How Hot is Ukraine Gonna Get?

[db_zero]

I agree with General Berger as well as former Sec of Def. Gates who commented on the limitations of computer war gaming and simulation.

It’s easy to pat yourself on the back weeks into the fighting and you know how to plug in the variables to get a result close to what’s really happening.

Try doing that before the shooting starts and make life and death decision…

Same thing goes on in the financial industry. They have clusters of computers grinding out potential outcomes before they happen and are consistently off the mark.

In either case it doesn’t stop people from trying, but in the end it has been demonstrated that monkeys throwing darts do better at predicting.

[akd]

@sburke, another VDV Major: Oleg Patskalev, deputy battalion commander, 331st Guards Airborne Regiment, 98th Guards Airborne Division.

 

[The_Capt]

14 minutes ago, db_zero said:

I agree with General Berger as well as former Sec of Def. Gates who commented on the limitations of computer war gaming and simulation.

It’s easy to pat yourself on the back weeks into the fighting and you know how to plug in the variables to get a result close to what’s really happening.

Try doing that before the shooting starts and make life and death decision…

Same thing goes on in the financial industry. They have clusters of computers grinding out potential outcomes before they happen and are consistently off the mark.

In either case it doesn’t stop people from trying, but in the end it has been demonstrated that monkeys throwing darts do better at predicting.

That is only because CM has not evolved an operational layer yet.

[DesertFox]

Seems like they hit a jackpot with this one:

 

 

[Artkin]

Just now, The_Capt said:

That is only because CM has not evolved an operational layer yet.

Yet? Lets GOOOOO

[BeondTheGrave]

30 minutes ago, Lethaface said:

* The CBR regulates the supply of Rubles but is not the only 'venue' holding Rubles in reserve (or other means of receiving Ruble) and selling them (although I guess lack of demand has it's impact on the availability of Rubles worldwide ;-)). 

On the Ruble question I think it would be very interesting to know who all has Rubles in hand already. Much like how the CBR has built up a warchest of currencies, how have western governments prepared to cut off Russia. If Europe is holding a $1b euro worth of Rubles, say, they can buy a lot of gas before this becomes an issue. Or if the US or even China has large reserves or Rubles they could swap currencies and make a tidy little profit along the way. 

[Lethaface]

25 minutes ago, DesertFox said:

Some humour is necessary after all. Enjoy Alexander Nezorov!

Uncanny predictions of Ukraine's war from April 2021 by former Russian MP Nevzorov

 

 

Have a like!

[BeondTheGrave]

edit double post, plz delete/ignore 

Edited March 23, 2022 by BeondTheGrave

[DesertFox]

 

 

[db_zero]

13 minutes ago, The_Capt said:

That is only because CM has not evolved an operational layer yet.

It also lacks a strategic layer and operates in a vacuum.

Decisions like war or any other major decision is made in the presence of many other people with competing views. Groupthink, organizational politics and a wide array of other factors come into play.

Look at the interaction between Putin and his head of his spy agency before the war and he was asking for opinions…

I can only imagine the reaction in a group decision when the head hancho asks Mister  X for his opinion on why he thinks a war will go this way and the answer is “it’s based on a computer simulation”.

I love CM and computer simulations in general but I wouldn’t base serious life decisions on them.

[dan/california]

45 minutes ago, DesertFox said:

Some humour is necessary after all. Enjoy Alexander Nezorov!

 

 

Bleep me, he called it chapter and verse!

[The_Capt]

2 minutes ago, DesertFox said:

 

 

War to the Knife

[akd]

Interesting map seen here:

Looks like it includes this:

 

[dan/california]

26 minutes ago, The_Capt said:

That is only because CM has not evolved an operational layer yet.

That sobbing sound you hear is Steve going "NOOOOOO", and then drinking more.

[The_Capt]

14 minutes ago, db_zero said:

It also lacks a strategic layer and operates in a vacuum.

Decisions like war or any other major decision is made in the presence of many other people with competing views. Groupthink, organizational politics and a wide array of other factors come into play.

Look at the interaction between Putin and his head of his spy agency before the war and he was asking for opinions…

I can only imagine the reaction in a group decision when the head hancho asks Mister  X for his opinion on why he thinks a war will go this way and the answer is “it’s based on a computer simulation”.

I love CM and computer simulations in general but I wouldn’t base serious life decisions on them.

Well you hit on a pretty important point here and one modern militaries are all watching out for...what happens when we can?

Some argue that war is too chaotic and non-linear to ever be able to create effective models that provide predictive analytics; we do it now with a collection of human brains called staff, and as we see in the current example they can fail too.

Others argue that war is "chaoplexic" and not truly fully chaotic because it is bounded by some hard rulesets and frameworks (e.g. physics) and as such the irrationality of human beings can also be smoothed out or at least made more predictable (e.g. Cambridge Analytica and Big Data stuff). 

Most of us in the business don't really know but we sure are interested if that second one turns out to be true.  Why?  Because the first side to take AI/Machine Learning/Quantum Computing and use it to create effective predictive analytical models that stand up, has an enormous advantage on the battlefield.  The peace-nicks are all on about "kill bots" but this application of technology to warfare in predictive modeling has frightening implications that really give weight to the ideas of "cognitive superiority" as a applicable and measurable concept.

Lotta skeptics wave it away as "never going to happen", which it might not.  Then again there were a lot of skeptics on powered flight as well, and mechanized, and PGM, and smart weapons and cyber...so there is that.     

Edited March 23, 2022 by The_Capt

[dan/california]

42 minutes ago, DesertFox said:

Seems like they hit a jackpot with this one:

 

 

I am sure the Ukrainians only asked for 4 or 5 C5s of very special toys.

[Bulletpoint]

2 hours ago, akd said:

@sburke, okay, this one is a bit sketchy, but fun.  37th Guards Separate Motor Rifle Brigade commander Col. Yuri Medvedev, hospitalized in Belarus after being intentionally run over by one of his own soldiers:

If you intentionally run over an unsuspecting guy with a tank, and he doesn't die, then you're probably not trying very hard.

But I guess with the level of incompetence shown so far by the Russians... could be true.

[White2Golf]

11 minutes ago, The_Capt said:

 

Most of us in the business don't really know but we sure are interested if that second one turns out to be true.  Why?  Because the first side to take AI/Machine Learning/Quantum Computing and use it to create effective predictive analytical models that stand up, has an enormous advantage on the battlefield.  The peace-nicks are all on about "kill bots" but this application of technology to warfare in predictive modeling has frightening implications that really give weight to the ideas of "cognitive superiority" as a applicable and measurable concept.

 

When I think about this, I always end up wondering how they get AI to process "The Commanders Intent"?  It is not always "occupy space x".  I'm just a Sergeant, but for me that was the most important part of an order.   Just love your insights, sir.

[dan/california]

 

My amateur map reading implies this brings them VERY close to cutting off everything south of this. It would bring what looks like the only major road junction leading south towards Kyiv well into artillery range among other things. The Ukrainians might really manage to pinch it off

Edited March 23, 2022 by dan/california

[db_zero]

18 minutes ago, The_Capt said:

Well you hit on a pretty important point here and one modern militaries are all watching out for...what happens when we can?

Some argue that war is too chaotic and non-linear to ever be able to create effective models that provide predictive analytics; we do it now with a collection of human brains called staff, and as we see in the current example they can fail too.

Others argue that war is "chaoplexic" and not truly fully chaotic because it is bounded by some hard rulesets and frameworks (e.g. physics) and as such the irrationality of human beings can also be smoothed out or at least made more predictable (e.g. Cambridge Analytica and Big Data stuff). 

Most of us in the business don't really know but we sure are interested if that second one turns out to be true.  Why?  Because the first side to take AI/Machine Learning/Quantum Computing and use it to create effective predictive analytical models that stand up, has an enormous advantage on the battlefield.  The peace-nicks are all on about "kill bots" but this application of technology to warfare in predictive modeling has frightening implications that really give weight to the ideas of "cognitive superiority" as a applicable and measurable concept.

Lotta skeptics wave it away as "never going to happen", which it might not.  Then again there were a lot of skeptics on powered flight as well, and mechanized, and PGM, and smart weapons and cyber...so there is that.     

If we don’t blow ourselves up soon then the future is going to usher in some wild advances, quantum computation is one. AI is another and crazy stuff like Elon Musks experiment with implanting micro chips onto human brains.

Science fiction becoming science fact…I’m old enough to to remember watching Star Trek and saying those communicators are cool. We now have them- it called the smart phone.

When I bought my first board war game I thought-it would be cool if this could be done on a computer…

On another note there seems to be a mad rush to monetize this war on social media. People who never had an interest in the military or war are now experts…

There was an earlier post about the experts and their predictions before the war and how they’ve been wrong and why do we listen to them…

IMO it’s about getting your name out there and name recognition. Doesn’t matter if you’re right or wrong. If you’re name is known you’ll get invitations to speak at events and monetize your expertise.

Also the defense companies and probably their stocks are going to be quite profitable. Countries are going to spend more on defense and the supply of weapons will need to be replenished.

Also fertilizers will become more expensive and scarce and the wheat shortage not only means more expensive food but starvation in many parts of the world potentially leading to more wars and unrest.

The US military may be very busy for the foreseeable future.

Edited March 23, 2022 by db_zero

[TheVulture]

10 minutes ago, The_Capt said:

Some argue that war is too chaotic and non-linear to ever be able to create effective models that provide predictive analytics; we do it now with a collection of human brains called staff, and as we see in the current example they can fail too.

The other important point is that modelling in general (thinking in terms of scientific modelling as a whole, rather than anything military specific) is often not about "predicting the future" in the sense that most people think of it. More often, i is about seeing how outcomes change with changing assumptions and input conditions.

You might find that parameter A barely matters at all - you can change it by a factor of 10 and it makes 1% difference to the outcome. So for parameter  A, don't waste too much time trying to evaluate it precisely. While parameter B might have a large effect on the outcome for relatively small changes, which means that your prediction is only as good as your ability to measure B accurately (and tells you that you need to know all of its interactions very precisely). 

So often it isn't about predicting the future, it is about determining which the critical parameters are in your model, and what information you therefore need to be able to find out in order to make any kind of relevant prediction at all. It is about identifying the critical factors and understanding how they interact with each other.

We've all seen factors in this war that probably wasn't in many military models before, or were only just starting to be appreciated. The willingness of Russian troops to abandon important equipment. The ability of light infantry with modern ATGMs to be able to hit high value targets. The use of drones in reconnaisance, fire control and as weapon systems. Crowd-sourcing intelligence from a friendly population. Modelling can (hopefully) be used to figure out how important each of these are and how they interact with each other.

[dan/california]

2 minutes ago, TheVulture said:

The other important point is that modelling in general (thinking in terms of scientific modelling as a whole, rather than anything military specific) is often not about "predicting the future" in the sense that most people think of it. More often, i is about seeing how outcomes change with changing assumptions and input conditions.

You might find that parameter A barely matters at all - you can change it by a factor of 10 and it makes 1% difference to the outcome. So for parameter  A, don't waste too much time trying to evaluate it precisely. While parameter B might have a large effect on the outcome for relatively small changes, which means that your prediction is only as good as your ability to measure B accurately (and tells you that you need to know all of its interactions very precisely). 

So often it isn't about predicting the future, it is about determining which the critical parameters are in your model, and what information you therefore need to be able to find out in order to make any kind of relevant prediction at all. It is about identifying the critical factors and understanding how they interact with each other.

We've all seen factors in this war that probably wasn't in many military models before, or were only just starting to be appreciated. The willingness of Russian troops to abandon important equipment. The ability of light infantry with modern ATGMs to be able to hit high value targets. The use of drones in reconnaisance, fire control and as weapon systems. Crowd-sourcing intelligence from a friendly population. Modelling can (hopefully) be used to figure out how important each of these are and how they interact with each other.

Out of likes, so I will actually post that I think this is excellent.

[The_Capt]

29 minutes ago, White2Golf said:

When I think about this, I always end up wondering how they get AI to process "The Commanders Intent"?  It is not always "occupy space x".  I'm just a Sergeant, but for me that was the most important part of an order.   Just love your insights, sir.

Trust a Sgt to cut to the chase.

So Conative Models is what we are talking about here aka Human Will.  Based on all the effort and study into the fields of Marketing and Advertising the answer is definitely not "zero".  Two side to this 1) understanding and shaping an opponents Conative framework, and 2) predicting how well your own will actually do on the battlefield.  This winds up being another layer in the model overtop Cognitive and then all the physical layers...so complex.

As to #1, I think we are a lot further along than we give credit.  We have seen all sorts of big data groups being able to determine both collective and individual intent with very high levels of accuracy in everything from social media trends, spending/what they will buy, and who they will date.  Applying those advanced models to a military collective is not really that far fetched, what is hard is getting all the data all those people are giving off.  We have OPSEC and all that good stuff but we give off data like skin cells in the modern age [I heard one expert say that a switched on opponent could tell whether or not we were likely planning an offensive by looking at our waste water.  Troops would bulk up on Red Bull and that crap before a major outing].  Pulling all that data in and applying it to a predictive conative model is really not too far out there.

As to #2, well same principles but a lot trickier.  How do you measure the effect of a Comds intent on an organization?  It changes for every organization, Comd and timeframe.  I would imagine it becomes a "pressure factor" in the model that will need to be continual monitored and refreshed to be honest.  But how well intent is translated among human beings is the question here, driving how well we will likely behave in a manner that supports it.

As to how one pulls that into a full up predictive analytics model, well I would not be on a government salary if I knew that one to be honest...such a misspent youth.   Obviously the Conative framework would have to run up against Perception/Cognition as the two constantly interact.  In reality I think plumbing some learning models may yield significant fruit as this sort of thing happens a lot in that environment.  And then there is the physical predictive analytics layer, which is enormous and also filled with a lot of random effects.

My best guess is any effective model would be a bit like meteorology, very accurate in the short term and less accurate the further out one goes. Then we get into a competition to see whose model can stretch further out in time, and we are basically in a form of temporal competition.  This comes complete with c-measures and c-c-measures, which all make sense.  

So can AI/ML/whatever, understand and compute human context?  Well the answer really is a solid "maybe".  As was mentioned economic systems have not cracked it yet but I actually suspect they are much more complex than warfare.

Edited March 23, 2022 by The_Capt

[Kinophile]

57 minutes ago, dan/california said:

 

My amateur map reading implies this brings them VERY close to cutting off everything south of this. It would bring what looks like the only major road junction leading south towards Kyiv well into artillery range among other things. The Ukrainians might really manage to pinch it off

Bloody hell.

Guess we should watch out for Ivankiv in the news (in any form). That seems to be the road hub (to my untrained eye). Interestingly, if so, UA is advanceing(?) along the NORTH side of the Teteriv valley. 

 

It's wide, flat and got a nice ridge line on either side. If I was a RUS btg commander trying to retreat across this, against UKR units in the tree line...yikes...

https://www.google.com/maps/@50.8856219,29.8775028,3a,75y,246.3h,90.4t/data=!3m8!1e1!3m6!1sAF1QipMyjAkBGgCJ5aDVfaMF5FHaUUWg2-7nxVS0WOdU!2e10!3e11!6shttps:%2F%2Flh5.googleusercontent.com%2Fp%2FAF1QipMyjAkBGgCJ5aDVfaMF5FHaUUWg2-7nxVS0WOdU%3Dw203-h100-k-no-pi0-ya297-ro-0-fo100!7i8704!8i2948

I'm sure more experienced and smart military/analyst minds here have already cobbled this, but I suspect that the entire Bucha/Irpin/Makariv/Hostomel fight has been the classic Grab Them By The Balls, with the knife coming in from behind, at Ivankiv. Ivankiv can provide functionally clear LOS in a 5km, 270 deg field.

Guess what weapons system can hit out to 5km?

This entire valley is one vast tank trap. And it crosses the whole axis of Russian advance. And it only gets wider and marshier as it goes north, so any fleeing/reinforcing RUS units would need to cross a mushy, wide open shooting alley. 

If UA can hold the northern side of Teteriv, the entire RUS western attack on Kiev will come apart at the seams.

Edited March 23, 2022 by Kinophile

[danfrodo]

I do computer simulations for a living, physics-based simulations, no human decisions involved.  I'd say the issue is not AI or quantum calcs or anything that fancy.  The issue is the data entered into the simulations.  If the russians and ukrainians were modeled accurately and the simulation 'physics' could handle the small scale unit effects, then it could be modeled.  The reality is that Russia was entered as steel when it's actually cheap plastic, while the ukrainians were entered as plastic instead of steel.