Silicon Valley should've been called Balloon Burg. America's tech industry often seems like it might pop under the slightest pressure. (Remember the dot-com crash from the early aughts?) Now, according to The New York Times, many of these companies are looking at looming U.S. Department of Commerce export restrictions on artificial intelligence like an inflatable animal would look at a porcupine.
Here's the problem: Congress voted in August to limit the export of "emerging and foundational technologies" to preserve U.S. national security interests. The New York Times said that a Commerce Department proposal would restrict the export of "several categories of AI-like computer vision, speech recognition, and natural language understanding" to countries the U.S. has sanctioned in the past.
Americans can voice their opinions about this proposal until January 10. Then it's up to the Commerce Department to determine what technologies and export destinations it wants to restrict. Naturally, tech companies have asked the government to go lightly on them, because they believe restricting the export of their AI tools would allow companies in affected countries to seize their local markets.
This entire back-and-forth is about as obvious as 99 luftballons--whoops, sorry, red balloons--at a wake. Of course, the U.S. wants to restrict AI exports to countries like Russia, Iran, and China. The first two have been repeatedly accused of using tech platforms to influence U.S. politics; the last one is America's enemy in an increasingly hostile trade war that could significantly impact tech products.
And of course, Silicon Valley opposes those restrictions. Tech companies often rely on regulatory oversights regarding privacy, competition, and other consumer protections in their quests to make even more money for their shareholders. (See: Facebook's data sharing, ride-hailing companies' flouting of traffic laws, countless anti-monopoly safeguards.) Why should that be any different with AI?
But The New York Times notes that AI is a bit different simply because it's such a collaborative effort. Most companies don't work on AI technologies on their own--they partner up with their rivals, advance open source projects, and involve people from around the world in their efforts. The tech is so new and has so much potential for growth that it's easier for competitors to work together.
This collaboration actually makes it difficult for the Commerce Department to enforce any export restrictions. You can't put the genie back in the balloon (or however that saying goes) so publicly available information can remain, well, available to the public. It would have to restrict AI in other ways, such as preventing hardware sales to affected countries or blocking their access to information.
The former would be bad news for companies that have prioritized the development of AI products in recent years like AMD, Intel, and Nvidia. The latter would further undermine the principle of collaboration upon which the internet was built, which could, in turn, slow the U.S. tech industry's AI research. Both would basically shut off these companies from vital revenue streams and sources of knowledge.
So even though this conflict is about as cliché as a balloon animal station at a kid's birthday party, it could still have serious repercussions for U.S. companies. The technologies lumped under "artificial intelligence" are varied, but they're all integral to these companies. Just look at the rise of voice assistants, natural language input, facial recognition, etc. in basically every tech product.
Stay on the Cutting Edge
Join the experts who read Tom's Hardware for the inside track on enthusiast PC tech news — and have for over 25 years. We'll send breaking news and in-depth reviews of CPUs, GPUs, AI, maker hardware and more straight to your inbox.
Nathaniel Mott is a freelance news and features writer for Tom's Hardware US, covering breaking news, security, and the silliest aspects of the tech industry.
Restrictions will make US companies less competitive. After January 10, we will see how many people understand it)Reply
The hypocrisy of Google not wanting to work with the US military for "ethical reasons" but wanting to work with PRC on AI projects and sanitized web browsers is astounding.Reply
I have mixed feelings about this, but it really comes down to the fact that the horse is already out of the barn. If we stopped exporting AI hardware to China, it would only be a minor speed bump for them, and would just give them more incentive to mature their own competitive solutions.Reply
And any market to which we deny exports would just become a ripe target for Chinese and other competitors. So, ultimately, it would just hurt US-based companies more than foreign governments or competition.
Heh, it's not Google's management that wanted to terminate work on military projects - their employees basically demanded it.21636947 said:The hypocrisy of Google not wanting to work with the US military for "ethical reasons" but wanting to work with PRC on AI projects and sanitized web browsers is astounding.
I'm sure a fair number of Google employees are also upset about working with PRC.
China is already developing their own solutions. Not understanding that is just wrong.Reply
China is quite advanced in AI, I'm not sure what the intent of this measure is when the U.S. cannot clearly call itself a leader in this field.Reply
Furthermore, denying collaboration would prevent companies like google valuable access to partnership and data from 1.4 billion people. There aren't as much wait time and regulatory approval in China and tech companies can benefit immensely from a market friendly to experimentation.
Lastly, think history has proven such protectionist thinking doesn't quite work. Look at the space station situation for example. China was denied access to the ISS and they simply built their own. Soon China will be the only country with a space station in orbit.