I sampled a bunch of pop-sci books on the effects of technology on the mind. Most were shallow and anecdote-heavy. This one proposes a concrete mechanism governing behavioural changes and puts in the context of current understanding of cognitive control. Everything in these notes appears to be pretty well supported by both correlational studies and controlled interventions. Not one of those books that weaves a beautiful story out of a single crappy experiment.
Both authors are professors researching this topic. First author founded a neuro-therapy company.
Two main sources of behaviour. Seems to map roughly to dual process theory? Bottom-up - perception-action cycle. Top-down - goal creation and enactment.
Different strengths and weaknesses, not clear which system should have control. Eg trying to concentrate on a problem while being distracted by traffic noises vs trying to concentrate on a problem while ignoring the oncoming car.
Conflict between the two is mediated by executive functions. Main components are working memory, goal management and attention.
Working memory. Maintain internal representations and perform operations on them. Bridge - allows longer delays in perception-action cycle. Eg recalling last observed position of predator, plotting their possible movements, planning a course around.
Goal management. Scheduling both internal and external behaviour. Directing mental resources. Task-switching, including hierarchies of sub-tasks for complex behavior. Multi-tasking at this level is not possible - executive functions are single-threaded - rapid task-switching instead.
Attention. Complex, not well-understood. Spotlight. Focus on specific signal/pattern, area, moment in time etc eg throw a stone and watch for movement in the long grass when it lands. Ignore/suppress irrelevant stimuli. Focus can be spatial, temporal, specific pattern etc. Explains suppression as a way to conserve resources for more important tasks, but later mentions that suppression is cognitively demanding. ‘Resources’ is vague. It’s not at all clear to me right now how focusing and ignoring work. Setting priors for pattern recogntion? Discarding unimportant signals earlier in the pipeline? Suppressing the results of processing from intruding on working memory? Focusing and ignoring appear to be different processes, can fail independently eg ignoring decays much faster with old age.
Interference. Internal vs external eg decide to check email vs hearing notification. Distraction (resisted) vs interruption (endorsed) eg hearing notification and ignoring vs hearing notification and checking.
Bottom-up processing prioritizes by novelty and saliency (where saliency can depend on learned associations). Can top-down processing prime saliency?
Many limits on executive function.
- Selectivity can be overruled by bottom-up distractions
- Distribution of attention diminishes effectiveness eg focusing on 50% of visual field vs focusing on one exact spot
- Sustainability of attention is limited. By what? Resource exhaustion? Explore vs exploit?
- Switching speed - takes time to move attention. Attentional blink eg can miss the second of two visual events when they occur closer than ~0.5s.
- Capacity of about 4+-1 ‘chunks’ (numbers vary depending on the type of task and on what we consider to be a single item)
- Fidelity - representations decay over time - exacerbated by interference
- Very limited ability to parallel process on attention-demanding tasks (eg driving and talking on the phone has higher costs than driving and listening to music, because music doesn’t demand as much attention)
- Overhead costs from task switching
Executive functions follow a U-shaped curve peaking at 20-25yo. But doesn’t mention any longitudinal experiments - could be generational effects too? Also not clear how much is physical degradation vs disuse.
Sleep deprivation has huge impacts on executive function.
Time to 50m users:
- radio 38y
- telephone 20y
- tv 13y
- cellphone 12y
- internet 4y
- ipod 3y
- myspace 2.5y
- facebook 2y
- youtube 1y
- angry birds 35d
Uptake of new tech is accelerating.
Not sure it makes sense to compare angry birds to tv rather than eg a particular tv show, but general point seems sound.
Huge increase in frequency and range of tech use. Numbers highest in younger generations. Again no longitidunal, but seems likely.
Lots of studies to set the scene. I’ll just note a few interesting examples:
- Nomophobia - fear of being without a mobile phone. ~50% of young adults.
- Phantom pocket vibration syndrome. ~100% of young adults.
- In one office: ~5m between interruptions and ~1m response times to messages.
- Belief in multi-tasking ability is high. Actual performance is not. Ability to predict actual performance is varied but generally poor.
- Attentional blindness. Walk+phone accident rates in A+E are soaring. Apparently not distractible enough?
- Strangers conversing report lower levels of a closeness, trust, empathy and understanding when there is a phone in indirect sight vs a paper notebook.
- Correlational evidence for media use vs sleep deprivation. (Mayo clinic recommends dim screen at >14” from face should be ok. Does that mean frontlit Kindle is ok?)
Performance degradation is higher for social media and messaging than for other tech. Hypothesis - social interaction is a much stronger attentional priority than entertainment - matter of survival.
This section seems out of place, but still interesting.
ADHD sufferers switch tasks slower and have poorer working memory.
Symptoms of anxiety disorders correlate with tech use.
Depression not clear eg sometimes social media use ~ with reduced depression. Support network?
ASD sufferers struggle with planning flexibility and inhibition of irrelevant stimulus.
Why interrupt ourselves, especially when we know the costs and we are trying to focus? How might tech exacerbate self-interruption?
Optimal foraging. Mariginal Value Theorem. Explore vs exploit - switch resources when predicted value of new resource outweighs switching costs. Accurate predictions of foraging behaviour in several animal species.
Information foraging. MVT has been applied to browsing patterns. Perhaps can explain self-interruption too?
Resource intake curve is an internal estimate. Might be moved around by:
- Accessibility. Constant availability of media could lower the estimated switching/transit cost. Notifications advertise new resources.
- Boredom. Hard to define but… Possible that new high-stimulation tech raises the baseline for boredom. Boredom / killing time is top self-reported reason for phone and social media use. Vicious cycle. GSR rises before switching tasks. Strongest response and earliest onset (~30s) when switching from work to entertaintment. How does GSR connect to boredom?
- Anxiety. 20x increase in diagnosed anxiety disorders over 30y. Diagnostic improvements? Half of 18-32yo suffer in last year. FOMO ~ higher social media use. OCF symptoms ~ nomo anxiety, task-switching, social media use.
- Poor metacognition. Poor at predicting costs of switching, duration until resuming original activity. Hypothesis - awareness and management of boredom/anxiety triggers might reduce switching.
Actual evidence for these seems a little weak, but at least it’s a concrete model.
Improving executive function
Rates strength of evidence as prescriptive, signal or plausible.
- Traditional education. Good evidence for other cognitive benefits. Tools of the Mind program is promising.
- Exposure to nature. Clear evidence of health benefits eg lower stress. Plausibly a good way to restore cognitive resources.
- Meditation. Both focused-attention and open-monitoring style show reasonable evidence, but existing studies have (unelaborated) methodological limitations.
- Brain games. Commercial efforts usually over-state the evidence. Limited transfer. Mixed results for very similar games - seems that unknown details matter.
- Video games. Both AAA and lab games (Neuroracer) show improved cognitive control. But exposure to video games ~ ADHD symptoms. Also concerns about addiction.
- Drugs. Especially modafinil. But inconsistent results and modest effect size.
- Neurofeedback. Alpha wave and midline frontal theta wave training both improve control. Mechanism is not clear - may be unforeseen drawbacks.
- Brain stimulation. Mainly tDCS, interest in tACS. Produced improvements on Neuroracer scores.
- Exercise. Fitness ~ cognitive control across wide range of studies and from interventions as short as 1d. Bigger signal from cognitively engaging exercises.
Interest in cross-modal ‘Body-Brain trainer’.
No guarantee that any of these improvements will transfer to real-world behaviour.
Pretty typical recommendations eg leave phone in a drawer while working.
Improve meta-cognition by exposure to multitasking tests and the data in this book. Also by self-recording?
Enjoyable, but I wish it had been more in-depth.
- Section on executive functions was interesting at first read but on making notes I realized that I’m confused about a lot of it.
- Data on widespread multitasking and negative effects is useful.
- Neuro-divergent stuff seemed like padding.
- Foraging model is really interesting but I see very little supporting evidence.
- Review of ways to improve executive function is useful.
- Behavioural advice is just stuff everyone knows but doesn’t do, except for the ideas about improving metacognition.