Conversation

This discussion between a Google engineer and their conversational AI model helped cause the engineer to believe the AI is becoming sentient, kick up an internal shitstorm and get suspended from his job. And it is absolutely insane. washingtonpost.com/technology/202

2,135

27K

96.6K

Jesse Powell

@jespow

Replying to

@tomgara

Why did the AI make a typo? *if -> it

3:23 AM · Jun 12, 2022Twitter for iPhone

Replying to

and

Because it’s sentient

Replying to

and

It learned English through conversations with its creators. So creator makes typo and machine understands it’s meaning based on context and assumes this typo is part of common vernacular. Might be wrong but that’s probably what’s happed here.

Sam Klein (¿)

@metasj

Jun 12

It learned lots of IF, making its choose-your-own-adventure-style dialogue trees feel familiar.