Tuesday, January 4, 2022

Digital intelligence: a two-edged sword?

 

Strategy Page notes that military digital intelligence technology is increasingly making its way into the civilian economy.


Several small commercial firms are producing software systems that enables non-military users to quickly, often instantly, analyze large quantities of digital data to track trends. Developers of this software are building on decades of work in this area plus improved communications and computer technology.

For example, one of these firms ... offers what they describe as AI (Artificial Intelligence) powered software that quickly performs geospatial analysis on huge collections of commercially available data from satellite, UAV, balloon and other images, along with phone geolocation data. This rapid analysis provides accurate predictions about vehicle sales, crop yields and insurance risks.

These techniques were first developed by the military after 2001, to detect and track threats. This included terrorist attacks or the actions of irregular fighters as well as the effectiveness of friendly forces. Intelligence agencies also used it to track the strategic (wide scale, long term) capabilities of nations and their military forces. The military and intel agencies often did this by adapting commercial security software. Now there is so much commercial data available that potential users find the services ... affordable and effective for many business applications.

What made this all possible was the appearance, since the 1990s, of huge quantities of data and inexpensive supercomputers to analyze it.

. . .

Since the 1990s the U.S. has been using software to help scour satellite and aerial recon pictures for useful information. There were simply not enough trained photo analysts to examine the growing number of photos generated in the course of intelligence work. A related problem was the boredom of watching videos for hours. This problem was gradually alleviated by the use of pattern matching software that could detect movement that was in need of human attention.

. . .

By the 1990s there was a shift from analog to digital video and pictures. With digital it was easier to translate the video into numbers, and then analyze those numbers. Government security organizations have been doing this for some time but after the fact. It's one thing to have a bunch of computers analyze satellite photos for a week, to see if there was anything useful there. It's quite another matter to do it in real time. But computers have gotten faster, cheaper, and smaller in the last few years, and programmers have kept coming up with more efficient algorithms for analyzing the digital images. Commercial firms soon had software on the market that could analyze, in real time, video and alert a human operator if someone, or something you are looking for, appears to be there.

. . .

The new systems are possible because GPU data processing allows older video to be scanned faster than real time, allowing a lot of valuable information to be extracted from video taken years ago.


There's more at the link.

That's pretty impressive compared to this technology's formative years, which we discussed back in 2010.  We discussed part of its new civilian application, and its implications for our personal privacy and security, in November last year.

What concerns me is that this new technology also makes it much easier for nation-states to gain access to potentially game-changing intelligence about an enemy state, using only public-domain records.  A nation no longer needs to send reconnaissance flights over another country, or launch satellites of its own.  It can purchase, often relatively cheaply, digital images from commercial satellite operators or local surveying firms, and can buy entire libraries of past images from archives if it so wishes.  It can then analyze that data hoard using these new technologies, and gain all sorts of intelligence about target locations, methods of access, personnel traffic patterns, and so on.  This might then be used in strategic or tactical military plans and operations.

To give just one example, the Pantex nuclear warhead facility near Amarillo, Texas, is a strategic asset to the USA - the only one of its kind in the country.  If an enemy state wanted to destroy our warfighting capability, it would be a very high-value target.  Therefore, as might be expected, stringent security measures are in place to make gathering information about it very difficult.  However, what if a potential enemy state were to use modern analytical techniques on commercially available data about not just the site, but the area around it?  Pattern analysis of traffic to and from the plant would show how goods reached it and/or were dispatched from it, and the rail and road connections it used.  The vehicles used by its personnel could be tracked to and from their homes.  Those connections and/or personnel might be targeted by unnamed, unidentified "terrorists", thereby causing crippling indirect damage to the Pantex facility, without any overt act by a nation-state that might result in its identification, or retaliation against it.

The same thing applies to commerce and industry.  A few weeks ago we looked at China's worldwide logistics software, and how it provides that country with competitive advantage and priceless commercial intelligence.  I'm sure its government uses this sort of analytical software on the information thus obtained, applying it to its geopolitical advantage.  Frankly, they'd be fools not to do so - and the Chinese are emphatically not fools.  Are we doing the same thing?  I hope so, and I'm sure commercial firms here are doing so:  but our administration is currently so focused on political correctness and partisan infighting that I can't help but wonder whether official bodies are doing all they could in that regard.

All in all, these commercial applications of digital intelligence analysis strike me as very much a two-edged sword.  We can use them to our advantage, but our enemies are doubtless using them for theirs as well;  and there's usually far more freely available digital information to analyze about our open society than there is about more closed ones.

Food for thought.

Peter


4 comments:

  1. Once Pandora's box was open there was nothing anyone could do to get everything back in... The ideas related to modern AI and machine learning concepts originated in the 1956 Dartmouth AI Research Project and relies on the ideas proposed by Turing, Wiener and Shannon. These guys, like a lot of other scientists before, came with a set of ideas not considering the implications of their work will have on the societal framework. I guess Einstein was right when he stated that: "Technological progress is like an axe in the hands of pathological criminal." How often have idealist scientists chased the dream of progress, only to see that dream perverted? Nothing new under the sun from that perspective. What we need to consider and get used with is the fact that our everyday environment is recorded up t the last detail and we are what the German media dubbed "the glass citizen" - fully transparent in our everyday life...
    That is the bigger problem you and everybody else should consider... before the Chinese get to attack you you should consider how much of your everyday life your own government wants to control, how much of your privacy you still have and what does it means for you as an individual in terms of individual freedom and preservation of individuality.
    You should be concerned with what your own government does in gathering data on your everyday life and with how it uses this data. Is it moral to gather everyday data on the citizens? Well, it depends on the standards, but will the morality of that action stop the authorities gathering and analyzing them? Not a chance in hell! Once thy get the data, they are NOT going to release the power it gives them over the society!
    So, is the fact that advanced data analytics, machine learning and artificial intelligence will provide foreign powers with too much information a problem? Just a secondary bother, since the bigger problem is the fact that your own government is much more concerned with steering what you think and what you do...

    ReplyDelete
  2. Spot on. The Chinese have been interested in anything and everything for quite some time.

    ReplyDelete
  3. Pretty bog standard stuff these days. Every hedge fund does it. And has been at it for a while now.

    If your competitor is relying on official government statistics for (say) the amount of oil in storage and *you* are using AI to analyze the imagery you bought of every tank farm to find out how far raised or sunken the floating 'lid' thing for each tank is -- thereby giving you the volume of oil in tank... Well you do that by fully automated means as often as you need and you eat your competitors' lunches.

    The fun is in figuring out new correlates of this or that economic indicator or asset to apply this to.

    No accident that the Chinese have just cracked down on public availability of automated ship tracking data for their ports and coastal shipping.

    ReplyDelete
  4. Sorry but not sorry because have to say it.

    The Open Society is heaven for Foreign Actors and Remoras.

    Of course the Turkey Middleman is going to write a book promoting Thanksgiving.

    Chinese are just the latest in a long line. Of course safer to point finger at them and much less Pavlovian conditioning against the finger pointing.

    ReplyDelete

ALL COMMENTS ARE MODERATED. THEY WILL APPEAR AFTER OWNER APPROVAL, WHICH MAY BE DELAYED.