Microsoft CEO pretty sure he can Keep AI from Escaping Human Control

[vtqhtr413]

While some employees at the company are sobbing tears of joy, Microsoft's CEO Satya Nadella took the opportunity to address concerns over safely developing its AI in an interview with CBS News. First, he defended the decision to release the AI to the public, even if it's still full of kinks. "The only way for any new technology to be really perfected is to be in the market with real human feedback," Nadella said. "If anything, in particular with AI, it has to get aligned with human preferences, both personally and societally, in terms of the norms and yes, we will have many, many mechanisms in place to ensure that nothing biased, nothing harmful gets generated," he added. At the interviewer's prompting, the Microsoft CEO acknowledged that the ominous possibility of an AI going rogue and turning against humanity is a valid concern, "Runaway AI — if it happens — it's a real problem," Nadella admitted.

Microsoft CEO Pretty Sure He Can Keep AI From Escaping Human Control

Microsoft CEO Satya Nadella took time to address concerns over AI safety in an interview with CBS News, and he seems confident about keeping AIs reined in.

futurism.com

 

[blackice]

Nope, humans are always trying to find ways to destroy themselves. Add this to the list.

 

[Zero Knowledge]

in terms of the norms and yes, we will have many, many mechanisms in place to ensure that nothing biased, nothing harmful gets generated

Right.... I just don't see humans being able to stop bias and harmful content in AI. There will always be bad actors who use the technology for no good.

 

[ForgottenSeer 98186]

Microsoft CEO Pretty Sure He Can Keep AI From Escaping Human Control

Microsoft CEO Satya Nadella took time to address concerns over AI safety in an interview with CBS News, and he seems confident about keeping AIs reined in.

futurism.com

Just wait until types and forms of AI have to be made illegal because they represent a real threat or danger. Oh yes. This scenario is coming.

By then it will be far too little, way too late. Like most things humans do.

 

[show-Zi]

AI is similar to religion.
Can humanity, which has no common ideology, create an AI that can be shared?

 

[Zero Knowledge]

Can humanity, which has no common ideology, create an AI that can be shared?

And that's why I don't think AI or machine learning will never be on the same level as human thinking. Sure, you can model and implement code for logic, maths, language and rational thinking/thought, but can you model human creativity or emotion? Can you teach an algorithm anger or sadness or what spirituality or faith is? Can you teach a computer what love is? AI will bring huge advances no doubt, but we will still need humans for the time being to manage it and we will still need people because as a society we need each for company, comfort, sex and emotion.

The matrix was ground-breaking, the second and third were rubbish but they had some interesting ideas. The reason why The Matrix failed in the end and a truce was called is that it just could not understand what love is and what it means to us humans and what we will do for love (the little girl is key, you will see it now) or what human emotion was. I think that is somewhere we will end up, a stalemate between humans and AI where we both need each other to survive.