LTAAA Symposium: Response to Surden on Artificial Agents’ Cognitive Capacities

You may also like...

6 Responses

  1. A.J. Sutter says:

    Beneath the eloquent expression, the gist of your argument seems to be simply that you have a preference for one legal fiction over another.

    @ Fiction OLD arises from “our law and moral systems often reflect[ing] assumptions about humans’ [cognitive] capacities that on closer inspection are often shrouded in obscurity”
    @ Fiction NEW is that AAAs that replicate some human cognitive abilities are legal persons.

    It doesn’t seem that you’re saying OLD should be displaced from the law; rather, your argument seems to be that a fortiori NEW should be accepted into it.

    An aim is “that the lens should be turned back on humans as well, and on [certain] uncritical assumptions we make …, which are then written into the law.” Moreover, acceptance of NEW would “do wonders to displace our sense of uniqueness, acknowledge the possibility of other modes of existence, and re-invoke the sense of wonder about the elaborate tales we tell ourselves about our intentionality, consciousness, autonomy or freedom of will.”

    A couple of points:

    1. ALL legal fictions involve writing uncritical assumptions into law.

    2. In the particular cases of OLD and NEW there’s a remarkable congruence between the uncritical assumptions being made. The argument for taking the intentional stance regarding AAAs relies on the obscurity of what’s going on inside the black box.

    3. Let’s assume for the sake of this paragraph that it’s desirable to “displace our sense of uniqueness” and to “re-invoke the sense of wonder” you describe. And in the same spirit let’s assume the somewhat shakier notion that these are suitable objectives for “the law.” Even assuming further that according legal personhood to AAAs would be sufficient to meet these goals, nothing suggests that it would be necessary. E.g., according legal personhood status to primates and cetaceans might also suffice.

    4. Yoking together legal and moral systems, legal and moral vocabulary, etc. is to mix apples and oranges in the same breath. The reasons we might or might not “disdain” some other vocabulary or a fiction in the law can be distinct from why we might make the corresponding judgment morally. E.g., judges who accept the arguments from efficiency and logical coherence your book proposes might adopt the fiction NEW mentioned above. But many persons in society at large might find such a change in law morally repellent, and seek to reverse it by political or other means.

    Is your argument that we ought to accord legal personhood to AAAs based solely on legal grounds, or are you also arguing that we should do so on moral grounds?

  2. Samir Chopra says:

    AJ: I think I can now confidently state a theorem pertaining to philosophical discourse. Let me term it the Chopra Theorem of Philosophical Argument: With probability one, if a philosophical argument carries on long enough, one of the interlocutors will suggest the other’s conclusions or premises are ‘merely preferences’. (This also needs Chopra’s Lemma: All philosophical premises are expressions of preferences). So your suggestion that my argument above is merely an expression of ‘preferences’ is not too perplexing, and is unlikely to cause me offence.

    I am however, puzzled by why you have responded the way you have to a post that doesn’t actually feature a single mention from me of ‘legal personhood’. My claims about the aims of AI and the need for a lens to be turned back on human cognitive capacities and our self-conceptions can be made independently of any arguments for legal personhood for AAs.

    But lets take the argument there anyway. And lets take it via the notion of “fictions” which seems to be doing a lot of work for you in your rejoinders to me.

    So, point by point:

    1. As I have noted before, a great deal of our so-called substantive discourse is ‘fiction’, just well-established ones. Consider the notion of a ‘person’ for instance. Despite thousands of years of philosophical blathering about this, it seems to me that a ‘person’ is a fiction, invented to reify our notions of responsibility, blame and agency. Only a society interested in blame, holding people responsible, and in describing some kind of unitary entity as a cause for an event would invent ‘agents’ and ‘persons’. So I think constantly suggesting that I’m trafficking in ‘fictions’ is to not really mount a critique at all.

    2. I’m glad you made your point 2 for it supports my position. Because just like “The argument for taking the intentional stance regarding AAAs relies on the obscurity of what’s going on inside the black box” the argument for taking us to possess magical properties like free-will and autonomy rely on similar obscurity. When AAs get to be so complex that they become authorities on what they say, we will find we have to treat them as persons. (cf. Rorty, Incorrigibility as Mark of the Mental; Dennett, The Case for Rorts)

    3. I agree again. If thinking about AA’s makes us revisit our conceptions of ourselves then our job will have been done. But I think it might not be so easy; primates and cetaceans might still leave us thinking there was something unique about the particular biology of this planet.

    4. When I say “legal and moral vocabularies” I am not yoking them together; I am merely relying on a grammatical convenience that lets me refer to them in the same sentence (but they are somehow related, otherwise, I would not spend so much time in my philosophy of law class talking about natural law theories). To mention them in the same sentence is not to imply equivalence, identity or any tighter a relationship than when you say “apples and oranges” or when you might say “moral and scientific vocabularies”. It’s merely conjunction, that’s all.)

  3. Harry Surden says:

    Thank you for your thoughtful reply Samir.

  4. A.J. Sutter says:

    Samir: Of course my rhetorical point in context wasn’t that they’re just preferences, but preferences between things that differ only slightly and/or not along a very significant dimension.

    As for your point 1, just curious: can you present a counterexample of an actual society that lacks any concept of person, while also lacking any interest in (i) blame, (ii) holding people responsible, and (iii) describing some kind of unitary entity as a cause for an event?

    As for your point 3, I’ll address the biology point under one of your later posts.

    As for your point 4, the fact that you spend so much time on natural law indicates merely that some people believed that they are related in a particular way. 😉

  5. Samir Chopra says:


    You’re welcome. That technical perspective was really very useful.

  6. Samir Chopra says:


    I think the point in suggesting ‘person’ as a fiction was just that: societies that reach levels of organization that require blame and responsibility settle on the notion of an agent, selected from the flux around us, as a cause for events. These become reified into ‘persons’. This is a pragmatic choice and does not mark out some privileged ontological category.

    I hear you on point 4 :)