Recently, yet another politician called for the United States government to initiate a “Manhattan Project for Artificial Intelligence.” I’m not going to name the politician. It’s irrelevant where he’s from or what party he’s affiliated with because politicians across the across the land and political spectrum resort to this high-sounding but fallacious argument. In fact, the argument is so bad that it deserves it’s own name, the Manhattan Project Fallacy.

What Is the Manhattan Project Fallacy?

Hardly a political campaign season goes by without a politician or activist saying “I think we need a Manhattan Project for something I think is the most important thing people can be doing today.” Let’s break down why this argument is seductive and what it’s fallacious.

First, let me start with an anecdote. In order to see how this argument works in action, I told a few colleagues about this politician’s plan for a Manhattan Project for AI. Most of them thought that sounded like a fantastic idea. It was also not the least bit surprising that each of them had their own idea regarding what such a project would accomplish. For some, such a project would use AI to better deliver social services to those in need. Others thought that an AI Manhattan Project would deliver battlefield supremacy to the United States military. Still others posited potential medical research benefits that more funding for AI could deliver. In short, phrasing something as a Manhattan Project allows the person making the proposal to let their audience fill in the program’s purpose with whatever they like.

Therein lies the first way the Manhattan Project Fallacy works. Most people gloss over the fact that the original Manhattan Project had a defined starting point and defined endpoint. The starting point was the work of theoretical physicists that showed it was possible to split an atom to release a tremendous amount of energy. In theory, this technology could lead to a massively destructive bomb, a bomb that the Germans were keen to develop during World War II. The clear end goal of the Manhattan Project was to develop this weapon before the Germans could. In a nutshell, one could say the Manhattan Project was about making a theoretical good into a tangible good. Once the United States achieved that goal, the project ended.

That’s where the Manhattan Project rhetoric doesn’t apply to AI (or any other pet project a politician or activist proposes to receive the Manhattan Project treatment). Industry already transformed artificial intelligence from theory to reality. The only question left is what can the world use AI to do. It may shock you, but the companies that spent hundreds of billions of dollars making AI a reality are already in the process of coming up with new uses for it.

There are companies using AI to look for new drugs to treat disease; companies looking to streamline logistics with AI to better deliver goods and services; some businesses are already pursuing ways to integrate AI into weapons platforms like drone swarms. In short, whatever seductive fill-in the blank pet cause that an AI Manhattan Project is supposed to support, there are already businesses devoting their own resources to solving that problem. Is there any reason to believe that the companies that spent more money than small nations have developing AI are now suddenly dragging their feet to deploy it? If not, what is government going to add to the equation?

That Manhattan Project Fallacy Is Just the “War on” Fallacy Repackaged in Positive Spin

If you look at it closely, most Manhattan Project proposals are just more positive versions of proposals for a “War on”—Poverty, Drugs, Terror (take your pick). It appears that when a politician wants an open-ended commitment to tackle a problem that society views negatively, then it’s time to declare war. When that same politician wants an open-ended commitment to something viewed positively, that issue gets the Manhattan Project treatment.

In either case, the hallmark is harnessing the power and finances of government to achieve—well, something. By hitching his wagon to the prestige and success of the original Manhattan Project, the politician, thinker, or activist making the proposal comes off sound like a visionary, when in reality their argument has the content of caveman thought, “AI good. Need more AI stuff.”

It needn’t be only AI. We could just as easily have a Manhattan Project for climate change. If that’s not your thing then perhaps you’re open to a Manhattan Project to cure cancer. Sure these things will be expensive, have unforeseen consequences, and amount to little more than costly action for the sake of action. At the same time whoever proposes them will get to feel warm and fuzzy, as well as all the people who end up getting a cut of whatever vast sums the government doles out to fund these nebulous projects.

When you add it all up, calling for a new Manhattan Project turns out to be little more than a cheap rhetorical trick that misunderstands history, plays on people’s emotions by promising them everything, and accomplishes little more than improving the image of the charlatan making the proposal in the eyes of the gullible. Enough already.

Advertisement

Questions About Your Options After an Accident?

Claim your free, no obligation consultation with a qualified attorney from Grossman Law.

Get Help Now