Floridi’s Fourth Revolution and the Demise of Ethics and the Demise of Ethics | Michael Byron in Knowledge, Technology, Politics (2010), issue 23: 135-147
First three revolutions implied reconceptualization of human nature. But there is a discontinuity with the fourth revolution:
The fourth revolution, in contrast, concerns not a finding but a making. What’s novel here is the prospect of engineering human beings: will we be immortal? artificial? digital rather than analog? (137)
Section 2: Teleology and Anthropology
Skipped
Section 3: Ethics and Autonomy
Kant's autonomy:
Autonomy in the Kantian sense is at bottom the capacity of the will to choose without causal influence from outside itself. (139)
Autonomy as basis for the categorical imperative:
Our Autonomy is what sets us morally apart from such animals (and I’ll bracket the issue of whether and to what extent non-human animals might have this capacity). Autonomy is the explicit basis of what Kant calls our dignity, which in turn is the ground of the respect due to each rational being under the Categorical Imperative. (139)
Idea still fundamental in our modern lives:
And despite a host of objections to that idea, the concept of a fundamental human dignity has been a cornerstone of modern political and ethical life, from the recognition of universal human rights to the practices of informed consent in medicine. (139)
The challenge by the fourth revolution:
If we are to follow Kant in identifying the “self” in “self-determination” with the will, then the idea of self-determination necessarily points to a causal history of action that begins with the will. And this point becomes a concern if Floridi is right about the fourth revolution, since it is fairly clear that ACs, as artifacts, will never possess a free will in the required sense. (140)
The Kantian argument againt artificial cognition:
In short, Kantians have an argument for thinking that human beings have Autonomy, namely Kant’s transcendental argument for the possibility of morality; they have no such argument for thinking that ACs can have Autonomy. (140)
But this makes this type of ethics completely sterile for thinking artificial cognition that nevertheless, is predicted to have a huge impact on our live:
And that result is a quandary for Kantian ethics: how do we deal morally with millions of pseudo-persons who have no moral status but seem at some level indistinguishable from human beings? Once again, engineering human nature seems to threaten the possibility of ethical theorizing. (140)
Floridi describes another kind of autonomy:
Floridi has deployed a distinct concept of autonomy (lower case “a” here) that he considers a necessary condition of moral agency for informational organisms, or inforgs (Floridi and Sanders 2004). According to this definition, “A transition system is autonomous when the system is able to change state without direct response to interaction, that is, it can perform internal transitions to change its state.” Without worrying about the other terminology here, the point is that an agent possesses autonomy in this sense when it can change states due only to internal causes. (140)
But this is not satisfying from a Kantian point of view. Floridi's "inner" is not really "inner" in the Kantian sense because it does not imply free will; it is de-moralized:
The autonomy Floridi describes is not Autonomy in the Kantian sense, and thus provides no basis for ascriptions of moral respect or agency. It might provide some basis for affording inforgs value, but this value directs us toward a consequentialist rather than deontological moral view.
Takeaway here:
- Kantian concept of autonomy is sterile for expressing our daily reality in case our environments are filled with smart technologies or forms of artificial cognition.
- Floridi introduces a mechanistic consequentialist view of autonomy
- Both concepts of autonomy are incompatible.
Floridi's Information Ethics ethics of information
@information @ontocentrism @entropy
Interesting: from biocentrism vs suffering, to ontocentrism vs. entropy:
IE emerges as the same type of theory but one that replaces biocentrism with ontocentrism. Being is more fundamental—and more fundamentally valuable—than life, according to this view; and entropy, not suffering, is the generic evil, where we should understand ‘entropy’ in an information-theoretic and not a thermodynamic sense (about which I will say more shortly).(141)
Being is more important, but moreover being is now understood in terms of information. This implies that 1) information is inherently valuable, 2) its destruction, i.e. entropy, is bad.
Finally, once we understand all beings as fundamentally constituted by information—whether physically realized or not—the “ontocentric” nature of IE and its axiology becomes transparent: “informational objects” are valuable, and their destruction (or entropy) is thus morally bad. (141)
Entropy concerns the corruption or perishing of informational objects (not information itself, right?)
Byron argues that if we take seriously that what constitutes being is information, then Floridi's ethics cares about beings-as-information:
So when Floridi asserts that IE is ontocentric, he means that it is info-centric: being as such isn’t valuable, rather beings-as-information is valuable. (142)
Reasoning:
Floridi regards beings as such as bearers of intrinsic value. In order to have intrinsic value, beings must have a nature or structure, something in virtue of which they possess value. Kant bolstered his claim that human beings have intrinsic value with his anthropology: in virtue of our Autonomy, we possess a dignity that is worthy of respect. This dignity grounds the claim of intrinsic value. What comparable feature do beings as such possess, whether or not they are physically realized? The answer is clearly information. Reflection on the fact that Floridi understands ‘beings’ to include non-physical, informational beings should lead us to the conclusion that what is ultimately valuable in his IE is not beings as such but rather beings-as-information. (142)
More concerns:
- More information is better?
- What is the concequence of saying information is inherently valuable?
- "Though Floridi denies that IE entails valuing rocks as much as human beings, he offers no calculus for comparing them. If I am confronted with a choice between destroying rocks and destroying human beings, how do I make that choice? Since IE demands minimizing entropy, the question must turn on the informational content (whatever that amounts to) of what is to be destroyed. So some finite number of rocks will equal one human life." (142)
"Instead, Floridi’s IE is based on the persistence and thriving of informational objects. The basic principles of IE are:
- entropy ought not to be caused in the infosphere (null law);
- entropy ought to be prevented in the infosphere;
- entropy ought to be removed from the infosphere;
- the flourishing of informational entities as well as of the whole infosphere ought to be promoted by preserving, cultivating, and enriching their properties (Floridi and Sanders 2004). " (144)
Interesting point, but revisit paper for details:
Using examples here is itself question-begging: only if we concede that moral good and evil can be so characterized prior to and independent of the concept of moral agency can we possibly define moral agency in terms of moral good and apart from moral responsibility. Even then, the given examples are answerable: we treat children differently at different ages, depending on their progress toward full moral responsibility and moral agency. To the extent that we engage in moral reasoning with children, it is precisely in virtue of extending them proleptic membership in the moral community in order to train them up in the practices of moral life. (145)
Main criticism of Floridi's information ethics:
- Inherits "all of the profound flaws of consequentialism" (145)
- "has a problematic axiology of informational objects" (145)
- "cannot make sense of its own conception of moral agency without an antecedently given theory of moral good and evil" (145-146)
- "which in turn hinges on a metaphysically puzzling and mysteriously teleological definition of information as something that strives to be realized." (146)
Conclusion: The Demise of Ethics
Floridi pleas for a smooth transition of old ethics into the new. Byron instead argues that this is not supported, while there are clear arguments on why the fourth revolution will disrupt our ethics.
To the contrary, I think that I have shown that the fourth revolution must be deeply disruptive of our dominant modes of ethical theorizing. By radically changing our concept of human nature, the fourth revolution will break eudaimonistic ethics. By introducing a class of beings who seem human yet are not autonomous in the relevant sense, the fourth revolution will bypass Kantian deontology. By urging us in the direction of a theory as unsatisfying as Floridi’s Information Ethics, the fourth revolution even threatens to disable consequentialism.
In the end, the question of an ethically smooth transition through the fourth revolution becomes a problem in engineering ethics. How can we regard the transition as ethically smooth without ethical constraints on the engineering of human nature? We are in the unprecedented position of remaking ourselves—or making ourselves into something else, perhaps many things. It is not possible to answer the question, “how should we be?” if this “should” has normative, ethical import. The disturbing conclusion is that, with or without ethical constraints, the revolution is coming in any case.