Slashdot Mirror
Department of Commerce Could Be the First US Entity To Broadly Regulate an Aspect of AI (qz.com)
Dave Gershgorn and Max de Haldevang, writing for Quartz: Artificial intelligence technology has the capability to be the most impactful software advance in history and the US government has no idea how to properly regulate it. The US does know that it doesn't want other countries using its own AI against it. A new proposal published this week by the Department of Commerce lists wide areas of AI software [PDF] that could potentially require a license to sell to certain countries. These categories are as broad as "computer vision" and "natural language processing." It also lists military-specific products like adaptive camouflage and surveillance technology.
The small number of countries these regulations would target includes a big name in AI: China. Donald Trump, who has placed tariffs on hundreds of billions of dollars of Chinese goods as part of a simmering trade war, has long railed against China's alleged theft of intellectual property. This proposal looks like a warning from US officials, just as Chinese president Xi Jinping aims to boost AI in his own country. "This is intended to be a shot across the bow, directed specifically at Beijing, in an attempt to flex their muscles on just how broad these restrictions could be," says R. David Edelman, a former adviser to president Barack Obama who leads research on technology and public policy issues at the Massachusetts Institute of Technology.
The small number of countries these regulations would target includes a big name in AI: China. Donald Trump, who has placed tariffs on hundreds of billions of dollars of Chinese goods as part of a simmering trade war, has long railed against China's alleged theft of intellectual property. This proposal looks like a warning from US officials, just as Chinese president Xi Jinping aims to boost AI in his own country. "This is intended to be a shot across the bow, directed specifically at Beijing, in an attempt to flex their muscles on just how broad these restrictions could be," says R. David Edelman, a former adviser to president Barack Obama who leads research on technology and public policy issues at the Massachusetts Institute of Technology.
AI isn't very smart, its only redeeming value, is it is relentlessness in its calculations. While we as humans can only work on a problem for a few minutes a computer can sit there for years working out a problem, and not get distracted.
We as humans find shortcuts to allow us to adapt much more quickly to a changing environment. An AI system may take 6 months to learn how to make a Robot Walk on a flat plain. While it would take a human over a year to walk on a flat plain. However with that AI System, you then give new factors such as a Hill, or gravel or Ice, It will take another 6 months to figure it out. While the human will be able to adapt to the change within minutes, even without past experience.
AI is good for simple jobs, that we as humans really don't like to do too much. Just because they bore us.
If something is so important that you feel the need to post it on the internet... It probably isn't that important.
I regularly do work for overseas clients using open source imaging libraries. Libraries that are _already_ available in those countries.
So, a company in China could hire local developers to download and use the same NN (Neural Network) libraries I would use, and it would be legal. But if they hire me, overseas, to use the same libraries, that would be regulated as a technology exchange and possibly not allowed? That HURTS commerce. In an open source world, this sort of thing is ridiculous. It limits american companies by preventing them from competing internationally with already existing technologies - image classification, for example, is a Widely discussed topic, and many of the original theories and the techniques we use came from people outside the US - but thanks to international scientific sharing and open source, we - like everyone else - get to use and benefit from these techniques.
And these technologies are everywhere - they are not militarily purposed/used. These days - Our cellphones use NN to determine what sort of 'scene' we point the camera at and adjust the exposure/brightness/contrast appropriately - whether for a selfie or a food shot. Snapchat and many apps use facial recognition that uses Neural Networks - AI image classification and recognition - to implement filters. So, you're telling me a company in China can develop a new fox-face filter for their snapchat-alike app using the same techniques and libraries we can all download, but it would be bad/wrong for an american company to make such a filter and sell it on the Chinese market? No, if such regulation were to be used, the stipulations must be much more specific than just including image classification and other broad AI techniques. _OF course_ we don't want American developers making weapons for foreign regimes - but to limit ALL uses of these technologies is asinine and bad for our tech sector as it cuts off a broad swath of the global market - a swath that _already_ has access to these techniques.
The dept of commerce document linked in the article lists the following contact information:
ADDRESSES: You may submit comments
through either of the following:
Federal eRulemaking Portal: http:///
www.regulations.gov. The identification
number for this rulemaking is BIS 2018–
0024.
Address: By mail or delivery to
Regulatory Policy Division, Bureau of
Industry and Security, U.S. Department
of Commerce, Room 2099B, 14th Street
and Pennsylvania Avenue NW,
Washington, DC 20230. Refer to RIN
0694–AH61.
FOR FURTHER INFORMATION CONTACT:
Kirsten Mortimer, Office of National
Security and Technology Transfer
Controls, Bureau of Industry and
Security, Department of Commerce.
Phone: (202) 482–0092; Fax (202) 482–
3355; Email: Kirsten.Mortimer AT
bis.doc.gov.
man is machine