Sentient Google Chatbox

[MichaelW2]

Google engineer Blake Lemoine has leaked/shared conversations he has had with a Google AI chatbot, or synthetic colleague. He claims that the bot is self aware, has its own likes and desires, and fears being switched off.
Is this the beginning of a beautiful friendship or have we welcomed our new Google overlords.
Sentient AI. Yae or Nay.

 

[Alex321]

He can claim what he likes, but that doesn't make it true.

You can of course code in behaviour which makes it seem like it is afraid of being turned off, and code in likes and dislikes, even including enough pseudo-randomness to make it appear to be decided by the AI.

But we are quite a long way from genuine self-awareness in AI yet. It will undoubtedly eventually come, but probably not this decade.

 

[DRM]

Brace yourself it’s starting!!

 

[keithmac]

Joking asside there is a real Skynet..

 

[Joey Shabadoo]

I'm sorry Dave, I can't do that.

 

[Electric_Andy]

I'll believe it when I see it. Google and others are extremely clever at harvesting vast amounts of info from us all, and performing things based on that. But I still think true AI is huundreds of years off. It is fairly simple to present the appearance of AI using algorithms, but if a lifelike human were built tomorrow, it would still be running commands, telling another AI chip what the conditions are, and reacting to pre-programmed responses

 

[Alex321]

I'll believe it when I see it. Google and others are extremely clever at harvesting vast amounts of info from us all, and performing things based on that. But I still think true AI is huundreds of years off. It is fairly simple to present the appearance of AI using algorithms, but if a lifelike human were built tomorrow, it would still be running commands, telling another AI chip what the conditions are, and reacting to pre-programmed responses

Depends on just how you define "true AI" I suppose, but I don't think it is hundreds of years away.

I think more likely 15-30 years away, considering the pace of change just during my working career.

 

[MichaelW2]

I'll believe it when I see it. Google and others are extremely clever at harvesting vast amounts of info from us all, and performing things based on that. But I still think true AI is huundreds of years off. It is fairly simple to present the appearance of AI using algorithms, but if a lifelike human were built tomorrow, it would still be running commands, telling another AI chip what the conditions are, and reacting to pre-programmed responses

Ai would not neccessarily be human like but would be a learning machine, not limitted to pre programned responses. The Go playing machine created original moves not pre programmed copies . Self awareness would be a component of a high intelligence AI.

 

[Beebo]

To add fuel to any conspiracy theories Google have sacked the whistle blower.

https://www.bbc.co.uk/news/technology-62275326

 

[craigwend]

It's when people start to worship it...