Will Tebbutt

> In practice, GP is always associated with some likelihoods, therefore always a latent variable model. This is not true, and it's not the case that incorporating one's likelihood in...

I'm also a bit confused now haha. > What I am suggesting is to place LatentGP inside AbstractGPs package. I don't have partcularly strong feelings either way here. Provided that...

> I am thinking that AbstractGPs can define the interface (abstract types and methods) for likelihoods, while LikelihoodFunctions provides the actual implementations. I agree that the likelihood interface in AbstractGPs...

@YingboMa the Cassette PR doesn't address Lyndon's question. It would be helpful to understand whether the long-term goal is to take the kind of SCT approach that Lyndon discusses above,...

Sure. So is the plan to have a single `Dual` type that can contain arbitrary data types?

> Dual can already contain arbitrary data types. The plan is to make Dual not visible. Well, yes, kind of, but it only makes sense with subtypes of `Real` at...

> With Dual, we know what functions to differentiate, so it is pretty simple to ignore functions like Is see. There's only a finite number of such functions though, and...

Great, it looks like we're on pretty much the same page then. Possibly our only difference of opinion is that I now can't see why you would ever want to...

I agree with all of the above. I like your `wrt` proposal a lot - it solves the problem in a much cleaner way than we're currently allowing (although not...

Another way to go about the equality problem is to avoid defining approximate equality on tangents entirely, and instead rely on 1. being able to add a tangent to a...