Saturday, January 27, 2018

Episode #36 - Wachter on Algorithms, Explanations, and the GDPR

s200_sandra.wachter.jpg

In this episode I talk to Sandra Wachter about the right to explanation for algorithmic decision-making under the GDPR. Sandra is a lawyer and Research Fellow in Data Ethics and Algorithms at the Oxford Internet Institute. She is also a Research Fellow at the Alan Turing Institute in London. Sandra’s research focuses on the legal and ethical implications of Big Data, AI, and robotics as well as governmental surveillance, predictive policing, and human rights online. Her current work deals with the ethical design of algorithms, including the development of standards and methods to ensure fairness, accountability, transparency, interpretability, and group privacy in complex algorithmic systems.

You can download the episode here or listen below. You can also subscribe on iTunes and Stitcher (the RSS feed is here).


Show Notes

  • 0:00 - Introduction
  • 2:05 - The rise of algorithmic/automated decision-making
  • 3:40 - Why are algorithmic decisions so opaque? Why is this such a concern?
  • 5:25 - What are the benefits of algorithmic decisions?
  • 7:43 - Why might we want a 'right to explanation' of algorithmic decisions?
  • 11:05 - Explaining specific decisions vs. explaining decision-making systems
  • 15:48 - Introducing the GDPR - What is it and why does it matter?
  • 19:29 - Is there a right to explanation embedded in Article 22 of the GDPR?
  • 23:30 - The limitations of Article 22
  • 27:40 - When do algorithmic decisions have 'significant effects'?
  • 29:30 - Is there a right to explanation in Articles 13 and 14 of the GDPR (the 'notification duties' provisions)?
  • 33:33 - Is there a right to explanation in Article 15 (the access right provision)?
  • 37:45 - Is there any hope that a right to explanation might be interpreted into the GDPR?
  • 43:04 - How could we explain algorithmic decisions? Introducing counterfactual explanations
  • 47:55 - Clarifying the concept of a counterfactual explanation
  • 51:00 - Criticisms and limitations of counterfactual explanations
 

Relevant Links


   

Thursday, January 25, 2018

Papers on the Ethics of Sex Technology (Index)


[Updated August 2019]


Much to chagrin of everyone I know, I've written quite a number of papers about the ethical, legal and social implications of sextech. Most of it focuses on the media-friendly topic of sex robots, but I've also written about virtual reality, consent apps and intimate self-tracking. Anyway, I thought it might be useful to provide links to all of this material in one post. With the exception of the first three, all are available in open access repositories. I'll update the list when full versions become available.


Books




Articles



  • Danaher, J. (2019). Building Better Sex Robots: Lessons from Feminist Pornography. In Zhou and Fischer (eds) AI Love You - Studies on Human-Robot Intimate Relationships. Dordrecht: Springer. (Philpapers)


  • Danaher, Nyholm and Earp (2018) 'The Benefits and Risks of Quantified Relationships', American Journal of Bioethics 18(2): W3-W6 (Academia; ResearchGate)

  • Danaher, J. (2018). The Law and Ethics of Virtual Sexual Assault. In Barfield and Blitz (eds) The Law of Virtual and Augmented Reality (Academia; Philpapers)








Monday, January 15, 2018

Episode #35 - Brundage on the Case for Conditional Optimism about AI


IzBNIN-z.jpg

In this episode I talk to Miles Brundage. Miles is a Research Fellow at the University of Oxford's Future of Humanity Institute and a PhD candidate in Human and Social Dimensions of Science and Technology at Arizona State University. He is also affiliated with the Consortium for Science, Policy, and Outcomes (CSPO), the Virtual Institute of Responsible Innovation (VIRI), and the Journal of Responsible Innovation (JRI). His research focuses on the societal implications of artificial intelligence. We discuss the case for conditional optimism about AI.

You can download the episode here or listen below. You can also subscribe on iTunes or Stitcher (the RSS feed is here).


Show Notes

  • 0:00 - Introduction
  • 1:00 - Why did Miles write the conditional case for AI optimism?
  • 5:07 - What is AI anyway?
  • 8:26 - The difference between broad and narrow forms of AI
  • 12:00 - Is the current excitement around AI hype or reality?
  • 16:13 - What is the conditional case for AI conditional upon?
  • 22:00 - The First Argument: The Value of Task Expedition
  • 29:30 - The downsides of task expedition and the problem of speed mismatches
  • 33:28 - How AI changes our cognitive ecology
  • 36:00 - The Second Argument: The Value of Improved Coordination
  • 40:50 - Wouldn't AI be used for malicious purposes too?
  • 45:00 - Can we create safe AI in the absence of global coordination?
  • 48:03 - The Third Argument: The Value of a Leisure Society
  • 52:30 - Would a leisure society really be utopian?
  • 56:24 - How were Miles's arguments received when presented at the EU parliament?
 

Relevant Links


 

Sunday, January 14, 2018

3 Podcasts about Bitcoin and the Blockchain




I have done several episodes of the Algocracy and Transhumanism Podcast on the philosophical and social implications of bitcoin and blockchain technologies. I thought it might be worth collating them together in this post for people who are interested in the topic.

Remember you can subscribe to the podcast on iTunes and Stitcher. The RSS feed is available here.


Episode #9 - Rachel O'Dwyer on Bitcoin, Blockchain and the Digital Commons:  
In this episode I talk to Rachel O’Dwyer about the digital commons, money, bitcoin and blockchain governance. We look at the historical origins of the commons, the role of money in human society, the problems with bitcoin and the creation of blockchain governance systems. You can download the episode here or listen below. 

Episode #14 - Aaron Wright on Blockchain Technology and the Law 
In this episode I speak to programmer and lawyer Aaron Wright. Aaron is an expert in corporate and intellectual property law, with extensive experience in Internet and new technology issues. He is a professor at Cardozo Law School and the Director of the School’s Tech Startup Clinic. I speak to Aaron about the issues arising from his forthcoming book about blockchain technology and the law (co-authored with Primavera De Filippi) that will be published by Harvard University Press. You can download here or listen below 

Episode #28 - Angela Walch on the Misunderstandings of Blockchain Technology 
In this episode I am joined by Angela Walch. Angela is an Associate Professor at St. Mary’s University School of Law. Her research focuses on money and the law, blockchain technologies, governance of emerging technologies and financial stability. She is a Research Fellow of the Centre for Blockchain Technologies of University College London. Angela was nominated for “Blockchain Person of the Year” for 2016 by Crypto Coins News for her work on the governance of blockchain technologies. She joins me for a conversation about the misleading terms used to describe blockchain technologies. You can download here on listen below. 







Thursday, January 11, 2018

The Utopia of Games: Intelligible or Unintelligible?




According to Bernard Suits’s classic philosophical dialogue The Grasshopper playing games represents the highest ideal of human existence. The Grasshopper is set in a hypothetical post-instrumentalist future. That is: a future of technological abundance in which humans can have their every wish granted to them by smart machines. Need a house? In the world of the Grasshopper, you just have think about it and the mind-reading machines in the environment will build it for you. In this world, instrumental activity has been effectively limited. You don’t have to do anything in order to get something else. Activities are to be pursued strictly for their intrinsic merits.

This is, consequently, a world in which humans have nothing left to do but play games. One of the main functions of Suits’s dialogue is to provide a philosophically tractable definition of a ‘game’. These are defined as activities that humans voluntarily choose to perform and that interpose artificial obstacles between them and some arbitrarily chosen goal. Since all obstacles are artificial in a post-instrumentalist society, all activities are games. They can be fun to play, but they are ultimately inconsequential: nothing of value hinges on success in the game. You don’t earn money or gain access to other opportunities. You are playing the game purely for the hell of it.

The desirability of Suits’s game-playing utopia has been criticised over the years. I have written about these criticisms before. In this post I want to look at another criticism of the idea. This one comes from a paper entitled ‘Endless Summer: What kinds of games will Suits’ Utopians Play?’ by Christopher Yorke. It claims that Suits’s utopia runs into problems because it is unintelligible. I’m going to offer a detailed reconstruction and analysis of this argument from unintelligibility over the remainder of this post.


1. The Argument from Unintelligibility
Yorke’s paper doesn’t set out the objection to Suits’s utopia with the clarity that I would like. If you read it, you’ll see that he presents a series of concerns, starting with one about the impossibility of designing games that would be sufficient to stave off boredom in the post-instrumentalist world, and building into a general objection to the very idea of a post-instrumentalist world. On my reading, these objections can be united under a common argumentative framework - the argument from unintelligibility. I think this captures the core of what Yorke has to say, but in presenting it in this way I am going to be deviating from the structure of Yorke’s article. So, be warned that from here on out you are getting my take on Yorke’s argument, rather than a pure summary, in what follows.

The argument from unintelligibility works like this:


  • (1) In order to be rationally/motivationally compelling, a utopian project/vision must be intelligible (i.e. give some clear sense of how we get from here to there).

  • (2) Suits’s utopia of games is not intelligible.

  • (3) Therefore, Suits’s utopia is not rationally/motivationally compelling.



The surface structure of the argument is unfussy, but there is complexity buried beneath. Premise (1) is not something that is explicitly defended in Yorke’s article but I think it is implied and does some of the argumentative work. As it stands, I think that premise (1) is plausible. Classically, utopias are conceived as ‘blueprints’ for the ideal society. The most famous utopian work in the Western canon — Thomas More’s 1516 book Utopia — is a detailed description of the organisation and functioning of a utopian society. The blueprint is supposed to guide us, to help us advocate for and realise political and social change. Of course, the blueprint model is not favoured by everyone. Some people prefer to conceive of utopianism as an ongoing process rather than a project that arrives at a stable end-state. But even if you prefer this procedural or horizonal approach to utopianism, you still need some clear sense of what the procedure is supposed to require. It’s not going to be compelling to ask people to sign up to something that is completely vague, unspecified or incomprehensible. So intelligibility looks like a must.

Suits’s utopia seems to fit, broadly, within the ‘blueprint’ class of theories. His book provides a sketch of a post-instrumentalist society and claims that it would represent the ideal of existence. Thus, it looks like he is following the classic utopian playbook and so should be held to the same standards: he should provide a vision that is rationally/motivationally compelling.

There is, however, an alternative interpretation of Suits’s book, one that is supported by Suits himself. Perhaps he is not trying to say that post-instrumentalism is desirable but, rather, that it is inevitable? If we look at general progress in technology to date, we see that it has primarily been about getting machines to do things on our behalf. If we perfect this technological progress, we will inevitably end up in a world in which machines can do everything we desire without requiring any effort on our part. We will then arrive at Suits’s hypothetical society. We will then be lucky that this society happens to coincide with what Suits calls the ‘ideal of existence’, but his goal is not to tell us how we are to get from here to there — that’s going to happen anyway. According to this interpretation, premise (1) does not apply to Suits’s argument.

The problem with this interpretation is that Suits is then trying to walk a very fine line. He is saying that his game-playing utopia is inevitable but that it will also be great. The latter makes it seem like something we should be getting behind: something that we should find rationally compelling. Furthermore, Suits suggests that there are certain things we should be doing to prepare for the post-instrumentalist reality, in particular we should start designing the games we will play to stave off boredom. This looks like a utopian project, one that needs some intelligibility if it is to be compelling.

If this is right, premise (1) does apply to Suits’s argument and we can move on to premise (2). This is the real nub of Yorke’s critique and it is defended in two different ways. We’ll look at both in what follows.


2. The Unintelligibility of Utopian Games
The first defence of premise (2) focuses on the nature of the games we will be playing in the post-instrumentalist society. Suits is clear that games will provide the difficulty and challenge that is needed to stave off boredom in this world, and he is clear that we need to start “the immense work of devising these wonderful games now”, but he is not clear on what these games will look like. This suggests to Yorke that the utopian games are epistemically inaccessible and hence unintelligible.


  • (4) In order for Suits’s utopia of games to be intelligible we need to have a clear sense of what utopian games will involve (particularly since we have to start designing them now).

  • (5) We have no clear sense of what utopian games will involve.

  • (2) Therefore, Suits’s utopia of games is not intelligible.



Before we get into the main part of this argument we have to clear up a potential source of confusion. The concept of a utopian game is somewhat ambiguous. Given how he defines games (as the achievement of arbitrary goals through the voluntary triumph over artificial obstacles) and how he characterises the post-instrumentalist society, there is a sense in which any activity undertaken in that society counts as a game. In other words, all activities in the post-instrumentalist society are ‘games by default’. This includes things like building your own house, or hunting for your own food, or doing scientific research: since machines can do these things for you any attempt to do them for yourself would involve the artificial imposition of obstacles between you and a goal. These activities would not be games in any other society. These ‘games by default’ should be contrasted with ‘games by design’, which are deliberately designed to function as games, and are always and everywhere game-like.

You might think that one way for Suits to overcome the unintelligibility problem is to fall back on ‘games by default’; to say that we do have some clear sense of what utopian games will look like: they will look like the instrumental activities that we perform today. For what it is worth, this is how I always interpreted the utopia of games. I imagined that it would consist of many of the activities that motivate us today, performed for fun (for their intrinsic merits) rather than necessity. But Yorke resists this interpretation of the argument. He thinks it doesn’t account for what Suits actually says. In particular, he thinks that if you focus on games by default you can’t make sense of Suits’s worries about boredom in a post-instrumentalist society, nor his injunction that we ought to start the work of designing the games that the utopians will play. This suggests that games by design are necessary too.

The problem then is that Suits provides no guidance as to what these games will look like. You can read between the lines of his book, however, and find some guidance as to what they will not look like. For example, Suits stipulates that in Utopia no one can be wronged or harmed. Yorke argues that this rules out immoral games:

So games that necessarily, in the course of playing them, incur irreparable physical harm upon their participants — such as Russian Roulette and Boxing — will be excluded from the utopian set. Similarly, games that require players to engage in immoral behaviors like bald-faced lying in order to play effectively — Werewolf and Diplomacy come to mind — will be off the menu. 
(Yorke, 8)

Similarly, Suits argues that all interpersonal problems will be eliminated in his utopia, which suggests that all cheating or unsportsmanlike behaviour will be impermissible:

…in Suits’s utopia, his utopian games will not be ruined by the bad behavior of their players. Cheating, trifling, spoilsporting, bullying, and grousing behaviors will all be preemptively (if somewhat rather mysteriously) eliminated.  
(Yorke, 8)

Finally, Suits says that the games cannot be boring. This, according to Yorke, rules out quite a large number of games:

Boring games would include soluble games (those with a dominant strategy that, once known, leave no opportunity for the exercise of meaningful player agency, like Tic Tac Toe), games of pure chance (those that leave no opportunity for the exercise of meaningful player agency because strategy and tactics cannot have impact on their outcome, such as Snakes and Ladders), and games of pure skill (like Chess — since utopians will have endless hours to devote to its study, and potentially live in a time ‘when everything knowable was in fact known’). 
(Yorke, 9)

The problem is that once you rule out all of these kinds of game — immoral ones, ones that involve unsportsmanlike behavior and ones that are boring — you end up with a space of possible utopian games that is epistemically inaccessible to creatures like us. The utopian games are clearly going to be very different from the kinds of games we currently play, and we have no concrete idea of what they will actually involve. This leads to unintelligibility. That’s the essence of Yorke’s first argument
I have some problems with this argument. I think Yorke sets too high a standard on what counts as immoral and boring.

For me, the immorality of conduct depends largely on the context and the consequences. Boxing might be immoral if it entails, as Yorke insists, irreparable harm (though you might dispute this on the grounds that people can consent to such harm). But why assume that harm would be irreparable in Suits’s utopia? If we have reached a state of technological perfection would this not also involve medical perfection? Think about the world of Star Trek. I remember several episodes where crew members were injured playing games on the holodeck, only to have their injuries quickly healed by the advanced medical technology. If that becomes possible for boxing-related injuries, then I don’t see how boxing retains its immorality. Likewise, I think that within the context of game, behavior that would ordinarily count as lying or deception, can lose any hint of immorality, particularly if it is required for success in the game. Indeed, I would go further and argue that it is not even possible to ‘lie’ when playing a game like Diplomacy. The standard philosophical definition of lying is to utter deliberate falsehoods in a context that ordinarily requires truth-telling. The context of the game Diplomacy is not one that ordinarily requires truth-telling. Deception is built into the fabric of the game: it’s part of what makes it fun. And since there are no serious long-term consequences of this deception, it’s once again hard to see where the immorality is. So, for me, the kinds of games that Yorke is quick to rule out would not be ruled out in Suits’s utopia.

I have similar issues with Yorke’s characterisation of boring games. I might concede that soluble games and games of pure chance and pure skill would eventually become boring, but that doesn’t mean that they couldn’t be entertaining for a period of time (e.g. while people are figuring out the solution or developing the skill) and it doesn’t mean that we couldn’t create a neverending series of such games to stave off the boredom. In other words, Suits’s injunction against boredom need not be interpreted as an injunction against individual games that might eventually be boring, but as an injunction against the total set of games that utopians might play. Furthermore, just because a game might be soluble does not mean that humans will be able to solve it using their cognitive capacities. Machines might be able to solve many games that humans currently cannot, but that does not mean that the machine-solutions will be epistemically accessible to humans. A world of post-instrumental technological abundance is not necessarily one in which humans acquire superior cognitive abilities…or is it? That’s where we come to Yorke’s second argument.


3. The Unintelligibility of a Post-Instrumentalist Society
Yorke has another strategy for defending premise (2). This one takes aim at the very idea of a post-instrumentalist society. He argues that there is a significant ‘cultural gap’ our society and the one imagined by Suits. This post-instrumentalist society is consequently unintelligible to us from our current standpoint. One way in which you can test this is to look to Suits’s descriptions of that society. Yorke argues that they are not truly post-instrumentalist.

Recall, the basic definition of a post-instrumentalist society: it is one in which the need for all instrumental activities has been eliminated. But when you read what Suits has written, it’s clear that not all instrumental activities have been eliminated. He still talks about the need to eat and to stave off boredom. These are instrumental goals that need to be addressed by different activities. Furthermore, since human beings are biological creatures they will always have such instrumental needs. In fact, human beings are, in many ways, the quintessential instrumentalist creatures: we are distinguished from all other species by the flexibility of our means-end reasoning:

Scott Kretchmar argues from an anthropological standpoint that human beings have always been a need-driven species, and that this helps to explain how we have become the problem-solving and problem-seeking animals we are today [reference omitted]. We needn’t completely buy into a Darwinist line to appreciate Kretchmar’s perfectly reasonable observations about our species’ unique relation to obstacles, which stands as one of our defining characteristics. 
(Yorke, 16)

The point here is that a truly post-instrumentalist society would be one in which humans are radically different from what they currently are, i.e. one in which we have achieved some ‘posthuman’ state of existence. But since we have not currently achieved this state of existence, the idea is clearly be unintelligible to creatures like us. To state this in argumentative form:


  • (6) Suits’s utopia of games rests on the idea of a truly post-instrumentalist society.

  • (7) A truly post-instrumentalist society would only be intelligible to posthumans.

  • (8) We are not posthumans.

  • (2) Therefore, Suits’s utopia of games is not intelligible.


You can resist this argument. Premise (6) is particularly vulnerable to critique. As Yorke himself points out, since the idea of a truly post-instrumentalist society is so difficult to imagine, it’s probably not what Suits’s was trying to describe or defend. We can think of the instrumentalist nature of a society as something that exists in degrees. We currently live in what we might call a ‘hyper-instrumentalist’ society — i.e. one in which instrumentalist activity is rampant and central to our existence. This represents one extreme end of the spectrum of possible societies. At the other end is the completely ‘post-instrumentalist’ society — i.e. the one in which the need for instrumentalist activity has been completely eliminated. Between these two extremes, possibly closer to the post-instrumentalist end, lies the ‘hypo-instrumentalist’ society — i.e. the one in which many, but not all, instrumentalist activities have been eliminated.



It’s likely that Suits’s utopia of games is intended to exist in a hypo-instrumentalist society, not a truly post-instrumentalist society. In particular, it is supposed to exist in a society in which any need for physical labour in order to secure the basic goods of life — food, shelter, material consumables etc. — has been eliminated by technology. Other instrumental needs — e.g. the need for status, friendship, entertainment etc — may well remain.

Moving Suits’s utopia away from the extreme of post-instrumentalism could make his world more relatable to our own, but it all depends on how far away it is. Yorke seems to think that Suits’s imagined world of plenitude is so ‘irreducibly foreign’ from our world of scarcity that even the hypo-instrumentalist society is unintelligible to us:

I post that utopian hypo-instrumentality, and the alien values and culture which must of necessity accompany such a state of being, constitute a fatal obstacle to the project of utopian relatability. 
(Yorke, 18)

I’m not sure that I agree with this. I don’t think the idea of a post-scarcity society is as unintelligible as Yorke claims. Although there is undoubtedly much poverty and immiseration in a world today, there are some pockets of post-scarcity. In developed economies there is a seeming abundance of material goods, food and energy. It may well be an illusion — and one that is revealed as such in emergency situations — but for all practical, day-to-day purposes it gets pretty close to post-scarcity. Furthermore, there are some groups of people — super-wealthy elites — who probably get pretty close to living a post-scarcity life when it comes to these material goods and basic services. We could look to them for guidance on what a post-scarcity society would look like.

There are however deeper philosophical problems with hypo-instrumentalism and post-scarcity. The main problem is that the human capacity for instrumentalism, and for finding new forms of scarcity, may be infinite. What were once ‘wants’ can quickly turn into ‘needs’ and can be pursued with all the vigour and urgency that entails. So while we may eliminate certain instrumental needs from human life (e.g. the need for food or shelter), I think humans would quickly reallocate their mental resources to other perceived needs (e.g. the need for social status). This new need could then be pursued in a hyper-instrumentalist way, i.e. become the major focal point in society. I could easily imagine this happening in Suits’s utopia of games. So I’m not sure that we can ever achieve hypo-instrumentalism.

That said, I don’t think this makes Suits’s claim that games represent an ideal of existence unintelligible. It might make particular reading or interpretation of that ideal unintelligible, but the basic idea that we would be better off playing games than dedicating our time to securing our basic material well-being, could still be salvageable.





Sunday, January 7, 2018

The Moral Duty to Explore Space (1): The Positive Case



Space exploration has long been a source of fascination and wonderment to me. I grew up reading the classic science fiction novels of the mid-20th century, many of which idealised the notion of space travel. I also grew up watching the various Star Trek TV series of the 1990s, which seemed to embody a positive and progressive vision of humanity’s future in space. Our best hopes for the future, it seemed to me at the time, lay in exploring space. Space would provide a near-infinite arena in which to discover new truths and expand the possibilities of being.

And yet, when I turned to academic life, I found very few philosophers willing to seriously debate the merits of space exploration. This struck me as odd given how prominent the idea was in popular culture. If philosophers engaged with technology at all, it tended to be in the fields of biology, warfare and the internet. All worthy inquiries, to be sure, but why the lack of interest in space? I had to turn to scientists and science fiction writers for that, but they would often ignore or develop superficial analyses of the philosophical questions — the questions about meaning and value — that interested me.

Fortunately, this great silence is coming to an end. There has been a noticeable uptick in writings about ‘space ethics’ in the past few years. Philosophers like James Schwartz and Tony Milligan are doing some stellar work (sorry!) on this topic. Schwartz, for example, has written an interesting moral defence of space exploration. Indeed, he has gone even further and argued that not only is it morally permissible for us to explore space, it may actually be morally obligatory for us to do so.

I want to analyse Schwartz’s argument over the next two posts. I start, by looking at the positive case: the reasons for thinking that we have an obligation to explore space. I follow this up in the next post by looking at the objections to this positive case.


1. The Basic Structure of the Positive Case
Before we get too embroiled in the details of Schwartz’s argument, it will be helpful if we can clarify its intention. Schwartz is trying to present a moral case for space exploration. He doesn’t offer a clear definition of what is meant by ‘space exploration’ but, reading between the lines, it seems to mean something like this:

Space Exploration: The use of manned or unmanned spacecraft (and other technologies) to explore our solar system (including its planets and asteroids) and beyond. ‘Exploration’, here, is not limited to simply observing and recording details about the outer space environment, but also includes using resources from that environment, intervening in the natural environment on other planets and space bodies, and potentially settling (or ‘colonising’) those environments.

This might seem like a long-winded definition, but I think it is important to be clear on each of its elements. In particular, it is worth emphasising the fact it covers both manned and unmanned space exploration, and includes exploiting and otherwise manipulating the outer space environment.

How can you make a moral argument for such space exploration? You start with a simple motivating moral principle:

Motivating moral principle: We have an ethical duty to ensure the survival of the human species, and to protect the environment that we inhabit here on Earth.

For many people, this principle will be uncontroversial. I suspect that if you asked most people whether we have a duty to ensure our own survival and the survival of the ecosystems we inhabit, they would probably agree that we do (at the very least they would agree to the former if not the latter). Philosophers, however, may be inclined to disagree, arguing that we don’t necessarily owe a duty to an abstract class like ‘the human species’ and others arguing that there is nothing intrinsically valuable about the non-human environment. There are significant debates here about how we conceive and understand value and duty. We’ll have to set most of those to the side for now and work with the moral principle as set out above. Nevertheless, rest assured that Schwartz does entertain some criticisms of the motivating moral principle. We’ll get to some in due course.

So what, then, is the positive case for space exploration? There are three prongs to Schwartz’s argument, each one set out as a distinct sub-argument:

The Argument from Resources: We have an ethical duty to explore space in order to gain access to valuable resources.

The Argument from Asteroids: We have an ethical duty to explore space in order to protect the Earth from existentially threatening asteroid impacts.

The Argument from Solar Burnout: We have an ethical duty to explore space in order to ensure our survival beyond solar burnout.


Let’s look at each of these arguments in more detail.




2. The Argument from Resources
Schwartz sets out the argument from resources in the following manner (I have shortened this slightly):


  • (1) There are plentiful resources in space (on asteroids, comets, planets etc.).

  • (2) We need basic resources to survive and we will eventually exhaust the resources available to us here on Earth.

  • (3) Therefore, we will need to acquire resources from space in order to ensure our long-term survival.

  • (4) We are obligated to do what is necessary to ensure our long-term survival and to protect our environment.

  • (5) Therefore, we are obligated to explore outer space in order to acquire necessary resources from it.


There are two potentially controversial premises in this argument. Premise (1) is potentially controversial because even if it is true that there are plentiful resources in space, they may not be the kinds of resources we need to survive. What do we know about the resource makeup of space that would justify this view? Schwartz notes that breathable air and drinkable water are probably the two most important resources for human survival. We know that water exists elsewhere in our solar system (and on asteroids and comets) and can be extracted from these bodies with more or less difficulty. Breathable air is, however, in short supply, though it may be abundant beyond our solar system. There are, nevertheless, other resources, particularly metals, available on asteroids and other planets. These are essential for various industrial processes and exist in such abundance that they could satisfy human needs in this regard for a long time. So while we may not find every resource we need in space, we can certainly find resources that are valuable.

Premise (2) is probably the more controversial premise. It seems to suggest that we will, inevitably, exhaust the resources available here on Earth. This is something that environmentalists and conservationists could reject. They might agree that we will exhaust the resources if we continue as we are currently doing, but if we could reduce the human population, and switch to more sustainable and environmentally renewable modes of production, we could preserve the resource base indefinitely. Schwartz concedes this possibility but offers three responses to it. First, he thinks that certain strategies for conservation could prove depressing and less than ideal. For instance, depopulating the earth by (voluntarily) reducing fertility could be depressing since procreation is such a deeply embedded need for so many people. Second, he thinks that even if we could conserve the Earth’s resources, this is a risky process and we should probably have a back-up plan in case it fails. Acquiring resources from space would be an obvious back-up plan. Third, he says that even if you ultimately accept the conservationist view, you don’t defeat the case for space exploration because there are other arguments to contend with.


3. The Argument from Asteroids
The second argument works like this (again, this is slightly modified from Schwartz’s presentation):


  • (6) It is very likely that at sometime in the future a catastrophic event such as an asteroid (meteorite) impact will render Earth’s biosphere uninhabitable.

  • (4) We are obligated to do what is necessary to ensure our long-term survival and to protect our environment.

  • (7) Therefore, we are obligated to prevent catastrophic events such as meteorite impacts.

  • (8) Developing planetary defence technologies is the only way to meet this obligation.

  • (9) Therefore, we are obligated to develop planetary defence technologies.


This argument is more philosophically interesting than the previous one. We can take it for granted that premise (6) is true. We have ample evidence from the past to suggest that such catastrophic impacts can occur and that these have radically altered the Earth’s biosphere. While they may not completely wipe out all forms of life, they have certainly been destructive and if we want to ensure our survival we would be best advised to avoid them.

But here is where we run into a philosophically interesting debate, one that has to with the exact nature of our obligation to protect ourselves and the environment (i.e. premise (4)). Schwartz’s notes that there is a range of preservationist views that dictate different approaches to this obligation. The two extreme ends of the range can be characterised in the following manner:

Weak Preservationism: We have an obligation to (roughly) maintain current levels of bio-diversity, but not all existing forms of life nor all existing habitats.

Strong Preservationism: We have an obligation to protect all (or most) currently existing species in their current environments.

Weak preservationism does not generate an obligation to create planetary defence technologies. If we have no specific obligation to protect the Earth’s species and habitats as currently constituted, we could discharge our preservationist obligation by creating ‘space-arks’ or off-world colonies/settlements. This still results in a obligation to explore outer space and so would still support Schwartz’s overall case. Nevertheless, it’s only if we embrace strong preservationism that we are obliged to create planetary defence technologies.

This however raises its own questions. Why embrace the strong preservationist view? Is there something intrinsically valuable about the current set of species and habitats that warrants their preservation? Schwartz looks at an argument from Holmes Rolston III that call this into doubt. Rolston observes, reasonably enough, that there have been many extinction events in the history of the Earth. The last great extinction occurred at the end of the Cretaceous period and resulted in the end of the dinosaurs. This extinction is thought to have been precipitated by a meteorite impact. But there is a paradox here. In a very real sense, we owe our existence to that impact. If it did not happen, dinosaurs could still be the dominant species. It was only through their extinction that our small rodent-like ancestors gained the ascendancy, and evolved into the many diverse forms we see on Earth today. There is, consequently, a ‘creative potential’ embedded in destructive events like meteorite impacts. This creative potential could be viewed as, on balance, a good thing: it’s good to have some evolutionary change and dynamism rather than stasis and preservation. So maybe we shouldn’t be looking to protect ourselves from catastrophic events like meteorite impacts?

Schwartz again offers three responses. First, he argues that we could accept the value of ‘creative potential’ in nature without accepting the need to permit catastrophic events. There may be other ways to preserve that creative potential. Indeed, human beings could, through genetic engineering, be a major new source of creative potential (Schwartz doesn’t say this but it occurred to me as I read his argument). Second, there is a danger in Rolston’s strategy since there is no guarantee that catastrophic events like meteor impacts always lead to future creativity. If the impact was of sufficient magnitude it might wipe out all creative potential on Earth. Prudence would, again, seem to militate against this approach. Finally, a basic concern for current sentient life, and an awareness of the danger of instrumentalising that life (i.e. treating it as a means to some other potential, but not fully specified, good) would tip the balance in favour of planetary defence.


3. The Argument from Solar Burnout
The last of Schwartz’s three arguments works like this:


  • (10) Living organisms from Earth require habitable environments in order to survive.

  • (4) We are obligated to do what is necessary to ensure our long-term survival and to protect our environment.

  • (11) Therefore we have an obligation to ensure that Earthlings have access to permanent habitable environments.

  • (12) One day, the sun will burn out and render the Earth uninhabitable; pursuing interstellar colonisation is the only way to secure permanent access to habitable environments.

  • (13) Therefore, we are obligated to pursue interstellar colonisation.


This is obviously a very long-term argument. The sun is not going to burn out anytime soon. But if it makes sense to think about our obligation to survive on billion-year timescales, the significance of solar burnout cannot be ignored. We will have to confront it sooner or later.

There are, of course, some objections to that argument. An obvious one is that all the hope underlying it — that we will be able to access permanent habitable environments — is forlorn. Avoiding solar burnout is just a stop-gap measure. According to our best available theories, the universe itself will eventually burnout (experience a heat death) and become uninhabitable. We cannot outrun our destiny. Schwartz concedes this point, but argues that even if the universe is going to undergo a heat death it is not going to undergo one for a very long time. The sun will burnout in about 5 billion years, but this only represents 5 trillionths of the currently estimated habitable lifespan of the universe. So we can definitely hope for a much longer survival through interstellar colonisation than we can by hanging around on Earth.

This is probably true, but I’m not sure how much moral (or persuasive) weight it should carry. I don’t think humans are well-equipped for moral reasoning on billion or trillion year timescales. Although the difference is huge mathematically, I suspect it all blurs into one when we start to think about it. At least, it does for me. Consequently, I’m not sure that the survival advantage highlighted by Schwartz is intelligible at the moment.

Setting that aside, there is still another problem: that interstellar travel is (currently) technically impossible. This impossibility could be thought to undermine any moral obligation to pursue it since, if we follow Kant, ‘ought implies can’. In other words, if we can’t do it, we don’t have any obligation to do it. Again, Schwartz concedes the current technical impossibility of interstellar travel, but argues that it is not logically impossible, and so we may have a lesser obligation to at least pursue its technical possibility. So while we may not currently have an obligation to travel between the stars, we may have an obligation to try to make it possible in the future.

Okay, that’s the end of Schwartz’s positive case for space exploration. I’ll look at the criticisms and replies the next day.




Thursday, January 4, 2018

Episode #34 - Lin on the Rise of Cyborg Finance

Tom-Lin-Photo-300x400.jpg

 In this episode I talk to Tom Lin. Tom is a Professor of Law at Temple University’s Beasley School of Law. His research and teaching expertise are in the areas of corporations, securities regulation, financial technology, financial regulation, and compliance. Professor Lin and his research has been published and cited by numerous leading law journals, and featured in The Wall Street JournalBloomberg News, and The Financial Times, among other media outlets. We talk about the rise of 'cyborg finance' (Cy-Fi) and the regulatory challenges it poses.

 You can download the episode here, or listen below. You can also subscribe on Apple Podcasts or Stitcher (the RSS feed is here).



Show Notes

  • 0:00 - Introduction
  • 1:30 - What is cyborg finance?
  • 5:57 - What explains the rise of cyborg finance? Innovation, Regulation and Competition
  • 9:00 - The problem of systemic risk in the financial system
  • 15:05 - "Too Linked to Fail" - The first systemic risk of cyborg finance
  • 19:30 - "Too fast to save" - the second systemic risk of cyborg finance
  • 23:00 - The problem of short-term thinking in the financial system
  • 27:15 - Does cyborg finance undermine the idea of the 'reasonable investor'?
  • 34:57 - The problem of cybernetic market manipulation
  • 37:44 - Are these genuinely novel threats or old threats in a new guise?
  • 41:11 - Regulatory principles and values for the age of cyborg finance
 

Relevant links