Kardaras’ loose talk on addiction is what makes his argument so tough to believe. About 1.6 percent of Americans use heroin, but a quarter of them wind up addicted, according to an estimate from the National Institutes of Drug Addiction. Something like 16 percent of cocaine users become dependent on the drug within 10 years of trying it for the first time, say scientists in the journal Neuropsychopharmacology. Digital devices, as well as video games, are far more widespread than either of these drugs. If they were, in fact, comparable, we should be seeing a lot more people displaying "catatonia," with their iPads dropped dramatically beside them.
The tech-as-addiction metaphor is sloppy, though it might not be wrong. The problem is we don’t have a good handle on what qualifies as tech addiction — if it exists at all, how common it is, and what kind of environmental and physiological conditions predispose someone to it.
Although it is true that labeling something addictive is a politically and socially symbolic act, it is also true that what we experience as addictive in a given time and place is determined by cultural beliefs.
Purdue Pharma, authors of the current opiate and heroin epidemic in the U.S., created the problem by ignoring the "why" behind the numbers:
Boots on the ground was not the only stratagem employed by Purdue to increase sales for OxyContin. Long before the rise of big data, Purdue was compiling profiles of doctors and their prescribing habits into databases. These databases then organized the information based on location to indicate the spectrum of prescribing patterns in a given state or county. The idea was to pinpoint the doctors prescribing the most pain medication and target them for the company's marketing onslaught.
That the databases couldn't distinguish between doctors who were prescribing more pain meds because they were seeing more patients with chronic pain or were simply looser with their signatures didn't matter to Purdue. The Los Angeles Times reported that by 2002 Purdue Pharma had identified hundreds of doctors who were prescribing OxyContin recklessly, yet they did little about it. The same article notes that it wasn’t until June of 2013, at a drug dependency conference in San Diego, that the database was ever even discussed in public.
Combining the physician database with its expanded marketing, it would become one of Purdue's preeminent missions to make primary care doctors less judicious when it came to handing out OxyContin prescriptions. (Source)
The result? The largest drug epidemic in the history of the United States, one which has literally reversed declines in all-cause mortality in many demographics. See Opioids, Alcohol, Suicide
Part of this is a warning about the morality of Big Data. But perhaps an even larger issue is the problem of data without theory. The reasons behind these trends mattered -- were these replacing other drugs due to efficacy or due to addiction? Were the super-prescribers more enlightened as to pain management or were they running cash for scripts businesses?
Marketing, in one sense, does not require answers to these issues; you use the correlations to make sales, and the why does not matter. But ethical marketing is a different matter.