Printed: 2020-02-26

Institute for Ethics and Emerging Technologies

IEET Link:

A Paradox of Enhancement

Phil Torres

Ethical Technology

August 25, 2010

Is it necessary first to enhance in order to decide whether or not enhancing is a good idea?

Many transhumanists are enthusiastic about the possibilities of cognitive enhancement. Such enthusiasts might say something like: “I want to use advanced technologies - from genetic engineering and psychoactive pharmaceuticals to neural implants and even mind-uploading - to increase my intelligence, to make mesmarter, wiser, or more creative’ [PDF], to produce a ‘smarter and more virtuousperson, to mentally and emotionally augment myself.”

But, as the italics suggest, talk of enhancement presupposes some conception of the self. Specifically, it assumes that the self is capable of enduring such modifications, e.g., as a pattern, or as an immaterial soul, or whatever. The resulting enhanced being would thus still be me, it would just be a different and “better” (according to some set of criteria) version of me.

As IEET Fellow Susan Schneider puts it in an excellent 2009 article [PDF]:

Should you embark upon this journey? Here, there are deep philosophical questions that have no easy answers. For in order to understand whether you should enhance, you must first understand what you are to begin with. But what is a person? And given your conception of a person, after such radical changes, would you still be you or would you actually bear little relation to the person you were before? And if the latter situation is the case, why would embarking on the path to radical enhancement be something you value? For wouldn’t it instead be a path that leads to your own demise, leading you away from your true self, ultimately cause you to cease to exist? In order to make such a decision one needs to understand the metaphysics of personal identity - that is, one needs to answer the question: What is it in virtue of which a self or person is supposed to continue existing over time?

Now, an interesting paradox arises when one combines the above claims with a specific (and controversial) stance on what the self is. For the sake of clarity, I will begin by outlining four traditional philosophical approaches to understanding the self, each of which may exhibit some internal diversity among its advocates (i.e., there may be debate about which particular conception of the self is correct among theorists who all agree that the general approach is right).

The first attempts to provide a reductive account of the self - that is, it attempts to explicate the self (non-circularly) in terms of something else, such as a pattern (Kurzweil and Moravec), psychological continuity relations (Locke), or “the myriads of attributions and interpretations [...] that have composed the biography of the living body whose Center of Narrative Gravity it is” (Dennett).

The second posits that the self is an irreducible “brute fact” about the world. This position exhorts us to “openly confess that the concept of a person is primitive, that there are no informative identity criteria to be given for persons, that persons are what they are and are not some other thing.” On this view, the self is an extra fact about the world once all the physical facts are listed, whereas the first view sees the self as fact included in such an exhaustive list.

The third offers a supernaturalistic account of the self, one that holds that “the self must be immaterial, tenuously linked to the body, and God must have been its cause.” One finds this conception in religions that adopt something like the substance dualism of Descartes.

Finally, the fourth option is to simply eliminate the self: the self is not a pattern or diachronic relation of psychological continuity, nor is it an irreducible or “magical” entity. Rather, the self simply doesn’t exist, like heaven or hell or phlogiston.

To the above four approaches, Colin McGinn adds an intriguing fifth one: maybe the self has a purely naturalistic (or reductive) explanation, but maybe this explanation lies permanently outside the realm of human intelligibility. In McGinn’s words, maybe we humans are “cognitively closed” to the problem of the self, given the conceptual limitations inherent to our evolved mental machinery. Just as a dog could never in principle come to understand and speak a human language, even with years of training and the appropriate vocal apparatus, so too can we never in principle come to understand the correct solution to the conundrum of selves. In other words, we simply lack the right cognitive stuff for the task!

(McGinn argues that this applies to most, or possibly all, philosophical problems; indeed, he claims that philosophy is a receptacle for such permanent insolubility, and thus that there actually exist no philosophical solutions - there are only philosophical problems!)

Now, the paradox appears when one recognizes that cognitive closure is a species-relative phenomenon. Again, McGinn maintains that there are solutions to philosophical puzzles, they’re just beyond the epistemic reach of actual humans - but not beyond the reach of some possible future beings. The human can understand aspects of the cosmos that the dog cannot, and, as Bostrom illustrates with a (somewhat risible) diagram from his “Transhumanist Values,” a superintelligent posthuman could make sense of phenomena that will forever puzzle us, or that we will never even know exists.


(Incidentally, I sometimes sit in my room and observe the occasional house fly struggle to get through the window, which I often leave open an inch or two at the bottom. It’s not so much that the flies are dumb, they just lack the right cognitive apparatus to solve the problem of escaping through the crack at the bottom of the window. That is, they never evolved the ability to find the solution to such problems, problems that were completely absent from their past environments. Thus, I think of us philosophers - or physicists, or technologists, or whatever - as the flies, and the window as, say, the problem of the self.)

Thus, it follows that a species of cognitively enhanced posthumans could provide us with exactly the sort of solutions that we are looking for - for example, the solution to the problem of what constitutes the self. But the original question was whether or not we should create such posthumans through the use of cognitive enhancements in the first place, given that the process of enhancing ourselves may actually result in our own destruction.

In summary, then: Should we enhance ourselves? Well, it depends on what the self is. But, if McGinn is correct, the problem of the self falls into a category of philosophical conundra with respect to which we are cognitively closed. Thus, we cannot in principle apprehend the correct analysis of the self. But a cognitively enhanced posthuman may be able to do precisely this, given his/her/its “superior” powers of cognition. That is, the problem of the self may be “open” to such a being just as the problem of, say, speaking a natural language is open to us but not to our canine companions.

It follows from this that in order to know whether we should enhance ourselves or not, we should enhance, since our enhanced “selves” would (or at least could) be in a much better position to say than us. This is the paradox of enhancement.

A couple of clarifications are in order. First of all, the argument rests upon McGinn’s contentious thesis that the self is a permanently insoluble mystery (that is, to us but not to our posthuman progeny). The paradox would, therefore, lose much of its force if McGinn were wrong.

More importantly, though, it must be pointed out that cognitive enhancement is only one route to the destination of greater-than-human-intelligence: the other is artificial intelligence (AI). Another option would thus be to create a superintelligent AI system that could help us deliberate about whether or not we should use cognitive enhancements. This would offer a way out of the paradox, since it doesn’t involve modifying ourselves.

The trouble is, however, that AI may turn out to be more difficult than enhancing the neurobiological core of Homo sapiens, which means that the paradox would remain intact: in this case, the most feasible way to engender a new species of ultra-smart posthumans would be through human enhancement and not AI.

Finally, one could generalize the basic idea to AI as well. That is, we might pose a general moral question about whether or not it would be good to create a species of posthumans through either method of enhancement or AI. Our ability to answer this question, though, is no doubt far more limited than the ability of a superintelligent biotechnological hybrid or completely synthetic posthuman to answer it.

But, of course, the only way to get advice from such sagacious beings about whether or not to create them is to first create them. So, again, there are elements of paradox present here too. What then should we do?


Contact: Executive Director, Dr. James J. Hughes,
IEET, 35 Harbor Point Blvd, #404, Boston, MA 02125-3242 USA
phone: 860-428-1837