In the relatively near future, you’ll be able to chat with a computer—without knowing it's a computer. I don’t mean automated-helpline type stuff. I mean, comparing hobbies and sharing your feelings about Strasbourg.
The best way to test if your friend is fake is to focus on very specific things. You can't ask a computer if they have feet and expect to trip them up. Too easy. No, you need to ask what it feels like to have feet. To be cutting one’s toe nails, and to see one fly off under a cabinet, or disappear into the bathroom rug.
A human will reference their guilt or shame. A computer will say it tracked the arc of the stray nail, and then incinerated the dirty outlaw with a laser eyeball.
If you ask a ‘virtual brain’ their favourite band, they will, of course, say Kraftwerk—another easy question. Try asking them what they imagine Elvis smelled like, or get them to draw Adele’s face on an egg.
I’m not saying computers won’t be able to perform these more nuanced tasks; but it will take them a couple of seconds, and they’ll make tell-tale beeping noises, as they process their ‘thoughts’—that's all the proof you need! Checkmate!
NB: If you’ve heard some beeping noises while reading the information above, then you’re probably a computer.
Robots, eh?