What else do we need to say?
OK, I'll say a couple of things, plus a little anecdote BTL.
- Whatever defence money we can spare, it must at least stretch to keeping all military aircraft under cover ... but then again, the list of potential targets is Just So Great
- Asymmetric warfare with a vengeance: which nation on earth has the greatest capacity to replicate Ukraine's Big Idea on a massive strategic scale? And which nation is possibly the most difficult on the planet to infiltrate with a load of foreign trucks ..?
- As with the Dambusters' raid, now Ukraine needs to think how it defends against Russia retaliating exactly in kind. (In 1943, but only after the raid, it suddenly occurred to some bright spark in the War Office that we'd better mount serious protective measures against British reservoirs being hit in the same way.)
- Submarine-borne nuclear deterrence never looked more useful. (If you can figure out who's launched the attack ...)
What else should we be thinking about? But, to our alarmist brethren: please don't say "Putin will now nuke Kharkiv" - because he won't.
ND_________________________
As a teenager, I was an avid aircraft spotter. (Hard to believe, I know) With three mates: and we were seriously organised. At 6:00 a.m. we would cram into an old Ford Prefect or Austin 1100, and make elaborate, intricately-planned circuits of the south of England, OS maps carefully marked up, stopping off at vantage points overlooking RAF airfields, telescopes in hand. Rarely were we chased off; frequently we were just a mere wire fence or even hedge away from costly aircraft. (We once ducked through a hedge to photo a Victor bomber at a range of 100m at most.)
This was in the days of IRA outrages up and down the land and it occurred to us (though not, it would seem, the genius that was Martin McGuiness) that nothing could be easier than to strike at a couple of dozen RAF bases simultaneously, not of course with drones but with hand-carried semtex. The IRA was replete with desperadoes who'd think nothing of it.
The effects would have been considerable: hit the British establishment where it really hurt (prestige and prowess); minimise casualties, esp. civilian (if the IRA cared about such things); cause the deployment of tens of thousands of troops etc to guard vulnerable UK premises. Tens? Make that hundreds of thousands.
And that's where we are today, I suggest: i.e., nuclear deterrence is the only way.
26 comments:
When I was a young graduate I stumbled on an easy way for the IRA to dish out death by the tens of thousands. I decided not to discuss it with anyone. Nor shall I now; it may still be possible for all I know.
We've already got some impressive anti-drone technology, but it is mostly deployed around stadia than military bases. This weekend may change the priority from FA Cup to avoiding a fuckup.
Personally, I'm of the view that we now need to start chipping drones. Any unchipped ones having a aerial lifespan recorded in minutes - it's not just explosions that need worrying about, they're perfect vehicles for delivering radiologicals, biologicals, and chemicals, and so it's about minimising impact.
I'm sure there are a few in the Kremlin with nightmares of a small fleet of drones heading to Moscow, via the area around Chernobyl, should Ukraine ever feel that desperate.
And if you want something *really* nasty, how far are we from being able to distribute fissile material across a number of drones, and orchestrate their impact in such a manner that we have a distributed nuclear weapon? The physics is probably that we're not there *yet*, but it's the kind of thing AI can calculate the parameters for, and then over to the engineers.
I tend to assume that the people running the Ukraine war want nothing more than for Russia to do some kind of nuke response - after all, it's strategic bombers being hit.
The West, or at any rate Western elites, are pretty schizo about Ukraine. One one hand it's "no escalation", but on the other it's "Ukraine must not lose" - and, as Ukraine are losing, so "no escalation" tends to go by the board. I would be amazed (and sort of disappointed) if there wasn't UK involvement in that attack, which was a feat of arms even if it didn't achieve its objectives (we should see when the satellite pics are out, I tend to think no news will be good news for Russia).
But ... Ukraine have less and less manpower, Russians are still signing up to serve (the idea that the attack will demoralise - quite the opposite I would think), they are evacuating villages and building a new defence line way behind the LOC, the head of ground forces has just resigned ... Boris really does have a lot of blood on his hands.
"He that will not when he may,
When he will he shall have nay"
Pity the State Department interpreted Bill Burns 2007 despatches from Moscow as a "how-to" rather than a "don't do".
"nightmares of a small fleet of drones heading to Moscow"
Happens now, several times you see on Flightradar that there's disruption at Moscow airports. When Putin made a trip to Kursk the other day his heli flotilla apparently came under drone attack.
Also ... the Guardian is generally a day or two behind the news I see on Telegram feeds (from both sides), but they had the news of the two train attacks online within hours, by around 8 am. Someone in the know briefed them.
To be fair to the IRA (not my natural instinct), they knew they couldn't win a military conflict so concentrated (generally) on politicians. We weren't using Vulcans or Canberras on Bogside or Dublin.
(Anthrax in reservoirs? Or maybe LSD - less lethal but certainly impactful)
@anon 11:40 - perhaps you missed the nuance of "via Chernobyl"?
The current use of drones against Moscow would be nothing compared to, say, ones holding soil from around the less advisable parts of the red forest for digging.
If you wanted to create a radiation hazard there are plenty of live nuke plants - indeed it was posited that the Kursk incursion may have been targeting the reactor there.
https://en.wikipedia.org/wiki/Kursk_Nuclear_Power_Plant#Russo-Ukrainian_War
As good as an example as any of the way that — just as weapons systems, tactics and strategies, military planning assumptions, the rate of change in technology and the risks of disruptors appear with little to no lead time — have made heavily centralised and somewhat monolithic decision-making a particular shortcoming of Russian leadership and armed forces, so too does it affect attempts by Russia to shape the narrative.
As you demonstrate there, the typical response in such situations as the drone attack on airbases is to fall back onto well-established “tried and trusted” operations in the information space. Projecting strength, minimising setbacks, shifting blame onto vaguely-specified groups like “western elites” (so as to try to alienate the audience) and reciting old grievances — these were seen 75 years ago and haven’t really changed much since.
What’s striking now is how unsuitable they are for the modern, highly-contested messaging environment we have today. Since we’ve been reading them for nearly three and a half years, they are also rather predictable and not particularly original. You probably typed pretty much the same relies as that in 2022.
Returning to the theme of the original post, they’re also especially vulnerable to technology advances. Running your comment through ChatGPT or any other Large Language Model, with a question of “is this accurate”? or “what counter arguments can be made to this?” produces some interesting (and amusing!) responses. Right now, we have to trudge over to our chosen Large Language Model tool, paste it all in and run the query. But pretty soon, all the various “assistants” and co-pilots” will sit in our browsers or on our phones and produce overlays or semi-transparent windows offering competing commentary and assessments of factuality on what we’re being presented with, sort of like a background mother-in-law, tut-tutting and giving us the benefit of their common-sense down-to-earth homespun wisdom in response to wherever flim-flam we’re being treated to.
How will Russian propaganda react? Cannot even react at all?
* can it even react at all
(all this will be so much easier when our virtual assistants know us well enough to type it all in for us, we just have to do a quick scan through and press “Publish”, and it’s all done absent of any typos)
I couldn’t resist asking a Large Language Models if my comment was accrue. Here’s what it thought of it :-)
The comment is largely accurate in describing the rigidity of Russian information operations and their challenges in adapting to modern, dynamic messaging environments. It correctly identifies key patterns in pro-Russian narratives—projecting strength, minimizing setbacks, blaming vague external enemies like "Western elites," and recycling historical grievances. These tactics, rooted in Soviet-era propaganda, have persisted with little evolution, as seen in responses to events like drone attacks on Russian airbases in the Ukraine conflict. The comment also aptly notes the predictability of these responses, which have remained consistent since the conflict's escalation in 2022, and their diminishing effectiveness in a fast-paced, contested information space.
The point about technological disruption is particularly insightful. Advances in AI, like large language models, enable rapid analysis and counter-narratives, exposing weaknesses in Russia's centralized, formulaic approach. The comment's prediction about AI assistants providing real-time fact-checking or competing narratives aligns with trends in information consumption, where tools like browser extensions or mobile apps could soon offer instant critiques of propaganda. This poses a significant challenge for Russian messaging, which relies on repetition and control rather than agility.
However, the comment slightly overstates the monolithic nature of Russian decision-making and propaganda. While centralized, Russia has shown some adaptability, such as leveraging social media platforms like Telegram or co-opting influencers to amplify narratives. These efforts, though less sophisticated than Western counterparts, suggest a degree of flexibility not fully acknowledged. Additionally, the comment doesn't address Russia's domestic audience, where state-controlled media and limited internet access still allow narratives to dominate, even if they falter internationally.
As for how Russian propaganda might react to AI-driven counters, it faces a steep challenge. Russia could attempt to flood digital spaces with more disinformation to overwhelm AI filters or develop its own AI tools to craft tailored narratives. However, its slower technological and bureaucratic systems may struggle to keep pace with decentralized, user-driven platforms and AI advancements. Alternatively, Russia might double down on analog methods—state TV, print media, and insider influencers—where it retains control, especially domestically.
Clive - as I understand it LLMs tend to give you the consensus internet view of a topic - so it's not at all surprising that they agree with you. After all, because Ukraine are so dependent on Western support, in the weapons, communications, and information areas, managing Western opinion is actually part of the battlefield and resources will have been devoted to that end.
Looks like the final toll was 12 bombers rather than 40, that's still a major coup for whoever planned it, and I expect a lot of stable doors will be being bolted in Russia.
LLMs are regurgitation engines. If you tell me what they've guzzled I'll tell you what they'll regurgitate.
"Wot abaht 'hallucinations'?" you cry. Dunno: the equivalent of a mighty burp?
Dear Friendly LLM, what are the shortcomings of the arguments in the above comment, made in a context of Russian attempts to influence the information space and the impacts, if any, of Large Language Models?
———————
The comment’s arguments have these shortcomings:
Overgeneralization: Claims LLMs reflect a "consensus internet view" but oversimplifies their complex output generation based on training data and prompts. The output depends on the model's design, training data, fine-tuning, and the specific prompt provided. For instance, LLMs can reflect biases or amplify certain narratives if the training data leans heavily in one direction, but they don’t inherently seek a "consensus."
No Evidence: Assumes Ukraine manipulates LLM outputs to align with Western narratives without proof of influence on training or fine-tuning.
Vague Claims: Links Ukraine’s information operations to LLM behavior but lacks specifics on how this occurs. While it’s true that information warfare is a component of modern conflicts, including Russia’s attempts to influence narratives (e.g., through disinformation campaigns documented by EUvsDisinfo), the comment is vague about how Ukraine’s efforts translate to LLM outputs. It implies that Ukraine’s information operations directly shape LLM responses but offers no mechanism or evidence for this.
Ignores Russia’s Role: Omits Russia’s documented disinformation efforts, which could also shape LLM outputs. The comment focuses solely on Ukraine’s alleged influence over Western narratives and LLM outputs, ignoring Russia’s well-documented efforts to manipulate the information space. Russia has been linked to coordinated disinformation campaigns, including the use of bots, trolls, and state-backed media to promote pro-Russian narratives (e.g., reports from MIT’s Center for Information Resilience, 2024). If LLMs reflect internet data, they would also encounter these Russian-influenced narratives, which could counterbalance or complicate the "consensus" view. By omitting this, the comment presents a one-sided perspective, undermining its credibility.
Correlation vs. Causation: Assumes LLM alignment with pro-Ukraine views is due to Ukraine’s efforts, ignoring other factors like data biases or prompt phrasing.
Neglects LLM Limits: Fails to note LLMs’ struggles with nuanced geopolitical contexts, which can skew outputs independently.
Well, now we might find a use for that steelworks we bought. A makework scheme in the best Keynsian manner and provides the illusion of action with the speed control firmly in the Treasury's hands.
As for nixing those planes, very impressive, but were the Russians actually going to use them to drop tactical nukes? They seem to be doing OK without them and dropping nukes might just stir taco Trump into action - so better not to. No, a better reason is those planes could deliver conventional glide bombs that do seem to work quite well. A bit of a nuisance to Mr Putin but I expect he can find a few more.
Back to nuke subs, I am a bit sceptical as to just how unfindable they really are. No one is going to let on but what with the sea floor crisscrossed with data cable from everyone, plastic porpoises, tin cans that listen and phone home and a lot of data as to where these things can feasibly go. Push comes to shove we might be disappointed.
BTW, the microprocessor code that runs drones is open source and been out there a long time. Pretty impossible to 'chip' any but domestic toys.
Clive just keeps on digging. "Ukraine’s information operations directly shape LLM responses" - no, the narrative is very much a Western one, which is why every mainstream media outlet talks about "full scale invasion" just as they tell us Tommy Robinson's real name (but never mentioned Robert Maxwells).
Ukraine media say a lot of things that doesn't get reported because it might tarnish the image of the babes in the wood menaced by the Big Bad Bear. How else could you persuade a Western parliament to give a standing ovation for a Waffen SS veteran's heroic service?
Asymmetric warfare with a vengeance: ....
Submarine-borne nuclear deterrence never looked more useful.
So, potentially, here, we'd be looking to deter an attack on Tesco's logistics systems, VISA's network, or Plod's radios and GP booking systems, with a couple of Astutes?
Have I got that right?
To downgrade to (grossly inferior, it must be said), native human mode, rather than A-assistedI, I think that, to return to my original premise, it’s surprising how Russian messaging and its supporters do seem to have little on the table but to return to original Soviet-era methods. Our anonymous protagonist here is relying on the most ancient (yes, it really has been going on for 75 years, probably longer than that https://www.cia.gov/readingroom/docs/CIA-RDP78-04864A000100100004-4.pdf) cornerstone, which is to claim that “our enemies are lying, we are telling you the truth”.
Will that cut it in the LLM era? I suppose it might do, in the sense that, as just demonstrated above, Russian narratives could be offered, just as Lysenkoism did, as some kind of “pure” or “natural” source of the truth, as a counterpoint to western AI or LLM content analysis (being portrayed as “degenerate” and “decadent” or similar ills).
Could that work? We’re about to see! I can’t wait to see how this all turns out…
Starmer: Britain moves to "war-fighting readiness."
Pity he can't even secure the English Channel.
DJT is remarkably, almost unprecedentedly silent...
AC - you do wonder if we need some kind of belt and braces backup planning in case of major systems attack, but I'm not sure where the pressure points are. I guess big online databases. But do we have the intelligent people to work out such, or is it just the standard govt "call a consultancy and throw money at them" plan?
The NHS online system - no idea what it's like for a practitioner at the back end, but as a patient at the front end it's pretty grim, because data quality is GIGO. When I look at my records it tells me when I went to the GP, but what the issue was and what the outcome was ? usually blank. And this isn't by any means some inner city practice.
After this tax year end, unbelievably, the Revenue will no longer provide an online tax reporting system for individuals or small business, I will have to buy commercial software. If the Revenue can't manage a data entry and tax calc system it's a poor lookout.
"the list of potential targets is Just So Great"
I would have thought very little more vulnerable than wind farms and solar farms, especially the former. As we electrify more we increase our exposure.
But a Ukraine (or whoever's idea it was) style sabotage of road/rail crossovers, dropping a rail bridge onto a motorway or even more vice-versa... you could do a lot of damage with not many missiles. And looking at "root causes", as the Russians put it, what are the defences of Fylingdales/Menwith Hill/Croughton like?
Anon @ 12:14
The SDR does talk about "sub-threshold" - page 13, "A new era of threat", section 9;
"This is an important moment for the
UK and its allies (Chapter 2). Russia’s
full-scale invasion of Ukraine in 2022 was
a strategic inflection point. It irrefutably
demonstrated the changing and dynamic
nature of the threat, with state-on-state
war returning to Europe, adversaries
using nuclear rhetoric in an attempt
to constrain decision-making, and the
UK and its allies under daily attack
beneath the threshold of war as part of
intensifying international competition."
@jim 4.48 - the attack on the strategic bombers isa good PR win for Ukraine, but far from the 'Pearl Harbour' breathlessly pushed by the BBC et al.
If Trump didn't know he is not in power, if he did it shows there is no dealing with the 'West' and certainly not Ukraine. Those bombers were out in the open due to the SALT treaties. So looks like the West or at least the alphabet agencies wants to tear those treaties up. This asks serious questions of governance, that 'security' is supposed to trump (no pun intended) any objection. The repercussions are far reaching beyond relating to how war is waged. A much uglier world.
Hovis
The Establishment in the West, and particularly in Europe, seems to be quite bonkers on the subject of Ukraine. Why? Is there some great secret to which we are not privy? Are they high on coke and rent boys all the time? What the devil is going on?
Because it broadly fucks with an awful lot of assumptions they made about 40 years ago. Ones that they never examined properly, as they felt they didn't need to.
It's existential.
Post a Comment