Tuesday, May 11, 2021

At The End Of The Rainbow ... Cloud Computing!

The military’s current fad is cloud computing – that nebulous, magic ‘thing’ that will seamlessly connect every sensor, every shooter, and every individual on the battlefield.  What’s the vision?

 

  • we’ll know everything and the enemy will know nothing
  • we’ll make all the right decisions and the enemy will be paralyzed with indecision
  • we’ll have every sensor and shooter connected and the enemy will have no coordination whatsoever

 

What’s the key to making all this work?

 

It’s cloud. It is having your data connected and in a place where I can run analytics on it in real time or near real time, and then render informed decisions to support war-fighting missions,” said Rob Carey, former Navy CIO and deputy DoD CIO, now president of Cloudera Government Solutions. (1)

 

What’s the key assumption - and flaw - in the preceding statement?  It’s the ‘render informed decisions’.

 

The reality of any battlefield intel system, whether human or machine based, is that the raw data will be badly compromised.  That being the case, we fall back to the well known axiom, GIGO (garbage in, garbage out).  Wait a minute … with gazillabytes of data, light speed networks, and artificial intelligence assisted analysis, how can our raw data be garbage?  It’ll be garbage for the same reasons it’s historically always been garbage:

 

Destroyed sensors resulting in only sporadic and partial sensor coverage

Enemy deception (decoys, misdirection, camouflage, obscurants) resulting in incorrect sensor data

Jamming

Misinformation via spurious or incorrect data (false signals) injected into the networks

Miscommunication resulting from incomplete and interrupted data streams or verbal reports

 

So, with all that garbage entering the system, the output is guaranteed to be garbage.  GIGO.  This completely negates the ‘render informed decisions’ assumption.  Whatever decisions are made will be highly suspect.  GIGO.

 

Now, if we were talking about just one minor aspect of our collective warfighting effort, we’d just make a mental note to question our intel predictions and decisions and continue on with fighting the war directly in front of us - as we’ve always done – using whatever intel we could get but tempering it with the understanding that it is flawed and suspect.  However, we’re not talking about one minor aspect.  We’re talking about the entire basis of our collective US military warfighting effort.  The military wants to base everything on this single, mammoth, all-seeing, all-encompassing, AI-assisted, decision making software program they refere to as Joint All Domain Command and Control (JADC2).  This is the ‘be all’ and ‘end all’ of our military effort.  This is to be our next offset.     ………  and yet it’s obviously flawed from the start at its most fundamental level which is the raw data.  GIGO.

 

You cannot ‘render informed decisions’ with garbage for raw data.  GIGO.

 

Now, let’s be fair and objective.  Some of the data will not be garbage.  Some of it will be legitimate.  This has been true throughout history.  Some intel reports are valid.  Some are not.  Some of the raw data is good.  Some is not. 

 

A ship(s) was sighted at the reported position but the speed, course, number, and type are almost certainly incorrectly reported to an extent.  Which bits are correct and which are not?  We don’t know.  That’s always been the problem.  How do you ‘render informed decisions’ with flawed data?  You don’t!

 

There are no ‘informed decisions’.  Intel doesn’t give us ‘informed decisions’.  It gives us indications and suggestions and possibilities and alternatives and leanings.  In the end, we don’t make ‘informed decisions’, we make tentative assessments and the operational commanders who use those assessments layer on the understood degree of uncertainty.  The commanders may or may not act on those assessments depending on the degree of uncertainty they have about it and the degree of risk they’re willing to assume if they do act on uncertain assessments.  In the meantime, they continue to wage the war that is directly in front of them and whose outcome is decided by firepower, not intel alone.  Unfortunately, we’re abandoning the firepower part and constructing a system that is wholly based on intel … intel that is going to be as suspect as ever, maybe more so because digital systems are even easier to fool than human systems.

 

We can see, then, that this system, this dependence on some almighty digital, cloud-based system is flawed due to the inevitable corruption of the raw data but is that the extent of the problem?  No … there is another level of unreality baked into the concept and that is the end user access.  Consider …

 

“When we want to do call for fires, when we want to do things [where] people’s lives are in jeopardy or they’re at risk, we want that information right there so that they can render their own decision ... at the level of the squad,” Carey said. “So that’s what cloud enables, that connectivity to that information.” (1)

 

So, there you have it.  It’s simple, really.  The squad, in the middle of battle, with explosives going off all around them, dug into the jungle mud and water, with enemy electronic warfare working full time to jam, disrupt, cyber attack any communications, and with enemy fires waiting to zero in on any communications transmissions, will, we assume, still be able to sit quietly and calmly, with full access to a cloud server somewhere and ‘render informed decisions’.  Does that seem even slightly realistic to you?

 

Now, multiply this two-way, intensive computer network communications by a factor of many thousands for all the individual squads, ships, HQs, commanders, etc. who will, presumably, be accessing the system on a continuous basis and you’ll quickly see that the electromagnetic airwaves will be ‘humming’ with non-stop transmissions or, as the enemy calls it, a roadmap to all our locations.  Anyone who thinks we can perform this kind of non-stop, grossly huge bandwidth type of communications undetected is a fool or an officer of flag rank.

 

This concept might have the tiniest, most miniscule chance of working against stone age, third world terrorists in a zero threat environment but it has no chance in any real world scenario against a peer enemy.

 

Just out of curiosity, given that the Chinese appear to have already penetrated every network we have, why do we think our cloud system is going to be immune from cyber attack?

 

I apologize for being crude but this is just mental masturbation on a military wide scale.


Pot of Gold at the End of the Battle Rainbow - Cloud Computing !



 

 

________________________________

 

(1)crisrnet website, “How cloud computing makes Joint All-Domain Command and Control possible”, Andrew Eversden, 26-Apr-2021,

https://www.c4isrnet.com/smr/cloud/2021/04/26/how-cloud-computing-makes-joint-all-domain-command-and-control-possible/

 

36 comments:

  1. I think it might be even worse than you think. Based on recent history, we'll also put it all into production and implement it without even semi-adequate testing.

    ReplyDelete
  2. Bravo! War is like life, only 1000x worse.

    ReplyDelete
  3. I assume that the authors of this idea never saw the remake of Battlestar Galactica. Being off the cloud was the only thing that saved the protagonists.

    ReplyDelete
    Replies
    1. And the entire east coast of the US which is currently experiencing a cyber attack shutdown of gasoline supplies would agree with you.

      Delete
  4. "This concept might have the tiniest, most miniscule chance of working against stone age, third world terrorists in a zero threat environment but it has no chance in any real world scenario against a peer enemy."

    This can be applied to most modern US military ideas, frankly.

    ReplyDelete
  5. Think of the upside of the cloud, it'll go down in the first few hours, compared to the weeks to knock out the current maze of semi-linked networks. Thus freeing the USN Captains from the GOFOs. The PLAN on the other hand will be managed from Xi's imperial Ipad. The PLAN is doomed.

    ReplyDelete
  6. I'm surprised the soldiers still need to call in fires, shouldn't the Chinese just immediately surrender? Maybe they are debating if they should surrender to us or to the cloud?

    ReplyDelete
    Replies
    1. I hope you're being sarcastic. Underestimating an enemy to this degree, will have disastrous consequences, as the US should've learned after Operation Iraqi Freedom made that country a quagmire, forcing us to relearn lessons forgotten after we withdrew from Vietnam.

      Sadly, I can easily imagine some generals and admirals echoing such bravado. A thorough cleansing will be necessary to remove such overconfident officers, before their incompetence costs service members' lives and the nation's victory.

      Delete
    2. If you read carefully online, you can find Chinese emphasize a lot on electronic warfare, include network capabilities. They proclaimed J-20 is one generation ahead of F-22 because F-22 doesn't have cloud network capability (F-35 has). F-22 relies on E-8 to integrate information but F-35 can do themselves. J-20 also can.

      Delete
  7. "is a fool or an officer of flag rank."

    Im not one to critique grammer and such, but shouldnt "or" truly be "and"???

    ReplyDelete
  8. This networking of intel, it makes me recall the Midway movie wedding analogy about the WWII SIGINT that the codebreakers were working with, and how todays info-centric mentality will be no different. Probably even worse. Between gaps in info, false or misinterpreted content, and enemy jamming/secrecy, the "whole picture" will optimistically contain 20% actionable material. Never mind how much our connectivity will allow the dissemination of that 20%, or if leaders across the chain of command make the correct decisions based on it!!! This is where the overwhelming firepower comes in and makes up for communication and intel shortfalls, but sadly nobody seems to believe in that any more...

    ReplyDelete
  9. Beauty of cloud computing is one computer's failure has minimum impact on the whole system. Every node receive data input, computers (not single one) combine all data together to form an integrated theme. Data are stored in multiple computers and many run calculation together (so one goes down, another goes foreground).

    In military, if it operates competently like Google in civilian world, it would be very effective in battle. Each node (a F-35, etc.) provides what it "sees" through all sensors. Even if some sensors gave only partial data, integrated together with other F-35's finding, then, integrate into one more completed picture.

    In today's BVR (beyond visual range) warfare, information is very (if not most) important. You must first locate precisely enemy far away, then, with precision guidance to deliver your weapons. Discover then destroy immediately is how one can survive in fighting another superpower.

    Lots of large naval guns to fire together? It only produce a fire work show without any usefulness for a simple reason -- range not long enough. This is why modern naval ships rare install many guns.

    Generals usually fight last wars as they are educated on what happened before. Many of them lack enough scientific knowledge to predict what future would be. many technologies are promoted by various interest groups buy top generals need to have knowledge to discern what really works in future war.

    ReplyDelete
    Replies
    1. "Beauty of cloud computing"

      Ugliness of cloud computing is if one computer is hacked/infected, they all are.

      Cloud based computing has many weaknesses. Feel free to list them along with what you see as strengths.

      "naval guns"

      Your dismissal of naval guns demonstrates a very elementary understanding of how naval guns are used and what they can provide.

      "military, if it operates competently like Google in civilian world,"

      If the military operates like Google, we're in trouble. Google is routinely hacked, attacked, degraded, and compromised. A cursory Internet search will list many examples. Military applications need to do much better than that.

      Delete
    2. In civilian world, Google faces far more users than military would be. Bottom line, Google runs most time and billions people are OK with it.

      For F-35, J-20, ... etc. network (cloud) based platforms, if one node "see", the whole team "see". If one's sensors are damaged, others provide information as their mission computers are integrated into one. No need to talk through radio as you watch in movie of wars of past.

      Before this, all naval fighters send their information to E-2 to be integrated so you see crews in E-2 (or E-8 for Air Force give direction to each plane. Now, there is no need to talk verbally among F-35C pilots.

      Delete
    3. "Google faces far more users than military would"

      The number of users is absolutely immaterial. The only thing that's important is network security and it only takes one penetration of a military network to cripple it and render it useless.

      "if one node "see", the whole team "see"."

      You've been reading too many sales brochures! The F-35 has severe comm/linking challenges. It can't share with other aircraft except by the old, limited, ?Link 16?. I'm unsure about linking with an E-2 but I've not heard that it can. The US Navy is having to rebuild its carrier facilities to communicate/link fully with the F-35. The F-22 and F-35 can't talk to each other because they use two different comm protocols. And so on.

      The ?Air Force? was working on some kind of aircraft 'universal translator' for comms but I've not heard anything recently.

      Delete
  10. Serious question;

    Could today's US Navy beat the 1991 US Navy in battle?

    Lutefisk

    ReplyDelete
    Replies
    1. Unlikely, if nothing else because EMCON isn't a thing anymore.

      Delete
    2. (Don McCollor)..."Silence is Golden"...

      Delete
  11. Two thoughts based on the Cloudera executive's quoted comment.

    Thought the first:

    The (perceived) benefits of any technology are always vigorously marketed by the vendor and a handful of committed adopters.

    The deprivational aspects, weaknesses, and debilitating tradeoffs of vendor's particular design choices are left for you and you alone to discover and may *never* be widely disseminated even after everyone in the industry has long since figured them out. (tacitly though - wouldn't want make whomever made purchasing decision look foolish - why p*ss off your current or future boss)

    Technology sales, such as our friend @ Cloudera is engaged in, works on social proof and information asymmetry.

    Point the second.

    I'm gobsmacked at the effect AI and autonomous tech has on the Western military mind. We're just accepting the claims of ungodly effectiveness which is paralyzing our ability to put steel, gun tubes and aircraft into the order of battle. Meanwhile, the competition just keeps building ships and planes with tech we mastered 30 years ago in a re-armament campaign reminiscent of 1930's Germany.

    We've been wholly defeated by the idea of AI. Not by any particular manifestation of it.

    And of course we can't produce any useful AI weapons of our own. Yet somehow this tangible evidence doesn't get weighed as an important refutation of the original claims of total AI battlefield awareness and dominance.

    ReplyDelete
    Replies
    1. Nice comment.

      "I'm gobsmacked at the effect AI and autonomous tech has on the Western military mind."

      You're correct. Now … why? Why have Western military minds bought into this with no real proof of effectiveness? These are not stupid people with low IQ's. So how/why have reasonably intelligent people latched onto what seems to the rest of us to be a badly flawed concept?

      I've offered my thoughts on this throughout the blog. What are your thoughts?

      Delete
    2. Great insight!! I think the great confidence bestowed on this networking concept isnt so much from academic analysis. More likely I believe its a byproduct of multiple generations of officers amd leaders that have been forged in the era of transformationalism. Its not that the concept has been neutrally looked at, been gamed and tested and found plausible. Its pushed and endorsed more by a "belief" in it. The idea of an omniscent network is the stuff of '3rd offset' daydreams. Truly new and revolutionary weapons dont just happen, and in grasping for the next transformational item, these leaders and officers have tried to create one, and are pushing forward without properly giving the red team, let alone 'ol Murphy his say in its functionality on game day.
      Im afraid much of this reflects the changing of our society over the recent generations. Trying to not be political here, but the oft-used modern cliche with my adulteration fits... The 'feeling' that clouds and networking will transform warfare are not supported by the 'facts'...!!!

      Delete
    3. "Why have Western military minds bought into this with no real proof of effectiveness? These are not stupid people with low IQ's. So how/why have reasonably intelligent people latched onto what seems to the rest of us to be a badly flawed concept?"

      Because it was easy. Because it eliminated the hard work of actually having to think things through.

      It's not just the military. See Colonial Pipeline.

      Delete
    4. "Because it was easy."

      Not that military personnel are immune to laziness but they are trained to do the hard things and, by nature, are more likely to embrace difficult challenges. So, there's more to it. Dig deeper. What is it?

      Delete
    5. "Not that military personnel are immune to laziness but they are trained to do the hard things and, by nature, are more likely to embrace difficult challenges. So, there's more to it. Dig deeper. What is it?"

      I think there are two classes in the military, just like every other government organization. The ones in the field work their butts off to do the hard things. The bureaucracy at the top sit on their butts and take the easy way out. That's one big reason why I agree with you that we need to get rid of the bloat at the top.

      Delete
    6. Wondering what you're searching or asking for CNO??? "Its easy" probably has some merit, but what about the network taking the decisions out of the hands of lower-level leadership? Are they growing to fear the responsibility of command in battle?? Or is it a grasping at straws trying to avoid casualties and "clean up" warfare?? Maybe removing people from the equation, turning it into a big video game thats not so personal?? Does it go hand in hand with the big push for unmanned everything?? I know there's an answer youre trying to coax out, but its just not crystallizing after only one coffee this morning...

      Delete
    7. Oops, I left out the obvious option... The higher level leaders want to be able to micromanage, TAKING the control from junior leaders in the field... Is this the mentality behind it??

      Delete
    8. "Oops, I left out the obvious option... The higher level leaders want to be able to micromanage, TAKING the control from junior leaders in the field... Is this the mentality behind it??"

      Yep. Make war an antiseptic video game that we can control from our LMDs (large metal desks) in DC, and the hell with how many people get killed or maimed in the process.

      Delete
    9. "I think there are two classes in the military,"

      I'm sure that's true to some extent - maybe a significant extent - but I don't think that's the explanation. I think the answer is a societal one. As a society, we've come to look to technology for the answer to all our problems. Thus, it's only natural for ANYONE in the military to look to technology as the answer rather than the hard work. After all, those people are products of their/our culture. I don't think it's laziness or dual classes or anything else; it's simply a conditioned dependence on technology that won't even allow people to consider a lower tech alternative. It's a paradigm that the leadership doesn't even recognize.

      Delete
    10. "Dig deeper. What is it?"

      Is it because of the lack of incentive? Nobody is rewarded for ACTUALLY demonstrating and ACTIVELY increasing the combat edge of our military. They are, however, rewarded when they can fund a new jobs program with promises of an easy retirement in any defense companies and happy senators friends.

      A better solution would be having an actual exercise maintain by a watchdog similar to DOT&E (outside of the chain of command and funded by Congress) and contains retired high-ranking staff officers with little conflict of interests (no or little political/defense companies ties). The only condition for any new weapons to get approved for IOC is to be put through the stress/combat tests designed by this board. They might even get blew up a few prototype to understand the capabilities of this ship thoroughly. And more to that, this is also the only way for program manager to get promoted to higher rank. We further incentivizes these high-ranking program managers to go in defense companies for guidance and let them identifies and proposes even more prospective ideas. That would be the dream!

      Delete
  12. Note that this was all predicted decades ago (during the days of vacuum tubes!) by science fiction writer Arthur C. Clarke. Check out his short story "Superiority", in which a more powerful empire decides to resort to new "wonder weapons" to shorten a war which they are almost guaranteed to win. Well, due to problems with the new technologies, they suffer defeats, then try more new technologies, then more problems and more defeats, and finally total defeat in the war (since they'd stopped building the old kinds of weapons to make the new wonder weapons, the enemy eventually overwhelmed them with their "inferior" weapons).

    ReplyDelete
    Replies
    1. Seems entirely prophetic...will have to find and read it!!!

      Delete
    2. There's a copy online, at http://www.mayofamily.com/RLM/txt_Clarke_Superiority.html

      It should be required reading at all our military academies.

      Delete
  13. Have we ever tried something like a Fleet Problem or a Springtrain with unlimited jamming? If so, how did it work?

    ReplyDelete

Comments will be moderated for posts older than 7 days in order to reduce spam.