Arguing the Iraq War with ChatGPT
For the past year or so, AI has been a hammer in search of a nail. People are very impressed by what it can do, but they're not sure to what uses it can be put.
I have an idea of how it can improve public discourse, and I'm going to test drive it in this post.
In the early age of blogging, there was a genre called the “fisk.” To “fisk” something was to write a hasty, hard-hitting rebuttal of some recently written piece you regarded as ridiculous and dangerous. A fisk would feature lots of direct quotations, often uncharitably interpreted and probably taken out of context, as the author tried to put something beyond the pale, to enforce a consensus against it.
Fisks tended to be rather unfair to their targets, but were often lucid and principled, and fostered principled agreement among sympathetic readers. Sometimes it takes a polemical duel to give hard and high matters of general principle enough urgency to make people express themselves clearly and make audacious commitments. The fisk tended to seize a high ground of moral clarity in order to anathematize its target. Fisk authors could then be held to those standards of moral clarity by others when they were tempted to compromise. Fisking could leverage acrimonious public debates to fuel the articulation and enforcement of good principles. But the natural response to a fisk is an angry “That's not what I meant and you know it!” – and the conversation to which that leads is not particularly valuable.
The fisk per se went out of fashion, and while no doubt it has its successors in some ways (the internet isn't exactly insufficiently polemical), I think the principled pugnacity characteristic of the fisk, the use of provocation by stupidity or moral vacuity as a stimulus to the articulation of principles, is something of a lost art. It was lost partly for the sake of fairness and civility, which are much needed, but the loss was real.
And here's where I think AI can help. We can fisk AI! Why not use AI as a kind of punching bag for public arguments?
Sometimes being fair to one's interlocutor is a distraction from expressing valuable thoughts which an interlocutor has provoked. AI is an interlocutor to whom there is no duty to be fair. You can get civility out of the way, and get all the benefit of lucid polemic.
Now, we care what other people think in a way that we don't care what AI thinks. If someone is flagrantly, indefensibly wrong, it may seem morally urgent to refute them. It's not morally urgent to refute a machine.
But the remarkable thing about ChatGPT is that it's remarkably human-like, and it achieves this by processing and mimicking huge quantities of online material. It follows that when AI is prompted to express an opinion, the opinion it comes up with usually is, and is readily recognizable as, a lot of people's real opinion, more or less.
AI, in short, can serve as the spokesman for groupthink. And in fisking AI, you fisk groupthink. That was always what the fisk was really about, and that's why the defensiveness of an unfairly treated author was an unfortunate detour. Maybe the author didn't mean that, but something they wrote was representative of a widespread error. The fisker caught them saying it and rebutted them fiercely to deter other people from making the same blunder. Now AI can save us from the tedious business of finding actual people to draft as spokesman for the fallacies we hate, and deal with their offended indignation at being so treated. We can use ChatGPT flush out fallacies that pervade public opinion and refute them with fierce efficiency, without obligating anyone to defend their personal honor.
But it's hard to assess the merits of this proposed new genre in the abstract. So let's take fisking AI on a test drive. As the topic, I'll use the 2003 liberation of Iraq.
Fisking ChatGPT on the Iraq War
Here's the prompt I used to start the conversation:
User: Can you explain why the Iraq War of 2003 is widely considered so disastrous?
ChatGPT answered, and I'll quote the answer in full, in italics, interrupted by my fisking responses.
ChatGPT: The Iraq War of 2003 is widely considered disastrous for several reasons. Here are a few key points:
Lack of clear evidence for the need to invade: The Bush administration claimed that Iraq possessed weapons of mass destruction (WMDs) and was a threat to the United States and its allies. However, this claim was not supported by solid evidence, and subsequent investigations found no evidence of any such weapons. This lack of clear evidence led to widespread skepticism and criticism of the war's justification. (Numbering mine)
Guess what, ChatGPT. People have to act on imperfect evidence all the time. Sometimes it's even wise to act as if something is true when you think it is false, e.g., when you buy homeowners insurance as if your house will burn down when you think it won't.
We didn't really know if Saddam had WMDs. How could we? It's hard to penetrate the secrets of a totalitarian tyrant. The logic was, both in international law and in realpolitik calculation, that we couldn't run the risk that Saddam would acquire WMDs, and we needed a weapons inspections regime to ensure that he didn't. Then he kicked the weapons inspectors out. Casus belli.
It's very odd, when you think about it, that Saddam took the risk of excluding the weapons inspectors when he had, as it turned out, virtually nothing, except a little bit of ricin gas, to hide. I thought, as I think did almost everyone else, that Saddam must have had WMDs, or he would never have taken the risk of kicking out the weapons inspectors. I recall hearing speculations that Saddam was more afraid of Iran than the US and the international community, and didn't want Iran to know that he had been successfully disarmed. Who knows? It doesn't really matter. Multiple UN resolutions obligated Saddam to verify that he lacked WMDs. He didn't comply, so we went to war. There were other reasons, too, but that was sufficient. International law doesn’t mean much if you won't enforce them.
Even if Saddam didn't have WMDs, he clearly wanted them. A harsh regime of sanctions had been imposed on the country for decades to prevent him from getting them. Sanctions made the Iraqi people suffer terribly. The war made them suffer too, but at least there was light at the end of the tunnel.
And the fact that Saddam didn't have WMDs made the war a lot less destructive than it would have been if he had.
Huge loss of life and resources: The war caused significant human and economic losses, with estimates of the death toll ranging from 100,000 to over a million people. The financial cost of the war was also staggering, with estimates of the total cost exceeding $2 trillion.
Obviously, war is expensive and causes death, so it shouldn't be undertaken lightly.
But America had suffered horrific terrorist attacks on September 11th, by a widely dispersed and secretive terrorist organization which was inherently difficult to attack directly. America was pursuing a global strategy to defray that threat.
By liberating Iraq, America presented Al Qaeda with what proved to be a fatal choice between letting a democratic transition in the heart of the Arab world create an appealing competitor to their paranoid ideology, or fighting the Americans in a theater of war where it would be inevitable that they killed lots of Arab civilians, thus ruining their reputation among their putative base. Unfortunately, they chose the latter course, to Iraqis’ cost. But that was the key to success in the US’s larger war on terror. By the time bin Laden was killed in 2011, it was an epilogue, not an energizing martyrdom.
Al Qaeda’s strategy on 9/11 made more sense than is sometimes realized. They cared less about the American “far enemy” per se than about the “near enemy” of oppressive secular regimes in the heart of the Arab world, of which they saw the US as a patron. No doubt, they would have liked to see America retreat, but the next best thing was for America to double down on the alliances with Arab dictators through which America had long hypocritically pursued its national interest in the Middle East, and thereby to further discredit itself and mobilize Arabs against it. That was the trap bin Laden was leading us into. That was the trap out of which Bush found an audacious moral high road with his “forward strategy of freedom.”
Al Qaeda largely burned itself out in Iraq, and ceased to be a global threat, because it could not plausibly champion oppressed Arabs after killing, in terrorist attacks, so many Arabs who were just trying to stand up a just, democratic regime in their country after the fall of a hated tyranny. It has been called the “flypaper” strategy. Iraq was where Islamic terrorists went to die, from Sadr through ISIS.
As for the financial cost of the war, America spends hundreds of billions of dollars on defense every year, for the sake of having a powerful and effective fighting force at its disposal. Question: will the US military, built and maintained at such expense, be more effective if it always sits on the sidelines and never fights, or if it sometimes gets tested on actual battlefields? It would be wrong, of course, to fight random wars periodically just for the sake of getting battle experience. But given that we continuously spend so much money maintaining a military force that rarely fights, I would question whether it makes sense to regard the one really intensive and successful fight of the last thirty years as a cost. Insofar as it's defensible on other grounds, it was a good opportunity to keep the military in fighting shape.
Destabilization of the region: The war destabilized the region and created a power vacuum that led to increased sectarian violence and the rise of extremist groups, including ISIS. The war also had a significant impact on the relationship between the United States and the international community, particularly in the Middle East.
The Middle East wasn't all that stable before 2003, but the real point is that what stability it had was bad. Peace isn't the only thing that matters. Freedom does, too. And America paid the price on 9/11 for being the patron of bad, oppressive stability.
Destabilization of the region was, in a sense, a war aim. In Iraq, civil war was a step up from one of the most wretched totalitarian dictatorships on the globe. Other Arab regimes weren't as bad as that, but they were pretty bad, and Arabs’ lack of the modern package of democracy, capitalism, and freedom was seen with good reason as a root cause of terrorism.
As a kind of stretch goal for the war, there were hopes that the liberation of Iraq would trigger the democratization of the rest of the Arab world. The case for war doesn't depend on that. It would have been nice if the fall of Saddam led to a sweeping democratization of the Arab Middle East similar to what happened in Eastern Europe after 1989. But there was plenty of justification for the war without that. The point is that destabilization can be good.
Instead, the Arab dictatorships remained standing, for a while. There was no immediate democratic contagion. Then in 2011, a revolution began in Tunisia, and this time there was contagion, especially to Egypt, Libya and Syria. It’s hard to say if there was any causal connection, or which way it went. What is clear is that the Arab Spring failed to durably democratize any country, where Iraq, for all its faults, holds regular, competitive elections to this day. That underscores how helpful it is, if you're going to transition out of tyranny, to be able to do so under the tutelage of a beneficent foreign occupier with strong democratic principles.
Neighboring Syria shows the likely counterfactual. In 2003, it was also ruled by Baathists, but of a milder stripe. It joined the Arab Spring in 2011 with an attempt at democratization, only to be plunged into a gruesome civil war which left millions dead or in exile, and the country still under the same, now far more blood-stained, dictatorship. Iraq could have been like that. But American intervention in 2003 secured it something better.
Failure to achieve objectives: The war failed to achieve its stated objectives, including the removal of Saddam Hussein from power, the establishment of a stable democracy in Iraq, and the spread of democracy throughout the Middle East. Instead, it left Iraq in a state of political and social turmoil, with ongoing violence and instability.
Perhaps it would be unfair, or unhelpful, to fisk the statement that “the war failed to achieve … the removal of Saddam Hussein from power.” surely that's just a ChatGPT fail, right? Public opinion isn't as stupid as that, is it?
And yet it suggests a haunting question. With someone who got their opinions about the Iraq War by a kind of uncritical osmosis, not researching it, but just reading pundits’ casual references to it, even know that it succeeded in removing Saddam Hussein from power? Or is this ChatGPT fail symptomatic of a larger fail in public discourse?
You can see why pundits who want to settle into a lazy-minded anti-war complacency would rather sweep under the rug the basic fact that we won the war and overthrew a gruesome tyranny once and for all. There's nothing to set in the scales against that which will really tip the balance against. Of course, it's possible for a war to be won and yet not to have been worth fighting. But that's a hard case to make. If you're determined to dismiss the war as a stupid blunder, it's best to keep silence about the removal of Saddam.
In having achieved its stated objectives, Operation Iraqi Freedom stands in striking contrast with other recent American wars such as Vietnam and Afghanistan, where America had to leave, after a long and painful struggle, with the enemy owning the battlefield. To simplify somewhat, we lost. In Iraq, we won.
There were, of course, some stretch goals that weren't perfectly achieved. World War II also left half of the European continent under the totalitarian domination of Soviet Communists. Victories are almost never more than partial, but partial victories are often well worth fighting for.
The war damaged the reputation of the United States: The Bush administration's push for the war was widely criticized as being based on flawed intelligence, and the United States' unilateral action was viewed by many as an overreach of its power. The war also led to the use of controversial interrogation techniques, such as waterboarding, which further damaged the United States' reputation.
Foreign policy is not a popularity contest. It's pleasant to be liked, and useful to be liked. But it's also useful to be feared. The liberation of Iraq was a dramatic display of American power and willingness to use it. It provoked a lot of moral disapproval in our European allies, but was there any strategic cost to that? It's not as if there was ever any risk of a civil war within NATO, or of France and Germany's momentary alignment with Putin turning into military cooperation against the US.
The liberation of Iraq made us a little less liked and, surely a lot more feared in the short run, but that’s worn off, because, unfortunately, the war’s opponents have won the aftermath, turning public opinion against the war.
Liberals could have leveraged the fear created by the fall of Saddam to promote international law and human rights worldwide. While still quibbling about the legal runup to the war, they could have used the precedent to threaten other bad regimes. “Look what we can do,” these brave, hard-headed liberals would say. “If you don't want to be next, respect international law and treat your people decently.”
Instead, the liberals’ priority was to declare “Never again” on the Iraq War, and they were willing to burn all their bridges to lock public opinion into that comfortable, ungenerous, isolationism-adjacent consensus.
As for waterboarding, Abu Ghraib, etc., that's not a critique of the war, but of certain aspects of the way it was fought.
The war led to a humanitarian crisis: The Iraq War resulted in a massive displacement of people, both internally and externally. Millions of Iraqis were displaced from their homes, and the war created a humanitarian crisis that persists to this day. Additionally, the war caused significant damage to Iraq's infrastructure, including hospitals and schools, which had a long-lasting impact on the country's ability to recover.
Again, the humanitarian crisis wasn't just the war but Saddam's regime, including its wars, its domestic killings and genocidal acts, and its aftermath. Iraq fared better than neighboring Syria where an attempt at self-liberation failed catastrophically, because it had the benefit of tutelage by a liberal occupying superpower.
ChatGPT concludes:
Overall, the Iraq War of 2003 is widely considered disastrous because it was based on faulty intelligence, caused immense human suffering and financial costs, destabilized the region, and failed to achieve its stated objectives.
The Iraq War was twenty years ago, but it still matters today because the victory of the wrong side in public opinion has made US foreign policy stupid.
How should international law be enforced? Is there a right of the international community to protect the subjects of totalitarian and genocidal regimes from their rulers? What place should the promotion of freedom have in US foreign policy? How can the US project a credible threat of force against a wide variety of potential adversaries? You won't have anything worthwhile to say about these questions if you're trying to design foreign policy principles compatible with the silly notion that the liberation of Iraq was a stupid blunder.
These days, most people accept the conventional wisdom that the Iraq War was a disaster, and then try to generalize from that. It's like trying to build a science of geography on the premise that the Earth is flat. It touches everything and throws off all the calculations. It tends to deflate any kind of liberal idealism, in favor of a kind of nihilistic isolationism. The world is a horrible place. Nothing can be done about it. Let's run away.
We may finally be shaking off this torpor in light of the Ukrainian people’s heroic struggle to save their nation from Russian aggression. But we have a long way to go. We can't articulate a Truman Doctrine, because it would remind us of the Bush Doctrine. Long ago, the US turned its back on the world in disillusionment after World War I, thus opening the door to fascism and World War II. We let bad arguments con us into similar lazy-minded and cowardly escapism after the liberation of Iraq, fueling the emergence of the Putin-Xi-Iran axis. But liberalism has a lot of strategic depth, and I think the Free World will recover and rise in triumph again, re-articulating the old cause of freedom for changed times, and finding the courage and sagacity to lead it to yet more victories. As we do so, I think the opinions about Operation Iraqi Freedom from the benighted 2010s, roughly captured by ChatGPT in the quotes above, will become unintelligible to us.
Fisking AI
So let's get meta. Is this a worthwhile exercise? Can it be applied to other topics, and contribute to public discourse?
I think so. An interlocutor is often clarifying and stimulating, and ChatGPT did that well enough, without the moral cost of offending and humiliating an actual human being.
Here are some other prompts I'd like to use to draft a ChatGPT bot to channel the groupthink opinion, and then punch it down:
Why do some people think the 1973 Roe v Wade opinion was justified?
What is the actual evidence that life as we know it emerged through a process of evolution?
What is equality of opportunity, and why do people think it's morally desirable for men's and women's opportunities to be equal?
Why do people think countries are justified in excluding foreigners who want to come to get a job?
What is Vladimir Putin trying to accomplish with his war in Ukraine, and why do some Russians think the war is morally justified?
Why do so many people think the 20/20 election was stolen?
How bad will global warming be? What can we do to stop it? And why do many people think it's worth it to do so?
I’ll be grateful if someone else takes on the challenge, and fascinated to read it.