China's Leaders Soften Their Stance on AI, Say They Will Be Sharing Their Findings With Other Countries (technologyreview.com) 45
China might be at loggerheads with the United States over trade, but it is calling for a friendlier approach to the development of artificial intelligence. From a report: Speaking at the World Artificial Intelligence Conference in Shanghai this week, China's vice premier, Liu He, said that AI would depend heavily on international cooperation. "We're hoping that all countries, as members of the global village, will be inclusive and support each other so that we can respond to the double-edged-sword effect of new technologies," He said through a translator. "AI represents a new era. Cross-national and cross-discipline cooperation is inevitable."
President Xi Jinping delivered a similar message in a letter presented at the same conference. Xi said that China would "share results with other countries in the field of artificial intelligence." He also called for collaboration between nations on AI topics such as ethics, law, governance, and security. This new, softer approach to artificial intelligence comes just over a year after the Chinese government announced an ambitious and aggressive AI plan. This blueprint called for Chinese AI researchers to lead the world by 2030, and for domestic companies to build an industry worth more than $150 billion. China's tech industry has already embraced machine learning and AI at an impressive rate.
President Xi Jinping delivered a similar message in a letter presented at the same conference. Xi said that China would "share results with other countries in the field of artificial intelligence." He also called for collaboration between nations on AI topics such as ethics, law, governance, and security. This new, softer approach to artificial intelligence comes just over a year after the Chinese government announced an ambitious and aggressive AI plan. This blueprint called for Chinese AI researchers to lead the world by 2030, and for domestic companies to build an industry worth more than $150 billion. China's tech industry has already embraced machine learning and AI at an impressive rate.
Couldn't they just steal AI tech? (Score:4, Insightful)
Stealing intellectual property has worked for them in every other industry. Why is AI any different?
Re: (Score:3)
That wasn't the issue this article was about. There was extensive international coordination between the rest of the world on AI to try to prevent the sorts of programs like the fictional Skynet, this included extensive discussion on arming AI and various other initiatives. Believe it or not the US is opposed to autonomous weapons that can make their own firing decisions.
The Chinese on the other hand rebuffed these offers and presented a front that they believed the weaponization of AI was imminent and they
No, they will not (Score:5, Insightful)
China has been lying/cheating on their building new coal plants, and how much coal they use.
They lie constantly about what their military is up to. The new man-made islands were NOT going to be militarized. Remember? Yeah. Only an idiot or a paid troll of China would claim otherwise.
Re: (Score:1, Interesting)
Shouldn't that be CUD? (Score:2)
Skynet shill to show up and spread FUD.
Since they are claiming there's nothing to worry about from AI because it is "useless", shouldn't that be spreading Calm, Uncertainty, and Doubt?
Re: (Score:2)
Calm, Certainty, Composure, and Peace - CCCP - uh oh...
Re: (Score:1)
Which means they will be sharing almost all of it since they have figured out that at a fundamental level there is little or nothing that is new and valuable in the latest "AI revolution".
Re: (Score:2)
Look at Boeing's 777. It was designed with Genetic Algorithms and made it the most efficient wing by a LONG shot (for when it was done).
Now, AI is being used for car driving, determining medical conditions (and better than most docs), etc. From a military POV, car driving could be tank and ship control. Dia
Re: (Score:2)
Xi said that China would "share results with other countries in the field of artificial intelligence.
Sounds like very careful wording to me. They won't share anything of value, just "results".
Did China take your coal job? (Score:2)
Re: Heâ(TM)s full of it (Score:1)
China agrees to share AI findings... (Score:3)
Sans serif fonts (Score:5, Funny)
Will anyone know? (Score:2)
if no one truly understands the AI algorithms, then how is it secure for anyone to be taking AI algorithms from nation-states?
It would be fairly simple to drop in code meant to snoop, or otherwise degrade performance in some way or another. No one would be wiser.
--
"Ollie ollie in come free" - Jan Brady
I don't think you understand how AI works now... (Score:2)
It would be fairly simple to drop in code meant to snoop
Sharing AI information would be sharing models and approaches to building/training networks, there's not really a way to put in any code.
In theory you could train a complex model to tweak the results in a way no-one would expect - like maybe weighting Chinese-looking faces to appear more innocent in general. But most places would be using their own training sets, it's more the setup and layout of models that is the kind of thing that would be shared i
Re: (Score:2)
2) We understand how it works, they same way it "worked" in the 1960s. See 1)
Re: (Score:2)
No, machine learning is not the way it "worked" in the 1960's. [arxiv.org]
Machine learning and AI is also a distinction without a difference. AI comes in categories not just limited to "general".
Chinese Restaurant Compliment Analogy (Score:2)
Makes Sense (Score:2)
Duh! China is the beta tester (Score:1)
Re: (Score:2)
"safety and security" is a good catch all, to be sure, but then you have to deal with all the people who love repeating the quote of some old dead guy who was involved in some kind of declaration or bill of something that's now completely irrelevant anyway.
Nah, just stamp "ADL" or "SPLC" on it [youtube.com], then it's above criticism and reproach. Unless you're a hateful bigot, of course, and who wants to be accused of being one of those. Then once it has "trusted flagger" status across the board it can enforce "commun
The way I hear it (Score:2)
On the consumer side, they are currently testing an Entertainment Droid, version 208 in their country. Again, whether that model will ship to the US for examination, or the next version, we'll just have to wait and see.
It is inevitable (Score:2)
"AI represents a new era. Cross-national and cross-discipline cooperation is inevitable." Inevitable AI: someone said something like that:
https://www.youtube.com/watch?... [youtube.com] @1:44