Chosen AI(s)
After a lot of deliberation, we have settled on 2 AI providers, although these are likely to change over time.
Claude
-
Overview: Claude is ideal for development via the CLI, and has several other features which have aligned with us.
-
Concerns: Decision making at the top level of the organisation, and the fact they still act as part of a perceived race, leaving safety of secondary importance.
-
Setup: We have a team license, so can add people to it as they need access, this also allows us some degreen of usage oversight.
Mistral
-
Overview: Mistral is understood to be one of the more ethical AI companies, they are also based in the EU, making the data storage side of things better than the big US companies.
-
Concerns: They still have military contracts with the French government.
-
Setup: Mistral is largely being used by our open web ui setup, which can be accessed via https://chat.agile.coop/
Last updated: