Wednesday, February 8, 2023

Uncoupling Cost and Benefit: How Technology Transforms Morality

How can technologies transform our moral beliefs and practices? One suggestion, made popular by a famous case study on technology and moral change, is by uncoupling certain costs and benefits, thereby altering how we perceive and prioritise values.

But how, exactly, does this happen? Can the opposite happen - can technologies 'couple' or bundle together certain costs and benefits - to the same effect? And can we use this idea of coupling-uncoupling to anticipate potential future technological moral transformations? These are the questions I want to consider in the remainder of this article.

1. Contraception and Uncoupling

The famous case study I alluded to in the introduction concerns the impact of contraception on sexual morality. I have discussed this case study in depth before and explained some of the supporting evidence. I don't wish to rehash it here. I just want to focus on what the case study tells us about the phenomenon of uncoupling.

In brief, the idea is that effective contraception uncoupled sexual intimacy/gratification from reproduction. In the past (roughly pre-1900) if you had (heterosexual)* sexual intercourse with another person, this carried a significant risk of unwanted pregnancy. This was true even if you used the forms of contraception available at that time. This meant that sexual intimacy was almost always coupled together with reproduction. It wasn't possible to pursue (heterosexual) sexual intimacy without also being forced to pursue the possibility of reproduction. This made extra-marital or premarital sex a risky endeavour, particularly for women (since they bore the main costs of reproduction), and consequently relatively few willingly engaged in it. Associated with this, there were very strong norms of sexual purity and chastity (primarily for women), and a corresponding condemnation of sexual looseness or liberty.

This changed, dramatically, when effective forms of contraception, particularly forms of contraception that women could control, became widely available. The contraceptive pill is the most famous example. These forms of contraception decoupled sexual intimacy from reproduction (to a high degree of probability and safety) and thus enabled people to access the value of sexual intimacy without necessarily being forced to pursue the value of reproduction. The social effects of this have been quite dramatic. Not only is premarital sex now normalised, but most people now ignore social or institutional norms that still enforce or favour sexual purity or chastity. A more liberal approach to sexuality has, as a result, taken root in many societies.

A few comments about this case study. First, maybe it is not correct to say that effective contraception 'uncoupled' sex from reproduction. All forms of contraception are subject to failure. The possibility of unwanted pregnancy is not completely eliminated. So perhaps it would be more correct to say that it significantly reduced the potential cost (unwanted pregnancy). Still, there is something to be said for sticking with the word 'uncoupled'. For most people, the perceived risk of unwanted pregnancy when using an effective form of contraception is so low that it doesn't feature much in their decision-making.

Second, to say that sexual intimacy has been decoupled from the risk of unwanted pregnancy is not to say that it is decoupled from all other risks. Depending on the form of contraception used, and the type of sexual practice, the risk of sexually transmitted infection could be quite high and this could, in turn, alter sexual beliefs and practices (there is an interesting story to be told about the history of HIV, both from initial panic -- perhaps 'moral' panic -- through to the invention of effective forms of treatment. I won't tell this story here though).

Third, it is of course true that, for many people, the link between sexual intimacy and reproduction is an important one, and many people want to pursue both at the same time. The critical impact of contraception, however, is that it gave people the choice of pursuing these things independently if they so wish (again, there is another interesting story to be told about the rise of assisted reproduction and fertility treatments that also serve to decouple sexual intimacy from reproduction. I won't tell that story here either).

Fourth, and finally for now, just because contraception impacted on how many people thought about the value of sexual intimacy, it does not follow that old norms of sexual purity and chastity are eliminated. They still linger. There are still double standards when it comes to social judgment of sexual liberty -- men, typically, being free from punishment and shame; women, typically, being subject to both. Nevertheless, social attitude surveys suggest that there has been a big shift in sexual mores over the course of the 20th and 21st century. That said, recently there evidence to suggest that younger generations (Gen Z etc) are less sexually liberal and less promiscuous than mid-to-late 20th century generations. There is survey evidence to support this regarding age of first sexual experience and number of sexual partners. For what it is worth, I don't think this is new trend (yet) represents a recrudescence of sexual conservatism. I suspect it may be overstated, and driven by other factors such as delayed adulthood, increased atomisation, social anxiety and the rise of a highly risk averse culture. I also suspect that some elements of this new trend represent a further compounding of the ethic of sexual liberty -- people should be free to not have sex, or to identify as asexual, if they so choose and should not be under any social pressure to have sex simply because that's what everyone else does.

These final comments are half-baked thoughts that require further research and development. They are also tangential to my main focus in this article. If we accept the contraception case study at face value, then we need to take seriously the idea that technology can transform our social moral beliefs and practices by uncoupling certain values and risks. How common is this phenomenon? Can the opposite happen? Let's consider these questions now.

2. Failed Uncoupling: The Case of Opioid Addiction

Contraception is a case study in effective uncoupling: technology promised to reduce the risk of unwanted pregnancy and it really did do so. Sometimes technology fails to deliver on its promise to uncouple. What are the potential effects of such failure? I'm sure there are many examples of this but one that springs to mind are the repeated attempts to create opioid-based painkillers that give us the benefits of pain relief without the associated costs of addiction and other related negative effects.

It has long been known that opioids are highly effective painkillers. It also long been known that they are highly addictive and that this addiction has negative impacts on individuals, their families and society at large. So, unfortunately, it seems that nature has bound both value and risk together in opioids: taking them can reduce pain and suffering (undoubtedly a good thing) but it can also create addiction, crime and other social costs.

As a result of this, there have been repeated attempts by drug companies to create safer forms of opioid that 'decouple' the positive effects from the negative ones. Heroin, for example, was created by the German company Bayer (in the late 1800s) and marketed as a safer alternative to morphine, on exactly these grounds. This is how Patrick Radden Keefe describes it in his book on the opioid epidemic in the US:

...The German company Bayer began to mass market [heroin] as a wonder drug -- a safer alternative to morphine...Bayer proceeded to sell the drug in little boxes with a lion printed on the label, and suggested that differences in the molecular structure of heroin mean that it did not possess the dangerous addictive qualities of morphine. It was an appealing proposition: throughout human history, opium's upsides and its downsides had appeared to be inextricable, like the twined strands of a double helix. But now Bayer claimed, they had been decoupled, by science, and with heroin, humans could enjoy all the therapeutic benefits of the opium poppy, with none of the drawbacks. 
(Keefe, 2021, 186)


Of course this turned out not to be true. Heroin was, in fact, much more powerful than morphine and just as addictive. The most recent episode in this saga has been particularly tragic and well-documented. In 1996, the company Purdue Pharmaceutical released a new opioid-based wonder drug onto the market: Oxycontin. Based on a preparation of oxycodone (also more powerful than morphine), Oxycontin promised to uncouple opioids from their addictive effects. How so? Oxycontin used a slow-release mechanism (a special coating on the pill) to ensure that people taking it didn't get a big 'hit' or 'high' from the drug when they initially swallowed it. Instead, the drug would be gradually released into their bloodstream over the course of 12 hours. This would, according to Purdue, minimise its potential for addiction and abuse.

Purdue Pharmaceutical aggressively marketed the drug. I won't get into all the details here -- they are well documented, for example in this article -- but suffice to say this marketing campaign was based on some suspect and misleading claims about the benefits of the drug and the low-risk of addiction. They also encouraged its use for non-acute long-term pain management as opposed to short term acute pain management. The net effects of this have problematic, to say the least. It turns out that Oxycontin did not achieve the longed-for decoupling. Its slow release formula was far from perfect and could easily be bypassed by addicts and abusers. There was, as a result, a huge increase in opioid-addiction and opioid-related deaths (approximately 500,0000 Americans have died as a result of opioid abuse since 1996), a lucrative black market trade in Oxycontin and other prescription opioids, an increase in crime and related issues in particular regions. Whether Oxycontin is solely responsible for this is, of course, unclear. Other factors played a role such as workplace related injuries, unemployment and economic decline, and the availability of other similar drugs. But Oxycontin was, undeniably, a significant part of the picture.

What were the moral effects of this failed uncoupling? This is tricky. Many philosophers and policy-makers think it is a mistake to moralise drug addiction and its consequences. We should not, in other words, view the decision to take a drug or to fall into addiction as a moral failing or moral vice. To do so, presumes freedom of choice which may be absent and may be counterproductive to the goal of reducing the harms of addiction. Furthermore, our restrictive drug laws often create additional or unnecessary moral problems associated with addiction. For instance, criminal wrongdoing associated with black market trading in prescription drugs (theft, violence, gang wars) may largely be the result of insufficiently liberal drug laws, not the drugs themselves. I am very sympathetic to this view and broadly in favour of drug decriminalisation. However, even if it is correct, the reality is that people often do moralise the choices of addicts and subject them to moral criticism. Likewise, even if addicts lack full moral autonomy, those associated with the production, prescription and trade of addictive opioids do not. Their choices and decisions can be moralised and subjected to moral criticism, if their consequences are individually and socially harmful.

So what has happened with the failed promises of Oxycontin? This is purely speculative but it seems that we have two waves of change. First, there is the euphoria associated with the promise of the drug. The pharmaceutical companies push the idea that we can now access the benefits of opioids without the risks. This leads to some initial softening of professional and social attitudes toward them. Doctors become more willing to prescribe them for a larger range of conditions, and become less worried about the potential harms of doing so. Similarly, the social taboo or shame associated with using the drug starts to dissipate.

Second, once the drug fails to live up to its promise, we get a retrenchment of attitudes. The drug companies become social pariahs and they are widely condemned for their false promises. In the case of Purdue Pharmaceutical, the family that owned the company (the Sacklers) have been 'cancelled' in many settings. They were major philanthropists and patrons of the arts. They have seen their names removed from buildings and exhibitions. People don't want to be associated with their morally tainted reputation. Prescribing the drug is no longer so liberally permitted. Doctors have to take greater responsibility for the decision to do so. Taboo and shame recrudesce around those that use the drugs. There is the sense that you lack moral rectitude or courage if you take the drugs; there is a virtue in resisting their temptations. There is a scene in the TV adaptation of Dopesick (a book about the Oxycontin epidemic) that revels in this moralisation. One of the lawyers in the case against Purdue undergoes treatment for cancer. After surgery, in intense pain, he is offered some Oxycontin. His gaze sharpens and he asks for a less effective drug instead. It's clear that the makers of the show view the decision as a courageous one. Commentators have, however, criticised it as unnecessarily shaming those that choose to take opioids. Perhaps they are right, but the point remains: the choice to take the drugs has become moralised. You exhibit moral courage and virtue if you do not. In fact, given what we now know about the negative effects of the opioid epidemic in the US, the choice to take opioids may be an even more morally contested decision than it once was.

As I say, these thoughts are speculative, but they give a sense of what might happen to social moral beliefs and practices in the aftermath of a failed uncoupling.

3. Apparently necessary coupling: the case of privacy

Technology doesn't always uncouple values from their costs. Sometimes, perhaps even oftentimes, technologies have unwelcome side effects that bundle together values and costs. Sometimes these effects don't become apparent until long after the technology has been adopted. As a result, the decision to use the technology forces us to confront a coupling together of values and costs that may not have been present before. A new value tradeoff comes into existence or becomes more acute and pressing.

There are many examples of this, but one that springs to mind is the coupling together of digital convenience and surveillance. As we all know, digital technologies have numerous conveniences. This is particularly true of software and services delivered via the internet or with the assistance of machine learning/AI. Most of us now spend inordinate amounts of time communicating via social media platforms, consuming digitally-mediated entertainment and news, and working with digitally-mediated tools. The problem is that these technologies, almost by their very nature, collect information about us. Every keystroke is recorded; every digital transaction is timestamped and placed in a memory bank. It's possible to destroy some of this information, but not all of it. Most of it is available somewhere, if you are willing and able to look for it. As a result, accessing digitally convenient services is coupled with a significant cost to individual privacy.

Early in the era of digital technologies, the scale and significance of this digital surveillance was not obvious. Many people did things online (wrote ill-judged emails; shared ill-judged photos) that they didn't expect to come back to bite them 20-30 years later. Many people participated in online forums without knowing that governments and corporations were silently collecting everything they did with a view to mining it for useful insights. Now, we all know this and the privacy-related cost of digital convenience is undeniable.

What is the moral effect of this? Tech enthusiasts and futurists have long supposed (or advocated) for the idea that 'privacy is dead' or, at least, dying. When faced with the choice between digital convenience and privacy, people seem to overwhelmingly choose convenience. As economists might put it, whatever we might say, our revealed preference is for convenience, not privacy. There is, however, a significant backlash against this idea. Privacy advocates claim that the value of privacy is even more salient and obvious in the digital era; that we are too quick to give up privacy in favour of convenience (is social media really all that great? Are AI-services all they are cracked up to be?); and that, to some extent, the choice between digital convenience and privacy is a false dilemma: we can have digital services that are less intrusive and surveillant. Significant legislative muscle has been brought to bear on this problem, particularly in the EU in the form of the GDPR. This puts more safeguards in place to prevent unwarranted collection of data and gives individuals rights over their personal data.

My own view, for what it is worth, is that these legislative efforts are valiant, and it is indeed correct that we shouldn't be so quick to give up privacy for digital convenience. Nevertheless, when it all washes out, I suspect the 'privacy is dead/dying' idea will be closer to the truth. As long as we keep using digital services (social media, internet-based communications, AI-based software) they will keep collecting and mining personal data. This means we will always forgo some of our privacy. Since we show no obvious signs of turning away from such services, it seems likely that privacy will continue to ebb away.

I could be wrong about this. There may be a major backlash and withdrawal from the use of digital technologies. There may also be some satisfactory technical solution to the privacy dilemma -- there are some already of course -- but I'm not sure these will be widely distributed or used.

4. Future Uncouplings and Couplings

Now let's turn to the future. If coupling and uncoupling can have the moral effects suggested in this article, then we might be able to use informed speculation about the future direction of technological innovation to anticipate future moral changes. Any such speculation must be taken with a heavy dose of epistemic humility, but let's try to have some fun with it.

Are there any obvious uncouplings on the technological horizon? One potential example is the uncoupling of meat consumption from factory farming and animal slaughter (for what it is worth, I spoke about this example with Jeroen Hopster in one of my podcast episodes). At present, if you wish to have a meat-based diet, you must do so in full awareness that this requires the slaughter of animals. And while some people consume meat that is relatively** humanely slaughtered, most people consume it from large factory farms with dubious animal welfare standards. In short, for most people, the benefits of meat consumption (gustatory pleasure; dietary need) come with the not insignificant cost of harm to animals, plus some other negative downstream costs (e.g. creating ripe conditions for zoonotic viral transmission). Most people seem happy to accept those costs; but some people (perhaps an increasing number) find them unbearable. [On a purely anecdotal level, I have now found that, among philosophers and ethicists at least, meat consumption is frowned upon. Few people will admit to it openly and many conferences and events operate with a vegetarian/vegan default. I fully accept that this is not representative of the general population.]

So, at present, we have the coupling of a value (meat consumption) with a cost (animal suffering). Developments in the field of artificial meat production may uncouple the value from the cost: we could produce meat without requiring significant and ongoing animal suffering. This may have two interesting moral effects. The first is that it may reduce or eliminate the moral pressure to be a vegan/vegetarian. This does not mean that veganism would be eliminated. There could be other benefits to a vegan diet that are unaffected by technological advances. The other potential effect is that it increases the moral 'taint' or 'sin' of traditional meat consumption: anyone that chooses to consume meat from a slaughtered animal, when there is another lower cost alternative, will face a larger burden of justification. This could have all sorts of interesting second and third order effects.

I won't go into other possibilities in as much detail, but here are a few other uncouplings that are worth pondering:

Virtual reality and human contact: VR may give us many of the benefits of in person contact without the associated costs. In many instances, I suspect people will prefer in person contact (at least for now) but in some cases this may not be true. For example, VR could give you much of the emotional excitement of in person contact sports(boxing; football etc) without needing to run the risk of serious injury. Some people like to run the risk of genuine physical harm -- that may be part of the pleasure for them -- but for many people these risks are not worth it and they may find the VR alternative compelling. This could lead to a moralisation of some forms of in person contact.
Automation and human risk/error: Many advances in automating technologies are sold to us on the basis that they uncouple the benefits of intelligence from the associated risks of human error. These benefits are often oversold, but in some cases they might be compelling and could affect moral beliefs and practices. For instance if (and it remains an 'if') automated driving is safer than human driving, then, as Nyholm and Smids have argued, human driving may become the morally inferior option and require greater moral justification.
Love/Sex robots and the complications of intimacy: Intimate human relationships come with all sorts of benefits: physical and emotional pleasure, mutuality, shared resources, support, comfort and so on. They also come with significant costs: betrayal, jealousy, anger, emotional and physical burdens, anxiety and so on. Some developers of love/sex robots have argued that the technology could uncouple the benefits of intimate relationships from their costs. Such claims are often overly-simplistic: some benefits of intimate relationships are inextricable from their costs. But I think Henrik Skaug Sætra is right to suggest that the technology could lead some people to prefer a different kind of intimacy ("deficient" to use his word) from traditional human intimacy. Again this could become a moralised choice.


Those are some possibilities. There are many more. I would encourage readers to consider them for themselves and to consider whether the 'uncoupling/coupling' idea is a useful way of thinking about technology and moral change.

*And, of course, non-heterosexual forms of sexual intimacy carried significant legal risks.

** There is a debate to be had about whether there can be 'humane' animal slaughter. I think there probably can be, but I won't get into that debate here.