AI: An Exercise in Analytical Philosophy

An excellent analysis of the issue! It seems like this is a problem for any interesting philosophical question. I’m always struck by how often philosophical disagreements are really just definitional disputes in disguise. It’s particularly troublesome for any discussion about the mind, about us at the most fundamental level, because people have intense emotions about the conclusions.

Blogging to Share Knowledge

The Question

I recently attended a computing group in which the following question was asked:

Can Software Achieve Human Level Intelligence?

We covered this question over the course of 3 meetings (7-9 hours total).  Those meetings didn’t go well.  We spent hours talking past each other, objecting to arguments, and accusing each other of missing the point.  In the end we gave up and agreed to talk about something else.

How can people spend hours talking about something without anything to show for it?Because we weren’t talking about the same thing and never settled on meanings first to discover that.

After reflecting on those meetings, I thought the following process would have been more productive:

  1. Each person precisely translates the sentence as they understand it.
  2. Replace any contentious terms to avoid arguing over semantics.
  3. Decide if there’s any basis for arguing.

So let’s continue by assuming two arguers: John and…

View original post 527 more words

44 thoughts on “AI: An Exercise in Analytical Philosophy

    1. Most AI researchers think that, in the grand scheme of things, the difference between the village idiot and world renowned genius is actually pretty small, at least compared with the differences between the average human and a chimpanzee. Of course, we’re still a long way from mouse intelligence, so there’s a long road ahead.

      Liked by 1 person

      1. Software for the world renowned genius could probably be written in MS-DOS Basic by a Junior programmer. It just requires of couple of loops with variations of statements about being the world’s greatest genius. 🙂

        Liked by 1 person

          1. We should turn to the “stable genius” for the answer to the “hard problem”. It will be an easy problem, maybe even a fake problem, or a hoax. He could solve it I’m sure.

            “People say it’s a hard problem. It’s not a hard problem for a stable genius like me. It’s only hard for Sleepy Joe and the low IQ Democrats.”


  1. Factor in “ontological antirealism” and those definitional issues, and it seems that while philosophy is a great exercise for the mind, it doesn’t actually contribute much, if anything, to science. It is, after all, mostly opinions. I’m not really sure one even has a “science” until one has definitions for what one is studying. As such, the whole field is filled with charlatans, wannabes, and the mystic.

    There is also that a lot of the people actively debating this should take debate class. It’s depressing how many people don’t really understand what constitutes a valid argument. The dialectic is a skill that, like any other, has to be explicitly learned and practiced.

    Liked by 1 person

    1. I have to admit that I often get fed up with philosophy, particularly the philosophy of mind. It can contribute by clarifying definitions and questions, but too often it devolves into pseudological rhetoric to justify preexisting biases, or outright speculative fantasy. I don’t mind fantasy, when it’s clearly labeled as such, but not when it’s taken to actually be proving something.

      The nice thing about science is it can challenge our biases with evidence that’s hard to dismiss. That said, a lot of science is influenced by the philosophy of the scientists, and all observation is theory-laden. So to some extent, we can’t get around the philosophy.


      1. The “philosophy of scientists” isn’t typically philosophy so much as just a world view. (As, for example, my life-long “philosophy” of always eating dessert first.) Certainly we are framed by our world view, but the dialectic and rigorous debate are tools to mitigate that.

        The alternative is to just give up and decide that, given our biases, there’s just no point in discussing anything.


        1. Many scientists do explicitly cite philosophers for guiding principles. Among cognitive scientists, Antonio Damasio often cites William James; Michael Graziano, Daniel Dennett; and Jon Feinberg and Todd Mallatt, John Searle. Of course, many scientists, such as Barnard Baars, Stanislas Dehaene, and Elkhonon Goldberg, along with numerous physicists, are dismissive of philosophy, so it’s a pretty mixed thing.

          Liked by 1 person

          1. And Judea Pearl cites David Lewis, and Sean Carroll cites David Wallace. Not surprising when you notice that a lot of (the best) philosophy lives alongside and among sciences.

            Liked by 2 people

  2. I sincerely hope that software can do much better than human level intelligence. How depressing it would be to labour long to create intelligent software then find out it had the same selfishness, prejudice, neuroses and self delusion as us.

    Liked by 2 people

    1. It’s worth noting that ANNs often acquire the same biases that plague humans. A lot of human reasoning is probabilistic, and it might be that that kind of reasoning inevitably leads to biases, which when you think about it, are simply erroneous associations based on insufficient information.


  3. The problem with AI intelligence is that the AI itself has no stake in the game. It has nothing to gain or lose. It doesn’t feel pleasure or get hurt if it makes a mistake. It has no reason to live or any reason to care if it dies. The concepts of living and dying are meaningless. It can have no real experience of living, no stake in the game, without a body that feels.

    Aside from that, anything can be simulated. It can be made to look and act human. The closer it approaches actual human the greater it will feel itself not human – the more acute its lack of a body will become. Humanity for AI will be like grapes for Tantalus always receding at the moment of clasping it.

    Liked by 1 person

    1. It seems like we could give an AI any stake we wanted to. Of course, it probably wouldn’t be productive to give it self concern as an overriding priority. (Assuming the people who say that’s the only path to general intelligence aren’t right.) But a mine sweeping robot’s stake would be in discovering as many mines as possible before its existence ends.

      I do agree we should think carefully before giving an AI human feelings. Until we can get them to the point where they might be accepted as fellow people, it might be an act of extreme cruelty.


      1. Unless you give it life itself, you can’t really give it a full stake.

        Living organisms are born, go through a growth and learning process, procreate, get sick, get well, experience old age, and die. You can simulate a lot but there are some things you can’t. The learning and growth process may be critical for general intelligence. Part of learning and growth is an individuation process, although I read recently this even begins in the womb.

        If you are only thinking of AI dedicated to specific tasks (like mine sweeping), of course, that will be possible but it sounds a little ridiculous to me saying you are giving it a stake by having it find as many mines as it can.


        1. I think most of what you describe for a living organism is irrelevant if we’ve succeeded in reproducing the functionality. It amounts to privileging the perspective you’re most familiar with. But our desire to survive and procreate amounts to programming, programming by natural selection. Privileging that programming over a robot’s programming to find mines is arbitrary, at least if the robot has capabilities for simulating multiple paths to achieve that programmed goal.


          1. Finding mines might be something human intelligence can do but it isn’t something human intelligence evolved to do. The question here, I thought, was about human intelligence. Human intelligence is generalist, creative, and goal directed. Will the mine finder AI also write poetry or enjoy music?


          2. Finding mines can be creative and goal directed. But that’s only an example, and how generalist the system is depends on how broadly we want to scope its primal impulses.

            Writing poetry and enjoying music heavily stacks the deck in favor of humans. Art requires profound insight into the human condition, so much so that most actual humans can’t create compelling versions of it. There’s no reason in principle machine won’t eventually be able to do it, but “eventually” may be a very long time from now.


  4. The author’s recommended procedure is a good one, but there is a big mistake when they say that agreement ends the debate. Just because two people agree on a claim, doesn’t mean they can’t productively question it. A simple example: they might find they have radically different reasons for agreeing on the proposition, and each may reject the other’s reason. Digging in, they might find that both reasons are inadequate.

    It’s also worth bearing in mind that the distinction between verbal and substantive questions is a fuzzy one. In the construction of knowledge, definitions come at a very late (high-level) stage, during theory development. Items that look clustered-together at one stage, fit for a single label, can come apart under further scientific investigation. Items that look separate can be unified.

    Liked by 2 people

    1. Your first paragraph reminds me that one the things I find most annoying is when someone gives a bad argument for a conclusion I agree with. Definitely, the reasons matter.

      On verbal vs substantive questions, I get what you’re saying here, but I think BIAR would point out that he’s talking about after there has been discussion and enough interaction to work through the initial superficial layers. But this reminds me why I’m often reluctant to except -ism labels. They almost always come with commitments I’m not comfortable with, even when I agree 80-90% of the -ism.


  5. I seem to have gotten caught in BAIR’s spam folder, though I’ll also say some things about this matter here.

    It’s notable that you and I have this same dispute going regarding “software consciousness”, though not hindered by means of definitional issues that I’m aware of — we simply disagree. If so then this implies that our skills in epistemology should be pretty good! Like BIAR, perhaps we could help others discuss contentious issues more productively. But do we indeed have such skills?

    Each of us believe that when my thumb gets whacked, associated information goes to my brain for processing. Your position as I understand it is that such processing then creates what I feel. Thus you reason that anything which is able to do such processing (like someone or thing able to quickly turn certain symbols on paper into appropriate “thumb pain” symbols), will thus create what I know of as “thumb pain”.

    I believe that you’ve got this right, except an additional step is needed as well to preserve causality (which of course you dispute). I believe that those symbols could only produce such phenomenal experience if fed into a machine that’s armed with the proper physics (or somewhat how computers are able to animate computer screens). And what sort of physics might brain information animate to produce various kinds of phenomenal experience? I suspect that certain neuron firing produces a conscious entity made up of associated electromagnetic radiation.

    Do you think this demonstrates that we’re extra skilled in epistemology, or have we instead been talking past each other?

    Liked by 2 people

    1. Eric,
      It’s typical for the first comment on a blog to require approval, so unless you’ve commented on BIAR’s blog before, you’re likely just caught in his moderation queue until he notices and approves it.

      I don’t think the difference between us on this is definitional. There have been periods where I suspected it might be, but I think we’ve discussed it enough, and your description of it above serves, to make me think it’s definitely an ontological difference. I don’t know I’d say we’re “extra skilled” in epistemology, although we might be in comparison to people who don’t usually have these types of discussions.

      Liked by 1 person

      1. Well I think we’re extra skilled Mike. Epistemology was actually the point of my comment over at BIAR’s. This gets into my position of how science (and especially on the soft side), suffers tremendously without a respected community of professionals who’ve developed generally accepted principles of metaphysics, epistemology, and axiology.

        On comment approval, I’m actually in a worse spot than that. My commentary has been banned from three blogs that I know of, and so WordPress demands first time approval from bloggers who aren’t expecting approval to be required. I guess I should change my email address or something.

        Liked by 2 people

      2. The WordPress policy that I read did specifically mention spam, though I suspect they realize that they’re casting a wide net here. Surely I don’t profile out as an actual spammer!

        As for what I’ve done to be banned at three different blogs, I’m sure you know that Massimo Pigliucci has always fostered spirited debates at his various blogs over the years. Also, I tend to make radical claims which may initially seem far less difficult to dismiss than they actually are. While Massimo himself quickly learned to grant me a wide berth to instead feast upon other prey, he’d certainly cheer for those who would challenge me. Two very different bloggers over there banned my commentary, not because of what I said at their blogs, I think, but rather because of how I was able to confound each of them at Massimo’s.

        I’ll accept far more blame for the first incident though. Some unfortunate circumstances existed, but in the end this was regrettably my bad.

        Liked by 1 person

        1. My question about spam came in the context of bloggers specifically reporting your comments as spam. From what I’ve heard from others who got caught in that net, it only takes a few bloggers to hit the spam button on your comments to put you in that category, although the algorithms seem to adjust depending on your history at each blog. I personally only hit that button for very obvious spam, but some hit it for off topic comments.

          But if those bloggers are blocking you with their Blocklist filter, then that shouldn’t, in and of itself, affect you at other blogs. (Unless has some system in place I’m not aware of, which is possible.) I’ve never outright blocked anyone myself (although I have flagged a couple of people for moderation before), so I’m not sure if it triggers anything.

          Liked by 1 person

      3. Checking around on this I noticed links to ask Akismet questions about any issues that one might be having. I’ve done so of course. They provide a reasonable questionnaire from which to simulate your last failed blog comment for their testing purposes. My technician Tracy looked into this and then made some adjustments. The results? Apparently unsuccessful. I’ve told them about this so maybe they’ll work it out in the end. I’m still not sure if others have flagged me for spam, though I suspect that wasn’t it.

        Liked by 1 person

      4. I just wanted to say that Akismet does seem to have come through in the end. It took several tries, and they once said that their logs showed that I’d leave a great deal of comments in a short period of time like a spambot, which seems strange to me. But I guess it’s all set now.

        Liked by 1 person

  6. Human brains aren’t trying to communicate when they speak. They are trying to coordinate. Language is a practical tool to achieve certain goals, not to attain “knowledge”. For instance, why am I commenting here? To engage in the argument would be a common answer. But if “why” means “what am I trying to get out of it?” then the answer, especially in connection to evolution, is not so clear. As Cioran put it, “the living being as such exists and is real only when threatened.” What does it mean for science to be the domain of the frightened?

    Liked by 2 people

    1. Wouldn’t you say coordination is an application of communication?

      The reasons why we engage in these conversations seem complex and multifaceted. They obviously trigger some instinctive impulses, although it’s possible those impulses, evolved in a hunter-gatherer context, are misfiring.

      Not sure if I follow that last question.


      1. Other animals, when faced with nothing to do, just go to sleep. Humans, in their anxiety, create culture and civilization. Culture is what happens when frightened beings, who have eliminated immediate threats but are still constitutionally fearful, attempt to assuage their anxiety through “meaning” when consciousness has no immediate practical focus (boredom). And so, humans are an aberration. But humans don’t consider themselves or what they do as aberrant, but as normal. This is the human, and scientific, stance: that it “normal” to do what we do and not the desperate attempt of a strange species to find something that will keep our fear at bay.


        1. That is one way of describing it. Another is that we’re the only species that can foresee our long term future, and the challenges and threats that may come in that distant future. We are probably the only species aware of our own mortality, that our death is inevitable. And I agree that much of our culture results from it.

          But if a dog had the same foresight, I suspect they’d be just as narcissistic.


  7. Mike,
    At the risk of going off topic and being flagged for spam 🙂 , I’ve been dumbfounded by the hysteria that we see today regarding this COVID-19 business. Though I can imagine pathogens which would warrant such caution, this one just doesn’t seem to add up. From what I’ve heard it’s not very fatal or debilitating, and the symptoms are pretty clear to someone who has it. Conversely we could be faced with sleeper pathogens that are ultimately quite debilitating. Now that’s scary!

    And yet today we should all bunker down with basic foodstuffs, bottled water, and a mountain of toilet paper? That seems to be the conventional wisdom regardless. The main question I currently have is how deep and extended will the resulting recession be? This should depend somewhat upon how long it will take for respected health officials to come out on popular media and tell the masses “Look, we’re working on vaccines for this, and as soon as possible all flu shots will be spiked with it. In the mean time please just go about your normal lives with a bit of caution.”

    Liked by 1 person

  8. Hey Mike, is there a place on your website for suggestions?

    Since this post is about philosophy of mind, here I’ll mention a new phil. mind book that seems right up your (and my) alley: “The Emotional Mind” by Stephen T Asma and Rami Gabriel. Discussed recently on

    Liked by 2 people

    1. Thanks Paul! No official spot. A comment is as good as anything else.

      I do know about Asma ad Gabriel’s book. They recently did a series of posts on The Brains Blog.
      Unfortunately, I find their writing too dense, verging on obfuscatory, requiring too much work to disentangle. They’ve identified themselves before as acolytes of the late Jaak Panksepp, whose writing is more clear, with The Archaeology of the Mind being worth checking out.

      Liked by 1 person

Your thoughts?

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

This site uses Akismet to reduce spam. Learn how your comment data is processed.