Considering these limitations, a cyber-Pearl Harbor or cyber-9/11 is the wrong metaphor for understanding these systems, Valeriano and other researchers say. Rather than battlefield-defining attacks, cyberweapons are better suited as weapons of espionage, sabotage and other covert operations. Like the cane or pen gun from the James Bond movies, they make for a neat spy trick but are unlikely to alter the international order the way aircraft carriers, precision-guided munitions or nuclear weapons have.Some have seen a more revolutionary role for them. In “The Perfect Weapon,” his 2018 book on the rise of cyberweapons, the Times reporter David Sanger argued that these weapons are unusual because they are very cheap, very stealthy and, unlike other novel weapon systems, available to a wide range of powers, from poor, isolated rogue nations like North Korea to former superpowers like Russia to criminal gangs and terrorist groups.Lucas Kello, an associate professor of international relations at Oxford who is a leading proponent of the idea that cyberweapons might be groundbreaking, has suggested that they could be as transformational for the international order this century as nuclear weapons were last century. He argues that cyberweapons will alter the world order by creating a state of relations between nations that he calls “unpeace” — a status somewhere between all-out war and complete peace, in which nations attack one another digitally in ways that produce major damage but don’t escalate to physical conflict.These predictions are troubling, but they haven’t yet been proved in battlefield conditions. The political scientists Nadiya Kostyuk and Yuri M. Zhukov studied how cyberweapons have been used in Ukraine and Syria. In both cases, they wrote in a 2017 paper, “cyberactivities failed to compel discernible changes in battlefield behavior.”Perhaps the fundamental reason for their failure, as Thomas Rid argued in his 2013 book “Cyber War Will Not Take Place,” is that cyberattacks are not inherently acts of war. Cyberattacks are rarely violent. They are rarely instrumental. While they may damage or annoy an enemy, they don’t often cause hardships that push an enemy toward a specific goal. And they are not always political. In many cases, perhaps the majority, cyberattacks are better suited to criminal or intelligence ends (stealing money, gaining information) than to shifting political calculus. In that sense, Rid said, cyberattacks are not weapons of war like nuclear bombs but are instead “sophisticated acts of network-enabled sabotage, espionage and subversion.”This, of course, is great. It should be cause for celebration that cyberweapons aren’t the second coming of nuclear weapons. But for some theorists, there is a sadness to this realization, too, because it suggests that war will continue to be as physically brutal as it’s always been.