Later On

A blog written for those whose interests more or less match mine.

How Autocracy Comes to America: Big Tech and National Security

leave a comment »

Matt Stoller writes in Big:

Today I’m going to write about a very dangerous theme floating around in military and big tech circles, which is that big American tech monopolies are good for national security and should be weaponized and controlled explicitly by the American national security apparatus.

The relationship between corporate power and global geopolitics frames the historical debate over antitrust, as I discuss in Goliath: The Hundred Year War Between Monopoly Power and Democracy. (Buy it. Yeah, I’m not going to be subtle.) The argument has been with us since before World War One, and is with us today. Last week, for instance, Mark Zuckerberg gave a speech on free expression, and yesterday he testified to the House Financial Services Committee on the need for Libra. Both times he implied that Facebook was essential to protect us from China.

What is interesting is how certain parts of the national security world may not be so averse to how Zuckerberg thinks, even if they don’t trust him specifically.

And so it is back to that debate we must go.

Incidentally, I was on CNBC yesterday to talk about Zuckerberg, and explain why I think Libra isn’t going to happen and is a “crazy idea.”

And now…

National Security Is the Last Resort of Monopoly Scoundrels

In his speech last week and in his testimony yesterday, Mark Zuckerberg put a choice to policymakers. Pick between Facebook’s domination, or China’s. “While we debate these issues,” he said, “the rest of the world isn’t waiting.” He explained this was particularly the case with his new currency Libra. “China is moving quickly” on its digital currency, and while Libra could “extend America’s financial leadership as well as our democratic values and oversight around the world,” that would only be the case if regulators allowed such innovation. The threat of Chinese dominance was implied.

As part of his narrative, Zuckerberg emphasized the importance of democratic values embedded in the American platform of Facebook. It’s easy to be skeptical of his framework, considering that Zuckerberg was so eager to get into China that he apparently asked Xi Jinping to name his child, blurbed a book of Xi Jinping’s speeches, and made his employees read Chinese propaganda. In June, when Facebook announced Libra, political leaders and regulators gave it a giant thumbs down. In July, during the first set of hearings, Facebook was embarrassed. Zuckerberg wasn’t persuasive yesterday either; political leaders were deeply skeptical, because Zuckerberg has been deceptive about his view of China as late as last week.

Less important than Libra is the overall “national champion” argument that Zuckerberg is making. Proponents of national champions believe big is powerful, and that national power is a function of the scale that a nation’s corporations can bring to bear on world markets and security regimes. In many ways, Zuckerberg is the worst possible messenger for this theory. There are in fact much better proponents, like former intelligence and DOD official Jon Bateman of the Carnegie Endowment for International Peace.

Yesterday, Bateman penned an op-ed in the Wall Street Journal on the strategic importance of big tech. “Despite their faults, tech companies contribute directly to American military and intelligence operations,” he wrote. “Their titanic scale can itself be an asset.” I’ve spoken with multiple important political leaders who believe this. It’s a seductive view, one based on the idea that wise technocrats need large amounts of centralized power to keep our systems stable and our society under control. Bateman also argued, interestingly, that big tech, by working with China and enabling bad behavior domestically, also weakens national security. He tries to thread a path, by essentially proposing a national security veto over antitrust cases against national champion-style big tech companies.

We’ve seen this argument before.

The idea of what is called a ‘national champion’ was and is a foundational element of big business conservatism and state socialism, and has been for over a hundred years. Teddy Roosevelt, who is mistakenly understood as a great trustbuster instead of the supporter of monopolies he was, used this quote in 1912 in his call to repeal the antitrust laws, that if we “do not allow cooperation, we shall be defeated in the world’s markets.”

It is what the monopolists at Alcoa (and their allies in the military) thought before World War II when they argued their engineering system of aluminum fabrication was a marvel of ingenuity that could not and should not have competition. It is what Secretary of War Henry Stimson believed during that war when he managed to tamp down antitrust cases against domestic monopolists.

And the frame, though discredited, continued after the war. In the late 1960s, U.S. large banks like National City (Citibank) argued that they needed to get our from under regulatory restrictions so they could promote American values globally. In the 1970s national security officials freaked out about Japan’s state sponsored planning for its technology giants, and hated that antitrust enforcers sued AT&T and IBM. How dare we go after the metaphorical Yankees of technology?!?

Under Reagan, the Pentagon tried to stop Bill Baxter from breaking up AT&T.

And yet there aren’t cases when antitrust action undermined collective security. The antitrust suit against Alcoa helped build a massive aluminum industry and the “arsenal of democracy” against the Axis. And the suits against IBM and AT&T resulted in the explosion of innovation in the 1980s and 1990s which happened in the U.S. and not Japan.

But since we decided to forego our heritage, it’s time to rebut this veneration of bigness, yet again.

The Case Against Antitrust

Bateman makes two basic arguments against antitrust action. First, economies of scale are necessary for modern technology. Cloud computing for instance requires large amounts of investment, and breaking up the cloud giants like Microsoft or Amazon would jeopardize this technology. Second, economies of scale preserve American market share and thus national security reach; smaller companies are just less competitive. It would undermine the ‘network effects” so critical to big tech success.

Bateman’s arguments are basically what big tech leaders think as well. Zuckerberg talks about the resources he has to fund national priorities, and Brad Smith of Microsoft in his book Tools and Weapons basically proposed a fusion of big tech and the state for the same reasons.

So let’s take them on.

Economies of Scale

Here’s Bateman:

Consider cloud computing. The Defense Department is planning a massive global cloud called JEDI. Unlike corporate clouds, the “war cloud” must support life-or-death missions on austere battlefields despite virtual or physical onslaughts…

Vast resources were needed to fund global networks of hardened data centers linked by undersea cables. The U.S. military’s unique demands required companies of unique scale. Yet one JEDI bidder faces a concerted breakup campaign (Amazon), and the other was nearly dissolved in 2001 (Microsoft).

There are many problems with this argument, the main one being that it doesn’t reflect the real experience of actual operations. The National Security Agency has its own government cloud. Years ago, the CIA went to Amazon for its cloud computing needs; reading between the lines, this article suggests the experience of being forced into dependency on AWS was problematic. The CIA is now trying to move to a multi-provider cloud. In general, while it’s true that economies of scale exist, they are almost always overrated.

More fundamentally, Bateman conflates technical economies of scale with financial conglomeration. There’s no reason AWS has to be part of Amazon or Azure part of Microsoft. What does book selling have to do with cloud computing? Nothing. In fact, there’s a fairly good argument AWS is less successful than it otherwise would be because it’s part of Amazon instead of an independent business. There’s also the very obvious problem of dependencies; Bezos is highly dependent on China through his retail arm, which gives China leverage over his cloud business.

There are some elements of Amazon or Microsoft that should technically stay together, but these companies are largely giant conglomerates whose divisions have no actual relationship with one another beyond a common hierarchy, legal status, and branding. Financial links are not technical links, but Bateman doesn’t distinguish between easily severable legal relationships and cemented together software-linked data centers.

Global Competitiveness and Big Tech

Bateman also offers this argument.

Splitting up Big Tech would reduce its intelligence value. First, smaller companies would lose global market share to foreign rivals such as Alibaba or Baidu, which can ignore FISA. Small U.S. sites can’t leverage the “network effect,” a gravitational force that helps large sites stay dominant. Intelligence collected from small sites would also be less useful. They see only narrow slices of online activity, whereas tech giants track users across sprawling internet ecosystems. Dismantling these ecosystems would put greater burden on intelligence agencies to “connect the dots” of potential threats.

This argument that big tech would lose market share if it were split up is 100% backwards. We venerate engineers in garages because that’s where good ideas come from, not from entrenched interests. Google and Facebook are less competitive because they are big, not more competitive. That’s why TikTok is gaining, because Facebook hasn’t improved its product in five years and it has killed all attempts at investment in social media.

Bateman also mischaracterizes network effects. A network effect is what happens when an additional user of a network adds value to a network in a geometric curve. A phone network with only one user has no value (no connections), with two users it has value because those two can call each other (one connection), three users can call each other (six connections), and so on and so forth.

But Bateman gets the dynamics of network effects and big tech wrong. There is a network effect for Instagram and one for Facebook, but there isn’t one for the Facebook-Instagram financial conglomerate. And again, network effects are also a result of political choices, not just physical ones; we’ve allowed Facebook to keep its network closed, but force AT&T to let Verizon phones connect in. Even the individual network effects are a function of a failure to organize interoperability mandates.

Bateman thinks that because Google and Facebook are big and powerful now, they will be innovative going forward. The opposite is likely true. They are like GM and Ford in the 1960s, doing extremely well because of market power and scoffing at foreign competition that destroyed them just a few years later. Or perhaps they are like Boeing in 2018, a company that passed $100B in revenue for the first time ever. It was on top of the world. Today, it is… not.

(And that’s just the commercial side of the argument. Without going into too much detail, let’s just say I’m deeply skeptical that more surveillance capacity means better intelligence. I suspect that more intelligent intelligence means better intelligence.)

A National Security Veto for Antitrust Over Big Tech?

Bateman isn’t just boosting big tech, he also brings some skepticism to the table. He writes, for instance, that big tech has “given terrorist groups a new way to radicalize recruits and enabled foreign governments to covertly influence the general public. These actors exploit algorithms that make sensational content go viral—a result of decisions by Facebook and YouTube to maximize user engagement.”

These companies have also subverted American interests vis-a-vis China. As Bateman writes, “‘AI and its benefits have no borders,’ as one Google executive put it—do not justify Google’s establishment of a Chinese research center. Dubious assertions and discredited theories can’t excuse Apple’s censorship of Chinese apps.” This presents a dilemma, in Bateman’s view. Bigness is good but big tech has misused their power. So what is the remedy?

His proposal is to incorporate national security officials to advise antitrust enforcers and Congress, especially where break-ups are concerned, as well as to reduce authority for state-level antitrust enforcers. Bateman, in other words, seeks to fuse the power of big tech with the national security apparatus, with the threat of antitrust as a stick for the intelligence world to force Google/Apple/Facebook/Amazon to do what they want. I’ve never liked the outrageous amounts of secrecy in our government, but an explicit veto over how we the people get to structure our markets should terrify all of us. This would engender a murky world of secretive government officials walling off political debate over fundamental questions about technology, commerce, communications and politics behind classified barriers. This framework seems to be, to protect us from China, we must become China.

Bateman’s piece is useful because it . . .

Continue reading.

Written by Leisureguy

24 October 2019 at 7:47 pm

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

%d bloggers like this: