Tried this for a legal use case and it was excellent, comparable to Opus in quality but much faster. AI is miles behind in law compared to coding: the output was similar to a law student intern. But coherent and directionally correct and beats starting from a blank sheet of paper. Impressed.
I'd like to use it for legal work too. Microsoft makes great hay about its ability to sandbox CoPilot's work and not train on or share company resources ("Look for the green checkmark."). It's largely for that reason that we've rolled out Copilot to most of the white collar positions in the company. Do you happen to know whether xAI has similar functionality?
xAI has that functionality in the business tier: https://x.ai/grok/business. It's got GPDR, HIPPA, etc. We are playing with Harvey, Legora, and Lucio, which can use various models. Harvey can use Grok: https://x.com/techdevnotes/status/2074956936701968652
Harvey is fairly impressive--it's the only one that seems to be built by people who know how LLMs work. :-/
How does Fable compare?
I haven’t tried it, didn’t have access at the time.