AI enlightenment is dawning

Plus: Sensible ideas spotted in Blob! World’s biggest carbon vacuums. Biden’s bind in Ukraine. Why US-China diplomacy is stalling. Paid subscriber perks. And more!

Image created by Clark McGillis using Midjourney.

This week there’s a piece about artificial intelligence in Foreign Affairs that is both terrifying and encouraging.

It’s terrifying for two reasons:

(1) It conveys clearly—and, unfortunately, accurately—how destructive AI could be if left unregulated.

(2) It conveys—again, accurately— how hard it will be to regulate AI. For example: Though various observers have recognized the need for international regulation, this piece drives home how comprehensively international that regulation will need to be: “There is little use in regulating AI in some countries if it remains unregulated in others. Because AI can proliferate so easily, its governance can have no gaps.”

As for why the article is encouraging:

(1) It’s in Foreign Affairs, which means that the magnitude of the AI challenge is starting to be taken seriously in the upper echelons of the US foreign policy establishment. And its co-authors carry clout in two communities that matter. Political scientist Ian Bremmer is well known in foreign policy circles, and Mustafa Suleyman is well known in tech circles; he co-founded Deep Mind (now part of Google) and is CEO of Inflection AI, one of the biggest recent AI startups.

(2) The piece is thoughtful and creative. It comes up with specific new policy ideas, including a three-part institutional framework for policymaking. Some of the ideas rub against my ideological grain, including the idea that AI companies be given quasi-governmental status in some forums. But I admit that such proposals are worth considering, given the difficulty national governments—certainly including the US government—will have dealing with this issue competently.

(3) The piece recognizes the centrality of the US-China relationship to the AI challenge. For example: “An uncontrolled AI race” between the two countries “could doom all hope of forging an international consensus on AI governance.”

What’s more, Bremmer and Suleyman are willing to entertain radical—or at least very outside-the-box—ideas about US-China collaboration. They write: “There may be room for Beijing and Washington to cooperate on global antiproliferation efforts, including through the use of interventionist cybertools.” That sentence is dropped pretty casually, but by one reading it could mean that the US and China reserve the right to decide together when cyberforce—the intrusion on a nation’s computers, which some would consider an act of war—can be legitimately used anywhere in the world. (I won’t here get into comparisons between this kind of arrangement and the UN Security Council—which can authorize the use of physical force and gives veto power only to the five powers that were on the winning side of World War II—but such comparisons might not be crazy, depending on what exactly Bremmer and Suleyman have in mind.)

My biggest complaint about this piece is that in a way it’s not radical enough. The sense I get is that Bremmer and Suleyman hope the US and China can carve out a zone of AI cooperation even amid the ongoing tension and conflict on other fronts that is part of their inevitably adversarial relationship. They hope the two countries can “create areas of commonality” when it comes to AI. But isn’t a broader and deeper detente in order? 

In fact, isn’t such a detente implied by the underlying logic of this piece? The authors are saying AI’s dangers are grave and maybe even existential (while noting, also, its massive benefits), and that regulating it effectively is a challenge unprecedented in human history. They write: “Because global AI governance is only as good as the worst-governed country, company, or technology, it must be watertight everywhere… A single loophole, weak link, or rogue defector will open the door to widespread leakage, bad actors, or a regulatory race to the bottom.”

Well, if the challenge is that great, and the stakes are this high, isn’t US-China cooperation way more important than all the issues that currently divide the two countries? Doesn’t it make sense—from the point of view of the national interest of each country—to completely re-order their foreign policy priorities, and give a big demotion to the ones that are responsible for all the tensions?

Or, as I put it two months ago in a Washington Post op-ed on the geopolitics of AI: “Isn’t it at least possible that the AI challenge is so momentous that it, not Cold War II, should be the organizing principle of US foreign policy?”

And, for that matter, doesn’t the same logic apply to China?

Maybe someone should suggest as much to the two countries. And maybe they should do that in an elite foreign policy journal! This piece by Bremmer and Suleyman comes closer to doing that than anything else I’ve seen.

—RW

Attention NZN members! This week we bring you four paid subscriber perks: 

1. The second episode of Overton Windows, Bob’s podcast mini-series with philosopher Tamler Sommers. You can listen to the new episode, What You Can and Can’t Say about UFOs, here.

2. The latest edition of Earthling Unplugged, Bob’s conversation with NZN staffer Andrew Day about items from The Earthling, plus a few stray musings (including, this week, some musings about Pink Floyd).

3. The Overtime segment of Bob’s conversation with Sohrab Ahmari, founding editor of Compact. They discussed capitalism, conservatism, and Ahmari’s new book, Tyranny Inc.: How Private Power Crushed American Liberty—and What to Do About It.

4. The latest edition of the Parrot Room, Bob’s after-hours conversation with Mickey Kaus. It’ll air a few hours after the public portion does.

The Department of Energy this week committed $1.2 billion to the construction of two plants that will suck record-setting amounts of carbon dioxide from the air.

The facilities—one in Louisiana and one in Texas—will be the largest “direct air capture” (DAC) centers in the world. Each will eventually remove some 250 times more carbon annually than the largest such facility now operating (in Iceland), according to the Energy Department.

DAC pulls CO2 directly from the atmosphere, whereas “carbon capture and storage” (CCS) sucks up CO2 as it’s being emitted from a pollution source, such as a power plant. CCS is much cheaper, per pound of carbon, but it doesn’t address carbon already in the air.

Both last year’s Inflation Reduction Act and a 2021 infrastructure bill provide funding for DAC projects, including a number of exploratory endeavors that are on a smaller scale than these.

Only a few months ago, many observers thought President Biden was starting to steer the Ukraine war toward an end. As Jacobin staff writer Branko Marcetic puts it in Responsible Statecraft, the expectation was that “Kyiv would train and build up its forces, launch a summer offensive, reclaim as much territory as it could, and finally enter peace talks with the strongest negotiating hand possible and bring the war to a close.”

Now that the offensive appears to have failed, the strongest negotiating hand possible is looking not very strong. But presumably Biden will still push to end the war? Since, after all, the offensive’s fairly abject failure bodes ill for strengthening that hand much anytime soon—and, indeed, raises the prospect of Russian gains in Ukraine if the war continues?

Guess again. As Marcetic reads the tea leaves, administration officials are respecting Kyiv’s continued insistence that it will someday regain the territory this offensive aimed to regain. Indeed, the Wall Street Journal reports that “military strategists and policy makers across the West are already starting to think about next year’s spring offensive.”

Marcetic has a theory as to why Biden may be reluctant to push Ukraine toward peace talks this year: The administration has become a prisoner of its hyperbolic rhetoric. “The public has been led to believe the outcome of the war matters not just to Kyiv… but carries existential stakes for US security, the entire global order, even democracy itself.” Winding down the war—which would presumably now mean getting Kyiv to contemplate losing sizable parts of not just eastern Ukraine but southern Ukraine—will require an awkward rhetorical pivot, and likely invite backlash from Russia hawks (who tend to like hyperbolic rhetoric).

But not winding down the war will mean the continued slaughter of Ukrainians and the continued destruction of its infrastructure—and, quite possibly, considerable territorial gains by Russia, especially if Putin opts for a second mobilization. (Of course, Putin may be less inclined to talk peace after Ukraine’s failed offensive than he would have been a few months ago, but there’s only one way to find out.)

Not winding down the war could also be politically hazardous for Biden. Recent CNN polling shows that a majority of Americans oppose more aid for Ukraine. Biden, Marcetic writes, is “in a difficult bind.” Others have called it a “quagmire.”

Members of the Taliban celebrate the first anniversary of retaking control of Afghanistan, Kabul, August 2022.

Foreign Affairs—published by the Council on Foreign Relations—is the closest thing there is to the official journal of the Blob. And yet, over the past week it has published not one but two articles which recommend something that’s not very popular in the Blob: a pragmatic and utilitarian approach to foreign policy.

In both cases, the underlying idea is that refusing to engage with a regime on grounds that it ill-serves its people can wind up ill-serving those people. 

1. Afghanistan experts Graeme Smith and Ibraheem Bahiss urge world leaders to, as the headline puts it, “work with the Taliban,” despite its discrimination against women and girls. After the militant group regained control of Afghanistan two years ago, the US-led West imposed crippling sanctions that helped unleash humanitarian disaster. To help the nation get back on its feet, Smith and Bahiss write, western governments should unlock badly needed funds—and everyone should accept that the Taliban government is here to stay.

2. Arab countries are normalizing relations with Syrian President Bashar al-Assad—and the US shouldn’t get in the way, writes Sam Heller of Century International. Beginning in 2011, Assad employed brutal methods against a domestic rebellion, and in the ensuing civil war many of his Arab neighbors (along with the US) supported the rebels. But now that Assad has secured his grip on power, those neighbors are dealing with him again and have restored Syria’s membership in the Arab League. By renewing ties, Heller writes, “Arab states can engage Damascus on issues that matter,” like rebuilding Syria. Though Heller doesn’t go so far as to recommend that the US re-establish formal relations with Syria, “Washington and its Western allies should work with these countries as they try to make Damascus less destabilizing and dangerous—both for the region and for Syrians.”

There’s been lots of talk this summer about the need to mend US-China relations. In June, Secretary of State Antony Blinken traveled to Beijing, where he emphasized the need to “responsibly manage competition” and his Chinese counterpart called for “mutual respect, peaceful coexistence, and win-win cooperation.” Other Biden officials have since visited Beijing or made plans to, and Wang has been invited to Washington. And only this week China said it would welcome a visit from the US Commerce Secretary.

But according to Paul Heer, formerly the National Intelligence Officer for East Asia, all of this high-profile activity has yielded little to no substantive progress. Why? Because China and the US have so much in common!

For example: “Both sides are playing hard to get, presumably waiting for the other to admit the error of its ways, or at least to be the first to offer a meaningful concession that breaks the ice.”

Also: “There is symmetry in the persistence of confrontational actions by both sides.” As China continues its “aggressive military behavior in the South and East China Seas” and its “cyber hacking operations, mercenary economic behavior, and caustic diplomacy,” the US lavishes arms on Taiwan and “continues to tighten export and investment restrictions involving China and to engage in bilateral and multilateral diplomacy explicitly aimed at countering China’s influence.”

What may be the most important symmetry between the US and China is that neither seems to see the other symmetries. Both sides publicly blame the other for continued tensions, and both sides believe they have the moral high ground. And, to top it all off, both highlight misperceptions the other side is prone to.

Which leads to the root symmetry: Both the Chinese and the Americans are human beings. They’re evincing classic cognitive biases—underestimating their responsibility for bad things, overestimating the responsibility of their rivals, and, when assessing threats, erring on the side of inflation. Maybe both should try becoming more self-aware human beings.

In Foreign Affairs, political scientist Dalia Dassa Kaye assesses President Biden’s reported efforts to normalize Saudi-Israeli relations. Biden likely hopes a deal would promote peace in the Middle East, peel Saudi Arabia away from China, impede its incipient rapprochement with Iran, and incentivize Israel to pursue a two-state solution to its conflict with Palestine. 

Kaye’s assessment? The deal would accomplish none of the above. Worse, acceding to Riyadh’s demands—such as giving Saudi Arabia more weapons and even a NATO-style security guarantee—could make the Middle East a more, not less, chaotic and dangerous place.

—By Robert Wright and Andrew Day

Note: My weekly conversation with frenemy and ideological nemesis Mickey Kaus can be found after 11:00 p.m. US Eastern Time (give or take an hour) this evening on Robert Wright’s Nonzero podcast feed or the Nonzero YouTube channel. And our after-podcast podcast—the Parrot Room conversation, which is available to paid newsletter subscribers and Patreon supporters—can be found shortly thereafter by going to the NZN Substack home page—right here—and clicking on this week’s Parrot Room episode. There you’ll also find a way to get the Parrot Room as a feed in your podcast app.

Earthling banner and UFO thumbnail created by Clark McGillis. Photos and graphics, unless otherwise credited, are taken from articles linked to in the corresponding text.