Perhaps the most important benefit of the surprising Kim-Trump summit was to end a rising crescendo of insults, threats, and provocations by both the United States and North Korea. Although the possibility of war seemed low in recent years, President Donald Trump threatened to start one. That, in turn, made it more likely that a nervous Kim Jong-un would imagine war was imminent and respond accordingly.

Yet the Kumbaya meeting between the two men had a major downside: the president’s unrealistic expectations. He appeared to believe that Kim would give up all his nukes, if not at the summit, then soon thereafter, in exchange for Washington’s promise to respond with love and kindness. However, Kim is no naif. He did not succeed his father in the snake pit of North Korean politics by trusting anyone, let alone his enemies.

In truth, Kim would be a fool to yield his weapons, and leverage, in return for empty promises. Anyone who negotiates knows that you don’t get the best deal by giving the other side what it wants at the start. Especially, the U.S. government, which routinely imposes its will on smaller adversaries. The unpleasant fates of other American targets, especially  Muammar el-Qaddafi—who gave up his missiles and nukes only to find himself on the wrong end of a mob—cannot be far from Kim’s mind.

This naturally fuels skepticism, which I share, that Kim will ultimately abandon his nuclear arsenal. But current U.S. policy presumes that he is willing to do so. In that case, Washington needs to address conditions Pyongyang has set for denuclearization.

Whatever Kim said, at the summit or elsewhere, the Trump administration should not expect him to act against his interest. It ain’t going to happen. Predicating policy on him doing so will ensure failure.

Moreover, Kim and his officials have been quite consistent in their message. I heard it when I spoke with North Korean diplomats at the United Nations in September. They insist that their government must be able to trust the United States before it abandons its deterrent. That hardly seems unreasonable.

Before the two leaders met, Kim was quoted by South Koreans as saying: “If we meet often and build trust with the United States, and if an end to the war and nonaggression are promised, why would we live in difficulty with nuclear weapons?” A more normal bilateral relationship would seem to reduce the likelihood of preventive military strikes by Washington, though no one knows what Kim would consider sufficient progress.

The summit statement, which the president may not have read though issued in his name, reflects Pyongyang’s priorities. First, the document does not define denuclearization; traditionally the Democratic People’s Republic of Korea included the entire peninsula and perhaps beyond. Again, that is a legitimate perspective: Washington routinely threatened the North and refused to rule out use of nuclear weapons. (This might have been good policy from America’s standpoint, but that is a separate matter.) The DPRK has little incentive to abandon nuclear weapons without comparable concessions on Washington’s part.

Second, the summit statement puts denuclearization last, behind establishing “new U.S.-DPRK relations” reflecting a mutual desire “for peace and prosperity” and building “a lasting and stable peace regime.” Only then does Kim promise to back denuclearization, however defined. Again, that order is logical. Why abandon your most powerful weapons if a state of war still exists?

The Trump administration appears to operate on the theory that American positions, if stated firmly enough, inevitably will prevail. If the United States demands something, then Pyongyang must provide it.

But no.

Washington’s leverage remains limited. Military action is a potent threat—no Third World country can withstand America’s armed forces—but would be foolish beyond measure. Kim would not likely acquiesce to an action he likely, and understandably, would see leading to regime change. Then full-scale war probably would be the outcome. The United States (and South Korea) would win, but at a potentially catastrophic cost. Imagine the DRPK dropping a couple of nukes on Seoul and Tokyo. Washington should focus on preventing war rather than starting one.

Sanctions are imposing significant costs on the Kim regime, which clearly desires economic relief. However, financial penalties alone are unlikely to force the regime to capitulate. For any such government, especially one once tagged as a member of “the axis of evil,” security transcends economics. Indeed, in the late 1990s at least a half million North Koreans died of starvation, which had no evident impact on the rule of Kim Jong-un’s father. The younger Kim is more committed to economic reform, but probably not at the price of his regime’s survival.

Moreover, the Trump administration’s “maximum pressure” appears dead. The Republic of Korea understands the importance of rewarding the North for steps taken toward normalization and denuclearization and has pressed Washington to be more flexible. The two governments have unconvincingly attempted to minimize their differences.

China appears concerned that the North might cut a deal with Washington and bypass Beijing. Thus, the People’s Republic of China appears to have loosened some economic controls. Russia backed relief for Pyongyang and is not inclined to do the U.S. any favors. These nations could undermine the sanctions regime developed at so much administration effort.

Indeed, Seoul, Beijing, and Moscow all appear to agree that peace is the most important objective. All desire denuclearization, but not at the price of war. If the North remains generally accommodating, then they are likely to encourage North Korean moderation even short of Pyongyang’s surrender of its nuclear arsenal.

Such an outcome would yield an impasse. The administration would have to choose among war, a nuclear North Korea, and negotiation. It shouldn’t be difficult to judge the last to be the best option.

There are plenty of imponderables. If the North Koreans act on their threat to restart their nuclear program (or do so publicly, since there is no way to confirm they have halted research and development), it might help unify China and Russia as well as the ROK behind the administration. However, Kim has abandoned threatening behavior—rhetoric about turning sundry allied cities into lakes of fire, for instance. Having demonstrated his deft diplomatic touch this year with multiple summits, he isn’t likely to toss away that advantage.

The president should see the current impasse as an opportunity. By bucking the conventional wisdom in holding a summit his more conventional predecessors would have avoided, he created new possibilities which didn’t previously exist. They should not be abandoned. But he must adjust his expectations to move forward.

It would be foolish to make the perfect the enemy of the good. Washington should consider objectives short of denuclearization, and what to offer to achieve them. Most important is receiving an inventory of the North’s nuclear assets, which has proved to be the stumbling block in the past. Last month a North Korean official explained that his government refused to do so because of continuing U.S. “hostility.”

Other, more limited, advances are possible. One would be to make permanent the North’s missile and nuclear test ban. That limits improvements in North Korea’s capabilities and is worth maintaining the suspension of military exercises. Other possibilities include destroying additional test facilities, inserting inspectors in even a limited fashion, and eliminating reprocessing facilities. All would be useful advances and could encourage future progress.

The ultimate result might be to leave the North a nuclear power. (In fact, much might be hidden even if full denuclearization is officially proclaimed.) But America has deterred far worse: Joseph Stalin’s Soviet Union and Mao Zedong’s China, both with nuclear weapons. Washington also deterred the North Korea of Kim Jong-il and Kim Il-sung. They developed weapons to target the United States only because Washington was “over there,” defending the Republic of Korea and threatening the DPRK. Otherwise Pyongyang would have preferred no contact with the United States.

Moreover, Kim Jong-un’s North Korea appears to be significantly different from that of his father and grandfather. There is no evidence that Kim is a liberal reformer. There is no reason to believe that he is not capable of starting a war if he believed it was in his country’s interest. However, he appears to have different, and more responsible, goals.

First, his byungjin, or parallel development, policy differs from that of his father’s military first policy. The former envisioned economic growth as well as nuclear weapons. At the start of the year Kim declared the nuclear prong to be completed, allowing his government to focus on the economy. Pyongyang has implemented genuine, if limited, economic reforms, and invested in education and science.

Moreover, the DPRK’s “Supreme Leader” has demonstrated both a knack for and skill in diplomacy. It could all be a fake, of course, but with seven summits so far—three each with China and the ROK, one with the United States—and more likely to come, Kim appears to be recreating himself as a global statesmen.

His government also has halted its hostile, militaristic rhetoric and eliminated anti-American propaganda from Pyongyang. That was quite an undertaking: the city is inundated with political messages, many of which, I can attest to from first hand experience, were directed against America.