[OrangeMusic]

The ad with the grandmas is cute and funny, but from the first 20 seconds you can see that the voice annoyingly interrupts people while they are talking. It's almost as if it tries to reply too fast - faster than a real person would, and the results is that it replies while you're still talking.

Oh and there was also a small fail in the live translate demo: the grandma says "tell him that..." which the bot translates verbatim, whereas a real translator would of course understand that this is an aside not to be translated.

Well I guess at least I should be happy that they're transparent in their ads :)

[simongray]

They don't act the way humans would.

It's normal to start a sentence after the other person in the conversation pauses, especially if you're eager to say something (like the chat agent is), but a real human is socially aware and would in most cases immediately cease talking if the other person starts another sentence at the same time. Humans do this so much in their conversations that the other person doesn't really register the interruption at all.

[rhet0rica]

Encoded in its design are the biases of its designers. It acts like a nervous twenty-something from San Francisco.

[OrangeMusic]

Exactly. If they have a setting somewhere for the minimum duration of a pause in the user's speech before starting replying, they should just crank it up.

In this other add [1], it's even worse, the user has to say "There's more to the story though" and almost looks annoyed + the agent overreacts to _that_ with a weird "Aaaaahh!".

Oh well... I'm sure they'll get it right eventually.

[1] https://openai.com/index/introducing-gpt-live/?video=1208099...

[ndkap]

Maybe they are trying to outmatch the world's worst translator: https://www.youtube.com/watch?v=foT9rsHmS24