The $1 Trillion Defense Budget will be Test Free.
But It Will Be Autonomous
The
late Admiral Hyman Rickover, father of the nuclear navy, once suggested
that the way to reform the Pentagon was to instruct the entire
workforce to go outside and form two circles around the building, and
then fire the outer circle. The Trumpian leadership now installed in the
military headquarters is clearly fearful that Elon Musk’s Red Guards
will storm the building and seize control as they have at other
agencies. Hence the memo issued
by Deputy Defense Secretary Steve Feinberg on April 7 ordering a
sweeping overhaul of Pentagon management. The memo is replete with
directions to clear away redundant positions and sprinkled with
invocations of “streamlining” and “innovation.” This may or may not be
enough to keep Musk’s minions at bay, but taxpayers should not expect
that the end product will produce “the most lethal fighting force the
world has ever known” - an aspirational pledge common amongst all factions of the war party, very much including Democrats.
The
word “test,” as in testing weapons to ensure they are operationally
effective, does not appear in Feinberg’s edict. Nor does it show up in
Donald Trump’s April 9 executive order mandating a streamlined weapons-buying process, thereby promoting innovation.
Stockholm Syndrome in the Weapons Labs
The
military establishment’s fear and loathing of operational testing -
i.e. subjecting weapons system to conditions likely to occur in actual
combat - has been evident ever since the creation of the Director of
Operational Test and Evaluation back in 1983 by a bipartisan
congressional coalition. The office, which for a time enjoyed the
protection of strong directors committed to the task, has subsequently
been progressively enfeebled thanks to the installation of leaders more
deferential to the priority of preserving the money-flow at all costs.
Now, as Dan Grazier of the Stimson Center has pointed out,
there is an insidious effort to combine operational testing with
development testing. Development testing involves officials working
alongside contractors building a weapons system to check that specific
components and sub-systems meet specifications. In Grazier’s words: “Injecting
operational testing personnel into the development process creates a
potential Stockholm Syndrome risk. They would have a stake in the
outcome because they will be involved in the development process when
they should be neutral arbiters evaluating results right before
graduation. In other
words, under the hallowed watchwords of “streamlining” and “innovation”
we will end up with a familiar roster of weapons programs unfit for
combat.
$1Trillion Up for Grabs!
Trump
did threaten to axe programs that are more than 15 percent behind
schedule, 15 percent over cost, unable to meet any key performance
parameters, or “unaligned” with the SecDef’s mission priorities. There
is obviously a lot of vulnerable low hanging fruit out there, notably
the infamous F-35 fighter (two hundred percent, at least, over budget)
or the Sentinel ICBM (81 percent over budget) along with many others.
However, assuming the possibility that one of these systems does get
axed (a cut-back of F-35 production has long been on the cards) there
are plenty of candidates pushing for a share of that $1 trillion budget promised by
Hegseth and Trump. One has already garnered a slice - Boeing’s F-47
fighter, already in receipt of a $20 billion development contract, with a
firm commitment for lucrative production contracts to follow.
Autonomous Maybe, But What Can it See?
The
Anduril corporation, a competitor for the Collaborative Combat drone
program slated to operate in conjunction with the F-47, is sure to reap
copious rewards, the firm being a poster child for everything that is
supposed to be new and different about the Silicon Valley way of doing
defense business. Rebooting the Arsenal of Democracy,
featured on the company website, has much to say about the sclerotic
state of U.S. weapons development, beset by lack of competition,
excessive bureaucratic hurdles, the need to develop weapons more
quickly. But central to the company’s mission statement is the assertion
that “Software will change how war is waged” thanks to the advent of
AI-driven systems that can coordinate with each other, detect and
destroy targets, attack enemy cyber systems, etc. Such religious belief
in the pending omnipotence of AI appears unmoved by the inability of AI
systems to overcome the incurable problem of hallucination - making
things up - nor of the more prosaic problem that an autonomous weapon,
however artificially intelligent, has to be able to observe the real
world through fallible sensors and make sense of what it sees.
Needless to say, the words “operational testing” do not appear in the Anduril mission statement.