have been added to AI Hub.
- Alphabet's cybersecurity firm Chronicle joins Google Cloud
- Entrepreneurs not threatened by AI and automation
- Intel reveals first AI chips
By adding a familiar sharing interface to AI Hub, data scientists and developers will have an easier time collaborating and now they will be able to share notebooks, trained machine learning models and KubeFlow pipelines with their colleagues and peers.
As part of Google Cloud's new update to AI Hub, users that are logged in will be able to access recent shared private assets and content and this should lead to faster model building.
Groups of users will now be able to share their work with colleagues by giving them the ability to either edit or view their work. However, users with permission to view one's work will be able to fork an asset by either downloading or opening a copy but they won't have the ability to edit or change the version originally shared on AI Hub.
Google Cloud is also giving AI Hub users the ability to copy and share the URLs of public assets on social media, once again to help facilitate greater collaboration.
Training and deploying machine learning models is difficult work but hopefully Google's update to AI Hub can help users work together to ease some of the burden.
Via ZDNet (opens in new tab)