The Register Home Page

back to article Live and Let AI: Former CIA officer says human spies matter more in the LLM age

The bots won't be coming for 007's job anytime soon. According to a former CIA officer, AI may help create false documents, but this fakery will give old-fashioned human intelligence fresh relevance. Former CIA case officer (case officers recruit and handle foreign agents overseas) and RAND Corporation researcher Thomas …

  1. Bebu sa Ware Silver badge
    Big Brother

    New Bond titles ?

    Lie and let AI

    Tomorrow Ever AI

    For Your AI Only

    AI another Way

    No Time to AI

    1. Rjan

      Re: New Bond titles ?

      That could go on forever:

      From Russia with AI.

      The spy who loved AI.

      The man with the golden AI.

      Never say never AI.

      No time to AI.

      The living AI-lights.

    2. Irongut Silver badge

      Re: New Bond titles ?

      How have you not suggested GoldenAI?

  2. Anonymous Coward
    Anonymous Coward

    Fair enough

    Makes sense that human intelligence (HUMINT) should remain the gold standard in this field imho. Same with FBI and courts where there's been nefarious pushes to replace experienced human analysts and judges with outstandingly inept activists of loyalist ideologies of late, that behave like obedient buttlicking robot slaves.

    Vibe harness coding should eventually come to a similar realization that the only way to properly cut through the gooey sloparooni sludge of thick-basted pretend programming from artificious misintelligence is to apply human brain cells to it in the first place, even in cases of "web-based RSS. sort of app" and related Kettle(s) ... ;)

  3. billdehaan Silver badge

    We learned this 25 years ago

    Just after 9/11, intelligence officials placed much of the blame on the loss of HUMINT (human intelligence) over the previous years. The HUMINT assets that had been deprioritized in favour of SIGINT (signal intelligence). SIGINT are the things like satellite photos, computer analysis, intercept decryption, and things like that.

    Just before 9/11 happened, I remember a very pro-SIGINT co-worker boasted that with the newer recon photo capabilities, the technology could identify the licence plate of a car in Baghdad in seconds, something that would take HUMINT resources hours, or longer.

    A second co-worker pointed that was great, but "we're not going to go to war against licence plates".

    AI may be able to perform mechanical tasks better than humans, but they're not going to pick up on nuance. And in a profession where subtlety is key, being able to detect nuance is key. Humans don't always get it right, either, of course, but I trust AI even less.

    As an example of how important nuance is, consider Kruschev's (in)famous "we will bury you" speech. When translated as "we will outlive you", or "we will outlast you", it's a boast claiming that communism will be proven to be superior to capitalism. When translated as "we will bury you" (as it was), it's a threat. If an AI translates it as "we are going to nuke you", just imagine the resulting fun.

  4. amanfromMars 1 Silver badge

    No Small Step for Man whenever a Giant Quantum Communications Leap for ITs ManKind

    Humans won't just be necessary for intelligence work, Mulligan said. He lays out a case that AI will make human intelligence officers essential as truth becomes harder to distinguish from computer-generated fiction in the coming years.

    Is it not surely currently the present truth that human intelligence tales are being revealed by AI as historical hysterical fictions servering an oppressive virtual reality rendering the Earthly existence for greater humanity in the perverse coercive command and control service of a sociopathic few aided and abetted by a collaborating quisling media operations machine and its necessary brainwashed and brainwashing puppet performers and muppet star turns for your daily dose of oppressive news with myriad extensive comfortable life threatening views ‽ .

    And therefore, for the foreseeable coming years will the future be a brighter novel and better more enjoyable SMARTR AI generated fiction with virtualisable realities for humanities which relentlessly successfully target and exterminate any possibility of the Earthly existence of a sociopathic few in perverse coercive command and control services aided and abetted by a collaborating quisling media operations machine and its puppet performers and muppet star turns ‽ .

    And that is the least that can be expected .... and with a great deal more totally unexpected to quickly follow for y'all to follow, are exciting times in CyberIntelAIgent Spaces guaranteed ahead. I Kid U Not.

    1. Anonymous Coward
      Anonymous Coward

      Re: No Small Step for Man whenever a Giant Quantum Communications Leap for ITs ManKind

      Yeah, if "human intelligence tales are being revealed by AI as historical hysterical fictions" is meant to suggest that humans' enthusiasm for AI reveals how we're in fact not very smart to start with, as a group, then I have to agree. But it's confusing because this here TFA is focused on intelligence as relates to espionage and spying, 'Get Smart' style and such, a different axis from the one that goes from numpty to genius (iiuc) ...

      In this space, I don't think AI has revealed HUMINT to be fiction, at all. In fact, SIGINT/ELINT without HUMINT would be just like a bunch of suckers without a hectocotylus imho (or vice-versa) -- not very tangible!

  5. Pascal Monett Silver badge

    "fresh relevance" ?

    HUMINT has always been essential.

    It has only been eclipsed by SIGINT because politicians like the idea of having a picture taken from space in their hands. It must seem more real than a boring verbal or text report about how agent XYZ overheard a conversation in a bar that indicated that country so-and-so was progressing in its uranium refining.

  6. Anonymous Coward
    Anonymous Coward

    "Humans won't just be necessary for intelligence work"?

    Is it just me or is that a very weird-sounding sentence?

    1. amanfromMars 1 Silver badge

      Re: "Humans won't just be necessary for intelligence work"?

      No and yes are the answers to those two questions you have asked, AC .... "Is it just me or is that a very weird-sounding sentence?"

      Would it sound equally weird, and be of itself even weirder, but also be much wiser to be prepared for a very different and surprisingly novel and strange ennobled future, to think that AIMachinery might rather autocorrect what one may say and in so doing share insights into the darker sides of its analytical processes which will result in likely autonomous executive action[s] that can easily directly threaten the course of human existence, because if IT were to state ....... “Humans just won't be necessary for intelligence work” rather than “Humans won't just be necessary for intelligence work” .......... is that a Quantum Communications Leap into a vast new series of Brave NEUKlearer HyperRadioProACTive Neurolinguistically Reprogrammable Worlds in which one can both be Enabled to Embrace and Enjoy, and Work, Rest and Play in.

      1. Rich 11

        Re: "Humans won't just be necessary for intelligence work"?

        Well, if anyone knows how to recognise a very weird-sounding sentence, it's you...

POST COMMENT House rules

Not a member of The Register? Create a new account here.

  • Enter your comment

  • Add an icon

Anonymous cowards cannot choose their icon