Winning the Super Intelligence Race

Is Super Intelligence (President Trump’s useful renaming of AI) a blessing or a curse? The same question was asked of nuclear weapons, and how we respond will determine who wins the ongoing Super Intelligence race.

Lately there has been a great deal of public concern over SI and its potential to do harm. The answer, say some critics, is place guardrails around it, limit it, slow it down. Senator Bernie Sanders has even proposed legislation that would impose severe penalties and lengthy jail sentences for researchers who exceed arbitrary limits imposed by nay-saying bureaucrats.

The idea that this transformational technology can even be limited is a stretch. And since technologists can always move to other countries, Sanders’ punitive Luddite approach would be about as effective as his other socialist ideas.

One suggestion for limiting Super Intelligence is that the United States should engage China in talks modeled on the SALT arms control process of the 1970s. Thought in this case if would be the Super Intelligence Limitation Talks, or SILT.

This problem with this approach is that is assumes SALT was successful when it wasn’t. Ronald Reagan correctly identified the main problem with SALT; only one party was adhering to the process.

Throughout the heyday of détente, arms control became something of a cult. Its adherents had an unquestionable faith in its importance and effectiveness. The American arms control community believed that they had unlocked the secret to global peace and stability, and through the patient and sometimes frustrating process of negotiations they would be able to educate the Soviet leadership in their basic truths.

However, the Soviets did not need an education. They understood that negotiations are simply part of the same political struggle. So, when the United States offered accommodation, the Soviets saw weakness, and they took advantage of it.

In 1970, at the advent of SALT, the United States had around 26,000 nuclear warheads, to the Soviet Union’s under 12,000. In 1986, after a decade and a half of the arms control orthodoxy, Soviet stockpiles had quadrupled to 45,000, while the US arsenal dropped to 23,000.

Reagan’s insight was that the Soviet Union would never consider reducing its nuclear forces while the United States it was unilaterally cutting back. In essence you had to build up in order to build down. So Reagan embarked on a multi-track effort to show that the United States was still in the arms race. This involved nuclear modernization, building up strategic forces, and of course missile defense through the SDI program. The United States demonstrated that it could dominate the process, because to have peace through strength you need the strength.

The same is true of communist China today. The United States has the lead in Super Intelligence. China is vitally aware of this and is attempting to steal as much US advanced technology as it can. They are also the main beneficiaries of the US slowing down, because they surely won’t.

It is hardly a coincidence that with President Trump meeting Chairman Xi we are suddenly hearing a lot about how Super Intelligence is going to destroy us, how it will get out of control, how it needs severe limits and guardrails. And it’s interesting that this narrative wis coming from the same doom sayers who had been warning us about the end of the world from overpopulation or global warming for nuclear war. SI is their latest apocalypse.

Unilaterally slowing SI research and development means ceding that ground to China. As with the Soviet Union, the Chinese may agree to cut back their efforts in a SILT agreement but there will be no way to verify or enforce it. The US will accept limits and China will seize the advantage. And as President Trump said, in the Super Intelligence race whoever wins, wins. The consequences of losing are being at the mercy of a totalitarian power who would use SI against us in ways we can only imagine.

The only way the United States can lose the SI race is by hobbling ourselves. Our technology is superior to any in the world, and will remain so if it is allowed to. And if we lose the SI, there is no way back. So the real threat is not that we’re going too fast, but that we would allow SILT to grind us to a halt.