By what year will an AI be able to competently walk consumers through moderate time horizon home improvement tasks? Some tasks (suggested by Claude):

• Install a dedicated outlet for an EV charger

• Replace a shower valve

• Hang a pre-hung interior door

• Regrout a shower

Caveats for resolution criteria:

• This task list isn't meant to be exhaustive.

• "Walk consumers through" means the consumer needs no additional help, and task success rate is high.

• The physical equivalent of vibecoding does not count: If some people are vibehandymanning that doesn't force a NO: as long as some people aren't this resolves YES.

• I will resolve YES if it can only assist people with some existing skills in these areas: this is about supplying missing procedural knowledge, not teaching someone who's never touched a screwdriver to rewire their house.

• If it can only assist people with professional experience that doesn't count.

I'll need either a number of trustworthy people to confirm that they do this, or for it to be broadly known that people do this regularly. I will also want at least some tradespeople to confirm that the work isn't all garbage (but do not require that this be consensus among the trades).