After years of allowing its military forces to atrophy (with the exception, perhaps, of its Strategic Rocket Forces), Russia has had ambitions to grow its defense budget and modernize its forces, announcing an increase to $75 billion as recently as 2014. However, these plans have largely been put on hold, with the devaluation of the ruble reducing Russian purchasing power roughly in half (i.e., to the equivalent of $35 billion annually). Subsequently, Russia announced it would be cutting its 2016 defense budget by 5 percent while it faces additional expenses due to its higher tempo of operations (in Syria and along the border with Ukraine). International sanctions are also likely to have an effect on Russia’s defense production as many high-tech items needed must be procured from international vendors.

Thus, fiscal reality doesn’t leave the resources for force modernization let alone the development of new advanced weaponry that DoD has raised concerns over. And, comparatively speaking, the entire Russian defense budget is a small fraction (7 to 8 percent) of U.S. defense spending.

Russia’s Limited Military Modernization. Most of Russia’s military equipment was produced in the Soviet era. Most of its twelve-thousand-plus tanks are in poor repair, and only roughly 2,800 are operational (mostly 1970s-vintage T-72s). Even the vaunted Russian air defense systems were developed under the Soviet Union and only been subject to incremental improvements. The Russian air force presently relies on combat aircraft designed (and mostly built) during the Cold War, and the Russian navy similarly relies on aging ships and submarines in a poor state of repair, manned by poorly trained crews.

While Russia has made some limited use of modern equipment in Syria and Ukraine, including UAVs, laser-guided bombs and missiles as well as electro-optical weapons, most of the Russian capabilities deployed in these operations are older systems from a prior era.

By all accounts, Russia’s efforts to modernize its defense capabilities to date have been limited at best. Despite the appearance of new products in the pipeline, Russia has been unable to develop a new main battle tank or infantry fighting vehicle, and still relies on Soviet designs inferior to those of the United States. Its future fighter program, PAK-FA, has stalled over cost and technical issues, and its promise of new carriers and heavy bombers have not been realized.

Russia has successfully modernized is its strategic nuclear force, which is seen as politically useful, both for intimidating its neighbors and maintaining Russia’s image as a great power. There has been some talk of Russia redeploying low-yield tactical nuclear weapons as a potential counter to U.S superiority in precision weapons, but whether this is a credible threat is debatable. Russia well knows that crossing the nuclear firebreak would bring the risk of uncontrolled escalation, something Russia would probably not be willing to do in pursuit of limited regional objectives.

The Russian military also has repeatedly attempted and failed to professionalize its military forces. Unable to attract volunteers, it still relies on unpopular conscription; draft evasion is endemic. As a result, Russia today has, in effect, two militaries: a small, professionalized force of perhaps one hundred thousand men, and a much larger force of poorly trained, poorly motivated conscripts with obsolete, poorly maintained equipment. The former is capable of maintaining internal security, conducting small-scale expeditionary operations, and perhaps seizing and occupying portions of neighboring countries in the face of light or disorganized opposition. The latter has very limited military utility.

There are serious questions whether Russia will be able, for the foreseeable future, to mount a sustained challenge to U.S. military power. Russia is most likely to be a regional power with aging forces and selectively modernized capabilities in discrete areas where cost-effective investments can be made (like the cyber domain). Thus, Russia has some capability to create havoc along its periphery and pose challenges in select other areas. But the notion of Russia as a “resurgent” great power and Russian capabilities as a frame of reference for an “offset strategy”—i.e., the development of our force of the future—is debatable at best.

IV. Autonomy as the Centerpiece of the Third Offset: Is This a Good Bet?

“The third offset strategy is based on the premise that advances in artificial intelligence and autonomy will allow the joint force to develop and operate advanced joint, collaborative human-machine battle networks that synchronize simultaneous operations in space, air, sea, undersea, ground, and cyber domains.”—Deputy Secretary of Defense Robert Work

The assertion that autonomy and artificial intelligence will be central to the Third Offset also bears considerable scrutiny.

Legitimate Areas for Investment, But Should We Bet the “Work(s)”? There is no doubt that these specific technology areas, in which the United States has invested for years, are advancing. Anyone who test-drives a Tesla, with its driverless autopilot feature, can attest to the fact that autonomous capabilities are arriving in the commercial sector. Certainly, as ideas go, these continue to be worthwhile ones for R & D investment—they are future technologies that are rapidly accelerating and warrant our focus, as one focus among others.

But the question is whether there is a basis for betting the house, or the “Work(s),” so to speak, on these particular ideas at this juncture? Where is the evidence that leads us to believe, with any degree of confidence, that these will be the enabling framework for the Third Offset, if there in fact is one.

A number of factors give one pause when considering this question:

• Are We Wise Enough to Know? At a conceptual level, who can identify today the centerpiece of an “offset” for a future force that will not be fielded for many years. Given the messy, unpredictable nature of innovation, is it bordering on hubris or chutzpah to take this view? On balance, developing a DoD, and service-centric autonomy plans—efforts now underway—appear premature.

• A Range of Capabilities Will Likely Sustain U.S. Military Leadership. In the future, we are likely to face a wide range of threats rather than a singular “Cold War” challenge, and will likely need a range of capabilities in different domains to address those threats. Thus, it is hard to believe that any single technology, including autonomy, will convey an overriding “overmatch” across the board. It is far more likely that any U.S. military dominance would be achieved in the future through a combination of different operational constructs and technologies that span a range of capability areas, from virtual reality to cyber technology to biotech.

• Why Will Autonomy Necessarily Yield a Sustained Overmatch? Even if these technologies become sufficiently mature to use on the battlefield, will they provide a competitive advantage, let alone one that can be sustained, given the proliferation of relevant commercial technologies and investments made by other countries?

• A Disconnect Between the Russian Threat and Autonomy Solution. While autonomy is a prudent long-term bet with China in mind, do we really need this capability to address Russia, given its inherent economic and demographic constraints, and its limited military capability and ability to invest for the future? There are other strategic approaches for dealing with what is essentially a regional power with the ability to conduct expeditionary warfare in certain limited situations.

• The Machine Interface: Ethics and Value-Added. A central ethical issue, which conjures up the Skynet program from the Terminator movies, is whether we would cede all control to the machines, removing the human interface especially over potentially lethal decision-making functions. Deputy Secretary Work has suggested we will always retain the option to keep human beings in the loop (Iron Man, rather than Skynet) and has highlighted the importance of man-machine interfaces. But will these innovations will in fact yield benefits to the war fighter if we leave humans in the loop and, by definition, automatically slow down the engagement cycle?

• Technical Challenges. Finally, there is a host of technical challenges in this area as well, including latency (the time for data to go from the unmanned platform to control and back again), situational awareness, unprogrammed situations that machine intelligence may not be equipped to handle, and the like, as well as cost.

In sum, many issues, including “unknown unknowns,” may derail autonomy as a Third Offset—we just lack the foresight to know if it will succeed. Whether this technology makes the difficult, long-term transition from the laboratory to the war fighter remains to be seen. In truth, an offset, if there is one, is something that only lends itself to description after the fact, when it is there for us to behold.

Put another way, if autonomy becomes the centerpiece of our overmatch going forward, it is likely not because of top down guidance from DoD, but because years of defense investments, trials and errors, including some failures, come to fruition. At best, DoD may be putting a punctuation mark on something that has been happening for years, perhaps giving it a little push toward the goal line.

Thus, some investment in autonomy certainly is warranted but there is no basis to bet the “Work(s).” Rather, we need a portfolio of sustained investments in different technologies, constructs and organizational structures—and considerable experimentation—to address the reality that we just don’t know what the future “offset” might be until it happens.