@cnn: Microsoft AI CEO Mustafa Suleyman speaks to CNN's Fareed Zakaria about why he thinks its time to “tone down the adversaial rhetoric” between the US and China when it comes to AI development. See the full interview Sunday at 10am ET on Fareed Zakaria GPS.

Sheena Hollywood Stylist :

The danger in America, is those AI are owned by the billionaires, they’re profit driven. Versus China it’s state controlled so they have a better monitoring system than America could ever be, and the Chinese are intelligent, they play chess ♟️

Jude Odhiambo :

The west is selling fear because they can't keep up

Nin :

Just leave China alone. 😅😅😅😅

Univarse ManOB 89 Q W lord :

🌹Soul i man not soleyman

Stephen :

I think a lot of fear around AI comes from putting completely different concepts into one box and simply calling all of it “AI.” The AI we actually have today can analyze information, generate language and images, write code, recognize patterns, plan actions and use tools, but those capabilities do not automatically prove consciousness, sentience, emotions, human-like free will or a desire to survive. AI means artificial intelligence; sentience is a separate question about whether something actually has subjective experiences. Then there’s AGI, which generally refers to a hypothetical level of broadly human-like general intelligence, although there isn’t even one universally accepted definition or test for when AGI has been achieved. Beyond that is ASI, artificial superintelligence, a hypothetical system substantially beyond humans across most important intellectual abilities. And then movies like Terminator, I, Robot and The Matrix add even more things together: extreme intelligence, consciousness, independent goals, long-term autonomy, physical control, access to infrastructure and sometimes weapons. Those things are not automatically the same thing. An AI can choose between actions or even write code without that proving it feels anything, and a non-sentient AI can still be dangerous if a human gives it dangerous objectives, tools or permissions. That’s why I don’t think the conversation should be “AI is evil” or “AI can never hurt anyone.” The better questions are: What can this particular system actually do? What evidence demonstrates that? What permissions does it have? Who gave it those permissions? What happens when it fails? And what safeguards can stop harmful actions? We should absolutely take real AI risks seriously, but we also shouldn’t turn predictions into facts or science-fiction scenarios into descriptions of current technology. And saying we’re “10% toward Skynet” doesn’t really mean anything unless somebody can define what 100% represents and show how they’re measuring it. 😂 Intelligence, sentience, autonomy, access and harmful intent are different questions. Show the evidence for each one instead of letting fear fill in the missing pieces.

Veins of Wisdom :

China is way smart

michaell3012 :

Don’t worry, be happy.

stella 💝💐 :

Patrick Ebert :

AI is becoming Amoral, indepth gaurdrails need to be put in place.

ام محمد 🤍🩷 :

I bet they fool him with AI clips and ratings and videos. That’s why he speaks like he really does! 🤣 🤦🏻‍♀️

lolavie :

368 days sober

𝒂𝒃𝒅𝒆𝒍𝒍𝒂𝒉𝒊 🕯𖣂ׁ :

yo

DrizzyJohnny :

First

Aurelia :

I think distance is nothing when two people enjoy the conversation. 😅

Lisa :

I’d happily trade a quiet night for this ✨

Katherine :

What’s something people underestimate? 👍

jon :

lol 😆

hydro :

yo😂

𝄞⨾💿✮˚.⋆ :

BLEHHHHHHH

Noah brooks :

fc

🌟FlyAwayFarAway🌟 :

Ai doesn’t know national boundaries. It is a global issues.

Lucas2022 :

If there is safety issue, China government won't ignore it just like the US. It is common sense.

Yashar Nafisi :

🥰🥰🥰

حسنین عباس :

😂😂😂

maiaspam :

💕💕💕