The Trouble With Targets

The threat of failure hangs over efforts to solve big, multidimensional engineering challenges, but that shouldn’t stop us from trying

Advertisement

Among technology leaders burdened with unmet and ill-defined goals, setting engineering targets is the rage, promising specific results to people, corporations, and government.

Whether the demand is for capping the rise in Earth’s temperature, creating therapies to halt memory loss in the elderly, or expanding farm output to meet the world’s growing population, the answer is the same: Set targets.

The urge to target is varied and insistent—and contradicts the widespread view that technologies and their underlying physics and mathematics determine outcomes, not people. A movement to put humans at the center of engineering is fueled by the popularity of “effective altruism” and humanitarian engineering. Increasingly, politicians and the public talk about the technologies they want rather than settling for what Technology—with a capital T—can give them.

The sensibility informs a range of urgent questions for engineers. Does artificial intelligence pose threats to humanity? Target good outcomes of AI. Might robots destroy human employment? Create robots that only help workers. To counter Ebola, cancer, and other lethal diseases, invent vaccines or cures.

Targets (think putting humans on the moon) are a clever means of holding technologists to account, charting their progress, and insisting on results. In an age of limited resources, great inequality, and growing uncertainty, clear aims trump the value of free-wheeling inquiry.

But while appealing, targeting masks complexity and encourages overconfidence, even complacency. Existential threats, it turns out, are easy to identify but difficult to resolve.

Terrorism, chaotic climates, cyberweapons, mysterious diseases, vanishing species: The list of fixable perils grows longer with each year. Technologies of abundance altered our existence but came at a cost that is only now being more accurately counted. This tension—between the glories of our engineered lives and the price to be paid for them—is the essential drama of our times. No one sits closer to the center of this gathering storm than the engineer. Only the engineer understands the contradictions of the human-built world and possesses the skills to craft solutions.

Yet engineering never occurs in isolation. Targets reflect the desires of masses of people. In a fragmented world, where cherished diversity spawns at times irreconcilable claims between factions, only existential threats generate the unity of purpose that in turn produces universally shared hopes for emerging technologies.

No surprise then that the technological landscape is littered with failed efforts at targeting, whether mounted by governments, corporations, or civil society.

Crafting targets is part art. Targeting seems least effective when goals are broad and fuzzy. Such targets as improving primary-school education, curing cancer, or preventing terrorists from using social media to win converts can seem impossible to reach. These challenges and others like them require pushing down multiple pathways toward many smaller targets, which then exponentially increases cost, complexity, and the chances of failure.

The multidimensionality of many goals has serious implications for targeters. Constraints on innovation aren’t limited by ambition or even resources. The prospects for bending the physical world to humanity’s wishes can never be fully tested in a lab or modeled on a computer. Because rising temperatures reflect many factors, halting warming will require many engineering projects, each with its own target. The interactions among the new targets raise the specter of an infinite regress.

Don’t despair. Setting targets makes sense, especially if targets are concrete, feasible, and widely desired. Yet ambitious technological campaigns demand enormous humility. Grand aspirations for engineering must be matched with an awareness of the potential for choosing the wrong target or messing up in pursuit of the right one. All we can know for certain is that our good intentions are never enough.

About the Author

G. Pascal Zachary, a professor of practice at Arizona State University’s School for the Future of Innovation in Society, is the author of Showstopper! (1994), on the making of Windows NT.

Advertisement