Blog
/
Intelligence
Pick a Pocket or Two
Your attention is constantly being exploited, its not easy to detect and harder to stop.
Posted at
Posted on
Intelligence
Stealing attention and driving behaviour online
Manipulation is not mind control. It is attention control. A pickpocket removing a watch and a fraudster talking someone out of their savings rely on the same thing, and it is not that the target was careless or stupid. It is that human attention has consistent, measurable and well documented limits. Everyone has them.
Apollo Robbins first came to notice as the man who picked the pockets of Secret Service agents while entertaining former President Jimmy Carter, leaving with the President's itinerary and the keys to his motorcade. He was later profiled in the New Yorker by Adam Green, who noted that neuroscientists had begun studying his methods [1, 2].
How Attention Fails
Attention works like a spotlight that can be moved independently of where the eyes are pointed, and its capacity is finite. Widening the beam dims it. You do not experience the world, you experience the portion of it inside the beam, and nothing tells you how much has been left out. In the classic inattentional and change blindness experiments, around half of observers fail to notice substantial changes happening in plain sight [3, 4].
The steering can be measured. Robbins believed curved hand movements misdirected better than straight ones, because a straight line invites the eye to jump ahead and back. The experiment confirmed it [5]. Curved motion held the eye in continuous pursuit. His working intuition was precise enough to be tested and published.
Choice, memory and compliance behave the same way. Johansson and colleagues covertly swapped the photograph a participant had selected as more attractive, and most people not only failed to notice but confidently explained a preference they had never held [6]. Loftus found that witnesses asked how fast cars were going when they smashed gave higher estimates, and more often recalled glass that was never broken, than those asked when the cars hit [7]. Freedman and Fraser showed that agreeing to a trivial request makes a person markedly more likely to agree to a large one later [8].
None of this requires a lie. A well built manipulation supplies true statements in a sequence that produces a false conclusion, and lets the target assemble most of it.
The Industrial Version
Advertising is the same discipline, funded, measured and run at scale, and its research literature is correspondingly large.
Robert Zajonc demonstrated in 1968 that simple repeated exposure to something increases liking for it, with no argument made and no claim advanced [9]. Later meta-analysis by Bornstein and D'Agostino found the effect was stronger when people did not consciously recognise having seen the stimulus before [10]. Hasher, Goldstein and Toppino showed the parallel effect for belief: a statement repeated is rated as more likely to be true than the same statement heard once, regardless of whether it is true [11].
The most useful finding for our purposes runs against intuition. Robert Heath's low attention processing model [12], building on earlier work by Krugman [13] and Ehrenberg [14], holds that advertising has more emotional and behavioural effect when it receives less conscious attention, because attention is what invites scrutiny and counter-argument. The title of the 2006 paper he published with Brandt and Nairn states the position plainly: brand relationships are strengthened by emotion and weakened by attention [15].
The aim, then, is not always to capture attention. Often it is to stay beneath the threshold at which someone starts examining what is happening.
Why This Matters for Research
An online recruiter, whether working towards fraud, radicalisation or exploitation, is running the same operations Robbins runs. Narrow the beam. Move it away from anything that would widen it, friends, family, any other source of context. Escalate commitment in steps small enough that none feels like a decision. Supply the interpretation, so the target reaches the conclusion themselves and remembers doing so alone. Repeat rather than argue, because familiarity builds trust without making a claim that could be challenged. A private message thread is close to ideal conditions for all of it.
The signature is therefore in the sequence, not the message. Any individual message looks unremarkable, because it was built to stay beneath the threshold at which anyone starts examining it. That is Heath's finding turned to a different purpose, and it means snapshot collection finds nothing. The pattern appears only across weeks of exchange, which requires sustained presence rather than a visit.
Kuro addresses two parts of this. The first is access. Researchers need to be present credibly, and for long enough to watch a sequence develop, without their presence changing what they are looking at. The second is scale. Sustained presence produces more material than any analyst can hold in attention at once, and the limits described above do not stop applying because someone is being paid to look. An analyst reading months of exchanges has a narrow beam and a finite capacity, inside material built to steer whoever reads it, sometimes by operators who know researchers are present.
Kuro's vision-capable language model is built for that gap. It does not share human attentional limits. It can review the same session as many times as required, assess a single post, a whole profile, or a profile's behaviour across months, and read imagery alongside text, which matters where much of the signal sits in screenshots and video stills rather than in what is written. Its beam is wider, its attention holds for longer, and it can be pointed at the sequence, not the message.
It supports analyst workflows rather than replacing them, surfacing what an analyst may have passed over, flagging patterns that appear only across time, and pointing attention back to whatever a subject worked to move it away from. The judgement remains the analyst's. The difference is that it is now made with everything in view.
References
Green, A. (2013) 'A Pickpocket's Tale', The New Yorker, 7 January. https://www.newyorker.com/magazine/2013/01/07/a-pickpockets-tale
Macknik, S.L., Robbins, A., Martinez-Conde, S. et al. (2008) 'Attention and awareness in stage magic: turning tricks into research', Nature Reviews Neuroscience. https://www.nature.com/articles/nrn2473
Simons, D.J. and Chabris, C.F. (1999) 'Gorillas in our midst: sustained inattentional blindness for dynamic events', Perception.
Simons, D.J. and Levin, D.T. (1998) 'Failure to detect changes to people during a real-world interaction', Psychonomic Bulletin & Review.
Otero-Millan, J., Macknik, S.L., Robbins, A., McCamy, M. and Martinez-Conde, S. (2011) 'Stronger misdirection in curved than in straight motion', Frontiers in Human Neuroscience. https://doi.org/10.3389/fnhum.2011.00133
Johansson, P., Hall, L., Sikström, S. and Olsson, A. (2005) 'Failure to detect mismatches between intention and outcome in a simple decision task', Science.
Loftus, E.F. and Palmer, J.C. (1974) 'Reconstruction of automobile destruction', Journal of Verbal Learning and Verbal Behavior.
Freedman, J.L. and Fraser, S.C. (1966) 'Compliance without pressure: the foot-in-the-door technique', Journal of Personality and Social Psychology.
Zajonc, R.B. (1968) 'Attitudinal effects of mere exposure', Journal of Personality and Social Psychology.
Bornstein, R.F. and D'Agostino, P.R. (1992) 'Stimulus recognition and the mere exposure effect', Journal of Personality and Social Psychology.
Hasher, L., Goldstein, D. and Toppino, T. (1977) 'Frequency and the conference of referential validity', Journal of Verbal Learning and Verbal Behavior.
Heath, R. (2012) Seducing the Subconscious: The Psychology of Emotional Influence in Advertising. Chichester: Wiley-Blackwell.
Krugman, H.E. (1965) 'The impact of television advertising: learning without involvement', Public Opinion Quarterly.
Ehrenberg, A.S.C. (1974) 'Repetitive advertising and the consumer', Journal of Advertising Research.
Heath, R., Brandt, D. and Nairn, A. (2006) 'Brand relationships: strengthened by emotion, weakened by attention', Journal of Advertising Research. https://doi.org/10.2501/S002184990606048X
Kuro supports lawful intelligence and investigative research for government agencies, law enforcement, journalistic and accredited private sector organisations. All use of the platform is subject to Kuro's Acceptable Use Policy and applicable legal frameworks.
