I’m trying to imagine what other interfaces could expose value from these existing AI tools value simply by making it easier and as low effort as possible.
Similar to how GitHub Copilot/Tabnine automatically suggests completions as unobtrusively as possible without requiring a lot of manual intervention. Just as an extension.
Once we have a google glasses type augmented reality this sort of light automated assistance can be fed automatically into our lives. Just by hearing or looking at something it could turn it into an automated prompts and be unobtrusively suggested or visually fed into our lives. Auto-translation of foreign languages is the obvious AI example with AR.
Maybe other interfaces such as direct MacOS integration the way the smart search box/Alfred style works or browser extensions.