Why define these patterns?

Design patterns are solutions that are commonly repeated. Iterated upon in product designs over decades, to find what works.
The successful ones make up what we call best-practice.

For example:

What do you expect to happen when you click a dropdown?
This:
Favourite Beatle?
John
Paul
George
Ringo

OR

This:
Favourite Beatle?
John
Paul
George
Ringo
Obviously, the second one.

Over time & across many products, various patterns were tried option selection & this is what has become ubiquitouse & familiar.

*Of course native browser controls influence this particular example heavily.

Enter AI...

Bringing a new paradigm in user experience design.

Moving user inputs from:

Command based

Configure controls & options.

Intent based

I would like this outcome please.

Previously...

Software was designed & built on known inputs & outputs.
⤏ ⤏ ⤏ ⤏

But...

AI is a powerful probablistic, natural language processor. The output however is not... predictable.

While generated results are getting better & better, they could range between:

⤏ ⤏

This new experience places the onus users to critically review their generated results.

So the question then becomes...

What existing UX patterns are applicable & what new patterns will we need to deliver effective AI experiences?

AI UX Patterns

Is an attempt to define & explore these patterns in an abstracted format. To serve as a useful guide as our AI adoption grows & matures.