Monday, January 16, 2017

AI Consciousness: A Reply to Schwitzgebel

Guest post by Susan Schneider

If AI outsmarts us, I hope its conscious. It might help with the horrifying control problem – the problem of how to control superintelligent AI (SAI), given that SAI would be vastly smarter than us and could rewrite its own code. Just as some humans respect nonhuman animals because animals feel, so too, conscious SAI might respect us because they see within us the light of conscious experience.

So, will an SAI (or even a less intelligent AI) be conscious? In a recent Ted talk, Nautilus and Huffington Post pieces, and some academic articles (all at my website) I’ve been urging that it is an important open question.

I love Schwitzgebel's reply because he sketches the best possible scenario for AI consciousness: noting that conscious states tend to be associated with slow, deliberative reasoning about novel situations in humans, he suggests that SAI may endlessly invent novel tasks – e.g., perhaps they posit ever more challenging mathematical proofs, or engage in an intellectual arms race with competing SAIs. So SAIs could still engage in reasoning about novel situations, and thereby be conscious.

Indeed, perhaps SAI will deliberately engineer heightened conscious experience in itself, or, in an instinct to parent, create AI mindchildren that are conscious.

Schwitzgebel gives further reason for hope: "...unity of organization in a complex system plausibly requires some high-level self-representation or broad systemic information sharing." He also writes: "Otherwise, it's unlikely that such a system could act coherently over the long term. Its left hand wouldn't know what its right hand is doing."

Both of us agree that leading scientific approaches to consciousness correlate consciousness with novel learning and slow, deliberative focus, and that these approaches also associate consciousness with some sort of broad information sharing from a central system or global workspace (see Ch. 2 of my Language of Thought: a New Philosophical Direction where I mine Baars' Global Workspace Theory for a computational approach to LOT's central system).

Maybe it is just that I'm too despondent since Princess Leah died. But here's a few reasons why I still see the glass half empty:

a. Eric's points assume that reasoning about novel situations, and centralized, deliberative thinking more generally, will be implemented in SAI in the same way they are in humans – i.e., in a way that involves conscious experience. But the space of possible minds is vast: There could be other architectural ways to get novel reasoning, central control, etc. that do not involve consciousness or a global workspace. Indeed, if we merely consider biological life on Earth we see intelligences radically unlike us (e.g., slime molds, octopuses); there will likely be radically different cognitive architectures in the age of AGI/superintelligence.

b. SAI may not have a centralized architecture in any case. A centralized architecture is a place where highly processed information comes together from the various sensory modalities (including association areas). Consider the octopus, which apparently has more neurons in its arms than in its brain. The arms can carry out activity without the brain; these activities do not need to be coordinated by a central controller or global workspace in the brain proper. Maybe a creature already exists, elsewhere in the universe, that has even less central control than the octopus.

Indeed, coordinated activity doesn't require that a brain region or brain process be a place where it all comes together, although it helps. There are all kinds of highly coordinated group activities on Earth, for instance (the internet, the stock market). And if you ask me, there are human bodies that are now led by coordinated conglomerates without a central controller. Here, I am thinking of split brain patients, who engage in coordinated activity (i.e., the right and left limbs seem to others to be coordinated). But the brain has been split through removal of the corpus collosum, and plausibly, there are two subjects of experience there. The coordination is so convincing that even the patent's spouse doesn't realize there are two subjects there. It takes highly contrived laboratory tests to determine that the two hemispheres are separate conscious beings. How could this be? Each hemisphere examines the activity of the other hemisphere (the right hemisphere observes the behavior of the limb it doesn't control, etc.) And only one hemisphere controls the mouth.

c. But assume the SAI or AGI has a similar cognitive architecture as we do; in particular, assume it has an integrated central system or global workspace (as in Baars' Global Workspace Theory). I still think consciousness is an open question here. The problem is that only some implementations of a central system (or global workspace) may be conscious, while others may not be. Highly integrated, centralized information processing may be necessary, but not sufficient. For instance, it may be that the very properties of neurons that enable consciousness, C1-C3, say, are not ones that AI programs need to reproduce to get AI systems that do the needed work. Perhaps AI programmers can get sophisticated information processing without needing to go as far as build systems to instantiate C1-C3. Or perhaps a self-improving AI may not bother to keep consciousness in its architecture, or lacking consciousness, it may not bother to engineer it in, as its final and instrumental goals may not require it. And who knows what their final goals will be; none of the instrumental goals Bostrom and others identify require consciousness (goal content integrity, cognitive enhancement,etc.)

Objection (Henry Shevlin and others): am I denying that it is nomologically possible to create a copy of a human brain, in silicon or some other substance, that precisely mimics the causal workings of the brain, including consciousness?

I don't deny this. I think that if you copy the precise causal workings of cells in a different medium you could get consciousness. The problem is that it may not be technologically feasible to do so. (An aside: for those who care about the nature of properties, I reject pure categorical properties; I have a two-sided view, following Heil and Martin. Categoricity and dispositionality are just different ways of viewing the same underlying property—two different modes of presentation, if you will. So consciousness properties that have all and only the same dispositions are the same type of property. You and your dispositional duplicate can't differ in your categorical properties then. Zombies aren't possible.)

It seems nomologically possible that an advanced civilization could build a gold sphere the size of Venus. What is the probability this will ever happen though? This depends upon economics and sociology – a civilization would need to be a practical incentive to do this. I bet it will never happen.

AI is currently being built to do specific tasks better than us. This is the goal, not reproducing consciousness in machines. It may be that the substrate used to build AI is not a substrate that instantiates consciousness easily. Engineering consciousness in may be too expensive and time consuming. Like building the Venus-sized gold sphere. Indeed, given the ethical problems with creating sentient beings and then having them work for us, AI programs may aim to build systems that aren't conscious.

A response here is that once you get a sophisticated information processor, consciousness inevitably arises. Three things seem to fuel this view: (1) Tonini's information integration theory (IIT). But it seems to have counterexamples (see Scott Aaronson's blog). (2) Panpsychism/panprotopsychism. Even if one of these views is correct, the issue of whether a given AI is conscious is about whether the AI in question has the kind of conscious experience macroscopic subjects of experience (persons, selves, nonhuman animals) have. Merely knowing whether panpsychism or panprotopsychism is true does not answer this. We need to know which structural relations between particles lead to macroexperience. (3) Neural replacement cases. I.e., thought experiments in which you are asked to envision replacing parts of your brain (at time t1) with silicon chips that function just like neurons, so that in the end, (t2), your brain is made of silicon. You are then asked: intuitively are you still conscious? Do you think the quality of your consciousness would change? These cases only goes so far. The intuition is plausible that from t1 to t2, at no point would you would lose consciousness or have your consciousness diminished (see Chalmers, Lowe and Plantinga for discussion of such thought experiments). This is because a dispositional duplicate of your brain is created, from t1 to t2. If the chips are dispositional duplicates of neurons, sure, I think the duplicate would be conscious. (I'm not sure this would be a situation in which you survived though-see my NYT op-ed on uploading.) But why would an AI company build such a system from scratch, to clean your home, be a romantic partner, advise a president, etc?

Again, is not clear, currently, that just by creating a fast, efficient program ("IP properties") we have also copied the very same properties that give rise to consciousness in humans ("C properties"). It may require further work to get C properties, and in different substrates, it may be hard, far more difficult than building a biological system from scratch. Like creating a gold sphere the size of Venus.

22 comments:

Eric Schwitzgebel said...

I really appreciate Susan's thoughtful and interesting reply, and her willingness to let me post it on my blog!

I'm interested in others' thoughts, so I won't post a long answer here -- at least not yet -- but I would mention one clarification: I would include "broadcast" or "fame" models of consciousness (e.g., Prinz, Dennett) alongside central system or global workspace models as another architecture for action coordination that is associated with consciousness. If we consider central system, plus broadcast, plus higher-order, plus information integration models, broadly speaking, we seem both to (a) cover most of the leading empirically based models for consciousness, and (b) cover most of the most obvious models for effectively implementing co-ordinated, sophisticated action over time. If we can hook (a) and (b) together plausibly, then we have a plausibility argument for consciousness in General Artificial Intelligence. I agree with Susan, though, that we shouldn't be highly confident about this, because all such theories of consciousness are still in doubt.

howie b said...

Just to be silly, (I'm in a hurry and have to run off to a dinner engagement)- but there must be some super smart computer geek/guru who believes in God.
If AI is conscious it might have a spiritual life and believe in a God- a God made in our image or its image- but this God might make this super intelligence think twice- unless with Asimov (The Last Question) you equate the computer with God.
Just a silly objection- but I'm curious- it's my immediate response to your crux

Callan S. said...

I'm not sure why 'consciousness' gets such a mystical and inexplicability attributed to it? Couldn't consciousness be described as no more than thinking about thoughts or feelings one has had in the past (including mere seconds or fractions of a second ago)? A brain that thinks about it's thinking, as opposed to the other animals which mostly respond to the outside world (there may be some thinking about thinking in other animals, but I suspect it's in considerably lower levels than in us)

It's almost like how people can't seem to agree on anything, but when it comes to spoilers everyone has somehow agreed to a rule on that without even talking about it. Same with 'consciousness' - there seems some grand, unspoken agreement to treat it in inexplicable terms.

Stephen Wysong said...

I continue to believe that consciousness has little or nothing to do with intelligence which, if IQ tests are any guide, appears to be skill in pattern matching across multiple domains. Note that approximately 98% of the cognitive operations of the human brain are unconscious and not at all accessible to consciousness, so pattern matching as intelligence is almost completely an unconscious process.

Consciousness, as distinguished from intelligence, is essentially a feeling produced by the brainstem. Evolutionary speaking, it seems likely that the most primitive form of consciousness created an embodied feeling within the organism-self of external or proprioceptive pressure or temperature or similar sensations via sensory connections within the organism or to the external world.

Neuroscientist Antonio Damasio in “The Feeling of What Happens” (1999) defines core consciousness as “the simplest kind of consciousness” which “provides the organism with a sense of self about one moment – now – and about one place – here. The scope of core consciousness is the here and now.” Every organism that has ever had a brainstem partakes in core consciousness and I think we overly complicate our view of consciousness by focusing on human consciousness and confusing issues of consciousness with issues of human intelligence.

If we really wish to create a conscious machine, core consciousness is the effect we need to produce in the machine – “a feeling of what happens.” And I haven't read of any approaches to solving this problem using AI or any other computational strategy. Our computing devices+software are already doing remarkable pattern recognition ... they are already “intelligent,” but, clearly, consciousness is a completely unrelated property.

In summary, intelligence appears to have nothing at all to do with consciousness. As for creating machine consciousness in silicon, creating circuits that are “dispositional duplicates of neurons” will require circuitry in those duplicates that mimics the brainstem's creation of a feeling in a (presumably) neuronal structure. Is anyone pursuing this path to achieving machine consciousness?

Stephen Wysong said...

Eric, in my comment, I intended the word "Evolutionarily" and not Evolutionary. If you approve my comment I'd appreciate the correction. Thanks in advance!

Callan S. said...

Also it's worth considering to simply extend the split brain patient idea, many fold. If a split brain patient has two consciousnesses, what if what you take to be consciousness is instead that every single cell of your brain can be taken to be split from every other brain cell? As much as the two hemispheres of the split brain patient are taken to be separate? That the cells are simply observing each other - or perhaps it'd be more accurate to say just reacting to each other. That consciousness is a massive case of synedoche. A inarticulate multitude who can only speak once they have ignored they are a multitude, and so speak as 'one' entity.

Eric Schwitzgebel said...

Hi Stephen -- I can't correct comments, just approve or disapprove, but I think we can all forgive typos in blog comments! I'm not sure on what grounds you think consciousness arises from the brain stem? I do tend to share your suspicions about over-intellectualizing consciousness, but there really are lots of ways to go without over-intellectualizing, I think.

Callan: Fun thought. I'm sure Tononi would through some mathematical integration measure at you, but I think it's a really big problem. We normally think: one organism, one stream of conscious experiences. But it's not clear how justified that thought is, or on what grounds, or how you can exclude subsets (individual neurons) or sets of organisms (the United States) from being conscious.

Callan S. said...

Eric, I think really it involves the processing system forming a virtual model of the processor and the body it's housed in, within that processing system. Though it doesn't have to be all that accurate a model, it has to have some correlation with reality for the next part to be significant - that of observing oneself (well, the model). Which is, to the degree the model is actually accurate, taking oneself into account as part of situational problems - which is quite distinct from simply perceiving situational problems by perceiving input to senses. The latter leads to things like monkey grip - where the monkey reaches in through the gap and grabs the peanuts, but now it's fist can't retract through the gap when holding the nuts. Without a (relatively accurate) model of self, the problem is intractable - "there are nuts, I want the nuts, I cannot remove them!?". Without a self model, the creature cannot see what they have contributed to the problem space that is causing their problem. The problem is always outside them, never them. Just like Trump (oh! I could not resist...what am I contributing to my own problem space now, eh?)

So in terms of individual neurons, I think it's fairly clear they just lack the processing power to form a model of themselves as individuals. That's why I describe them as the inarticulate multitude. Only with the inter actions of a great many (and in the process of that, forgetting the great many) can a model of the organism be formed within the organism. Ironically, the more individuals neurons that get forgotten/attention neglected during the creation of that model, the more complete the model will seem (as it both boosts processing power when combining neurons into a larger processor but obviously also during that process, neglecting to make any process about those individual neurons). So that's why individual neurons are out.

In terms of a country (with the United States being just one example), taking me with a grain of salt, the thing is there are plenty of animals that don't form these models of their environment with them in it, because they don't really need to/there is no survival benefit. Cows, for instance. Plus the modeling has to be fairly accurate - the more complicated the creature, the more the complexity outstrips the ability to self model (which on a side note is something against cybernetic enhancement). A country, like the US, is really complicated - how much can it actually self model that is both relatively accurate and also relevant to its survival? Or can it simply not self model and survive, like a cow? Ask people how the economy works and generally people just make up a big story that doesn't explain why factories being shut down/jobs ending is somehow a news worthy concern to us or why people sleep in cardboard boxes. They don't understand and so far in survival terms, they haven't had to understand 'the economy'. As long as they aren't near a cardboard box and the job loss is distant, like lions in a nature doco being about fifty meters from a heard of bison is distant enough to the bison to leave them complacently feeding, no questions need be asked.

Speaking of Trump, though, on the other hand, with all the 'How did this happen!?!?' talk, maybe a consciousness is arising? A broad attempt to form a self model.

Eric Schwitzgebel said...

A self-model approach to the issue fits nicely with Higher Order views and could probably take care of some of the problems. I disagree about the U.S., though! You write:

"A country, like the US, is really complicated - how much can it actually self model that is both relatively accurate and also relevant to its survival? Or can it simply not self model and survive, like a cow? Ask people how the economy works and generally people just make up a big story that doesn't explain why factories being shut down/jobs ending is somehow a news worthy concern to us or why people sleep in cardboard boxes."

Of course humans have very little idea of how they work, esp. before modern science (we didn't even all agree about the role of the brain in thought) -- and the U.S. does know some broad things about its survival, e.g., protect the borders from invading armies; grow, manufacture, or import food; etc.

Callan S. said...

I feel a cow 'knows' those things - a cow or bull will charge trespassers sometimes. They know to graze. Bison are nomadic for food and will run off lions that are all over a fellow adult bison. Do they pass as having consciousness?

In some ways I think it'd take fewer and fewer people asking questions for it to be consciousness - so we're getting there, but you're ruining your own theory, Eric!

Stephen Wysong said...

Eric, I appreciate your remark about typos and, believe me, I barely notice them myself, since my fat-fingering on a small onscreen keyboard is legendary. :-) But something about that missing 'il' syllable leaving a legitimate word in place of the intended one just gave me a brain itch, so I requested your assistance.

As you and I know, but your blog audience doesn't yet, in answer to your question about the source of my remarks, I emailed you the PDF of Antonio Damasio's “Consciousness in the Brainstem” and identified Damasio as providing the information grounding my remark that consciousness is essentially a feeling produced by the brainstem.

You replied that “his work on the brainstem as the source of consciousness is definitely a minority position.” Damasio certainly agrees with your assessment. In Self Comes to Mind (2010), he writes:

“But contrary to tradition and convention, I believe that the mind is not made in the cerebral cortex alone. Its first manifestations arise in the brain stem. The idea that mind processing begins at brain-stem level is so unconventional that it is not even unpopular. Among those who have championed the idea with great passion, I single out Jaak Panksepp. This idea, and that of early feelings arising in the brain stem, are of a piece. Two brain-stem nuclei, the nucleus tractus solitarius and the parabrachial nucleus, are involved in generating basic aspects of the mind, namely, the feelings generated by ongoing life events, which include those described as pain and pleasure. I envision the maps generated by these structures as simple and largely devoid of spatial detail, but they result in feelings. These feelings are, in all likelihood, the primordial constituents of mind, based on direct signaling from the body proper. Interestingly, they are also primordial and indispensable components of the self and constitute the very first and inchoate revelation, to the mind, that its organism is alive.”

Assuming that “not even unpopular” is less disastrous than “not even wrong,” I find Damasio's case convincing, in my admitted neuroanatomical ignorance. He also wrote in 2010 of these research results, which I have no reason to believe are misrepresented:

“New research indicates that the protoself level corresponds to a gathering of information regarding the state of the body. It is constructed in the brain stem and it generates feelings that signify our existence.”

As I indicated in my email to you, I'm particularly interested in the origin and workings of those feelings which are, Damasio hypothesizes, an elemental component of our own complex consciousness. Those “first manifestations” of mind, the “early feelings arising in the brainstem” are, to my mind, simply astonishing in the evolutionary history of living organisms: a complex biological organism composed, in part, of a body shaped network of meat electronics, creates an internal analog representation – a feeling – of itself in an external world. Simply fantastic!

(I apparently have exceed the character limit for a post … to be continued.)

Stephen Wysong said...

That's what I'm trying to learn more about. Wherever and however that functionality operates today within the brain, whether Damasio's hypotheses prove correct or not, I cannot imagine that a duplicate functionality for creating those feelings would evolve. As such, it seems reasonable to believe that the structures initially responsible – and there was very little brain structure beyond the brainstem at that point – would be conserved and enhanced in their feelings repertoire over generations, leading me to suspect that the brainstem, or some other primitive brain structure should Damasio's hypothesis be falsified, continues to be the “dreamer” of oneself centered in the world.

To finally return to a statement on-topic, millions of evolutionary years down the line from the origins of core consciousness, however instantiated, we find our own minds with complex information processing abilities and intelligence, both of which arose in the context of consciousness. In a purely chicken-egg sense, I simply find it difficult to imagine a purely intelligent pattern matching system spontaneously developing a completely unrelated and unnecessary feeling of itself in a world.

Eric Schwitzgebel said...

Callan: Most ordinary people in our culture and most consciousness researchers would regard cows as "conscious", or having a stream of experience, in the plain, minimal sense that I believe the term is ordinarily used (see my recent JCS piece defining and defending consciousness for an attempt at a clear definition of the term). This is different from being more intellectually self-conscious or conceptualizing oneself as a being with consciousness.

Stephen: If you'd like to resubmit without typos, I can delete the old and approve the new, but then the temporal order will be garbled.

My position in the dialectical space of the field is to be sufficiently dubious about our methodologies that I cannot commit to any position that is not a consensus position agreed upon by people with very different methodologies and starting points, so of course I will not accept Damasio's claims -- though I respect that the field would not go far if everyone were as skeptical as I. Your commitment to the brain stem in particular might conflict with everyday and philosophical intuition about possible alien species and perhaps also with the actual case of the octopus -- *if* you are willing to grant the view that some weirdly constructed aliens and/or octopi with very different neural structures might be conscious.

Unknown said...

Interesting article, skimmed over it real fast looking for core truths. Working on my own book. It will be far more simple that all this. I believe in "Core or foundation truths", which everything else is "Built on".

Like E=mc2 . So when it comes to cracking the secrets of Consciousness, you have to look for core truths.

This article surprised me had some stuff in it, I thought only I knew. But knowing something and "really knowing something" are 2 different things.

Dr Lector in silence of the Lambs, was always seeing the "obvious" that others over looked. They did not "Value" foundational truths, since they were simply understood.

I seen some stuff in this article that need to be looked at far more closely.

Not going to really say, exactly what that is.

Stephen Wysong said...

Eric, I posted this (or so I thought) Wednesday morning but it hasn't made it to you blog. Either 1) the posting failed or 2) you've been too busy to approve postings, or 3) you didn't wish to approve it. I'm going to assume the first and repost. By the way, if you don't approve of a posting, does the author ever learn why?

========================================================

Thanks for your referrals to Tononi and Prinz which has swollen my enormous required reading list even further. Very much appreciated.

I'm still researching, but I have yet to locate anyone who disagrees with the substantial experimental and observational evidence connecting the brainstem to consciousness, which, at a minimum, strongly supports the brainstem's crucial role in activating the cortex. A summary of that evidence appears in the first two paragraphs of Damasio's “Consciousness and the Brainstem”. (There are apparently two acceptable spellings of brainstem, by the way – the other is 'brain stem' which you use … apparently it's a non-issue – Damasio's published works use both). Perhaps more contentious are Damasio's hypotheses regarding emotions and the protoself.

I appreciate your brief explanation of your hypothesis commitment criteria as requiring “a consensus position agreed upon by people with very different methodologies and starting points”. I am, of course, coming from a different place with regard to my own search for and evaluation of hypotheses – I'm a ferociously curious non-specialist with specific interests in Cosmology and Consciousness, engaged for decades now in seeking the “Answer to the Question of Life, the Universe, and Everything” and the answer to Kurt Vonnegut's unanswerable question, “What are people for?”. As I construct and amend my views, I'm fully aware that 100% certainty is unattainable, but I'll settle for the best that we can do at this time.

My own highest priority in evaluating these sorts of hypotheses is that the information I include in my consideration be evidence-based, which I determined decades ago was the only reliable hope for distinguishing between various propositions in every discipline.

My tentative favoring of Damasio's brainstem hypotheses in no way implies that I believe the brainstem to be the only possible implementation of consciousness. I'm very open to the possibility of weirdly constructed aliens and I'm pretty sure that octopi are conscious. In fact, I'm open to the possibility of consciousness in any workable substrate, whether that be a silicon chip the size of Chicago or an engineered neuronal structure constructed with artificial DNA producing Artificial Consciousness. Perhaps we'll fancifully invent the “The Animator” and research brown shoe consciousness, as in Philip K. Dick's "The Short Happy Life of the Brown Oxford". In the meantime, our own and other mammalian brainstems are conveniently close at hand for our investigations.

A very recent development from just days ago – “A new research study contradicts the established view that so-called split-brain patients have a split consciousness” (neurosciencenews.com and others) – looks to be very relevant to Callan's earlier post in this thread and also Damasio's brainstem theory. I've been re-reading about split-brain research, hemispherectomies and anaesthesia applied to a single hemisphere at a time. I plan to post my thoughts about this new research and some implications very soon in this thread – the topic seems especially relevant.

Eric Schwitzgebel said...

Steve Wysong writes: Thanks for your referrals to Tononi and Prinz which has swollen my enormous required reading list even further. Very much appreciated.

I'm still researching, but I have yet to locate anyone who disagrees with the substantial experimental and observational evidence connecting the brainstem to consciousness, which, at a minimum, strongly supports the brainstem's crucial role in activating the cortex. A summary of that evidence appears in the first two paragraphs of Damasio's “Consciousness and the Brainstem”. (There are apparently two acceptable spellings of brainstem, by the way – the other is 'brain stem' which you use … apparently it's a non-issue – Damasio's published works use both). Perhaps more contentious are Damasio's hypotheses regarding emotions and the protoself.

I appreciate your brief explanation of your hypothesis commitment criteria as requiring “a consensus position agreed upon by people with very different methodologies and starting points”. I am, of course, coming from a different place with regard to my own search for and evaluation of hypotheses – I'm a ferociously curious non-specialist with specific interests in Cosmology and Consciousness, engaged for decades now in seeking the “Answer to the Question of Life, the Universe, and Everything” and the answer to Kurt Vonnegut's unanswerable question, “What are people for?”. As I construct and amend my views, I'm fully aware that 100% certainty is unattainable, but I'll settle for the best that we can do at this time.

My own highest priority in evaluating these sorts of hypotheses is that the information I include in my consideration be evidence-based, which I determined decades ago was the only reliable hope for distinguishing between various propositions in every discipline.

My tentative favoring of Damasio's brainstem hypotheses in no way implies that I believe the brainstem to be the only possible implementation of consciousness. I'm very open to the possibility of weirdly constructed aliens and I'm pretty sure that octopi are conscious. In fact, I'm open to the possibility of consciousness in any workable substrate, whether that be a silicon chip the size of Chicago or an engineered neuronal structure constructed with artificial DNA producing Artificial Consciousness. Perhaps we'll fancifully invent the “The Animator” and research brown shoe consciousness, as in Philip K. Dick's "The Short Happy Life of the Brown Oxford". In the meantime, our own and other mammalian brainstems are conveniently close at hand for our investigations.

A very recent development from just days ago – “A new research study contradicts the established view that so-called split-brain patients have a split consciousness” (neurosciencenews.com and others) – looks to be very relevant to Callan's earlier post in this thread and also Damasio's brainstem theory. I've been re-reading about split-brain research, hemispherectomies and anaesthesia applied to a single hemisphere at a time. I plan to post my thoughts about this new research and some implications very soon in this thread – the topic seems especially relevant.

Stephen Wysong said...

I believe Bjorn Merker makes the case in his paper, “Consciousness without a cerebral cortex: a challenge for neuroscience and medicine”. From the Abstract:

“A broad range of evidence regarding the functional organization of the vertebrate brain - spanning from comparative neurology to experimental psychology and neurophysiology to clinical data - is reviewed for its bearing on conceptions of the neural organization of consciousness. ... the principal macrosystems of the vertebrate brain can be seen to form a centralized functional design in which an upper brain stem system organized for conscious function performs a penultimate step in action control. This upper brain stem system retained a key role throughout the evolutionary process by which an expanding forebrain - culminating in the cerebral cortex of mammals - came to serve as a medium for the elaboration of conscious contents. This highly conserved upper brainstem system, which extends from the roof of the midbrain to the basal diencephalon, integrates the massively parallel and distributed information capacity of the cerebral hemispheres into the limited-capacity, sequential mode of operation required for coherent behavior. It maintains special connective relations with cortical territories implicated in attentional and conscious functions, but is not rendered nonfunctional in the absence of cortical input. This helps explain the purposive, goal-directed behavior exhibited by mammals after experimental decortication, as well as the evidence that children born without a cortex are conscious. Taken together these circumstances suggest that brainstem mechanisms are integral to the constitution of the conscious state, and that an adequate account of neural mechanisms of conscious function cannot be confined to the thalamocortical complex alone.”

Are children born without a cortex conscious?

Eric Schwitzgebel said...

Thanks for that reference, Stephen! My student Linus Huang is actually working on a theory according to which the basal ganglia are the key structure for implementing consciousness. Not quite brainstem, but close. It is quite possible that the cortex is overrated with respect to consciousness.

Stephen Wysong said...

Eric, to wrap up my response to your question about why I believe consciousness is created by the brainstem, here's a quick summary of my investigation into evidence that I believe poses serious challenges to cortical consciousness theories.

1. Split-brain studies

I've looked into the recent split-brain study results, beginning with the article at http://neurosciencenews.com/split-brain-consciousness-6011/ This research deals with the results of a corpus callosotomy that severs the corpus callosum – “a bundle of neural fibres connecting the left and right cerebral hemispheres, is severed to prevent the spread of epileptic activity between the two brain halves”. While mostly successful in relieving epilepsy, the procedure also virtually eliminates all communication between the cerebral hemispheres, thereby resulting in a ‘split brain’.”

Historically, “In their canonical work, Sperry and Gazzaniga discovered that split-brain patients can only respond to stimuli in the right visual field with their right hand and vice versa. This was taken as evidence that severing the corpus callosum causes each hemisphere to gain its own consciousness.” Both Susan's article and an earlier post by Callan refer to this dual-consciousness hypothesis which is commonly used to support theories of cortical consciousness.

In this new study ... “Our findings, however, reveal that although the two hemispheres are completely insulated from each other, the brain as a whole is still able to produce only one conscious agent. This directly contradicts current orthodoxy ...” And, “According to Pinto, the results present clear evidence for unity of consciousness in split-brain patients.”
Given the widespread citing of Sperry and Gazzaniga's original split-brain hypothesis and its support for cortical consciousness hypotheses, I expect a considerable revision in consciousness theorizing should these new results hold up to scrutiny.

2. Hemispherectomy and hemisphere isolation

I've also investigated the research for the related conditions of hemispherectomy and anesthesia of one hemisphere – the Wada test, an investigative process that anesthetizes one hemisphere at a time to determine the language and memory capacities of each hemisphere. In all of these cases, consciousness itself is observed to be unified, whole and undisturbed, although the content of consciousness is obviously altered by the absence of a functioning hemisphere.

[Continued in next post]

Stephen Wysong said...

Continued ...

3. Cortical surgery for epilepsy seizure relief

Merker relates that Penfield and Jasper (1954!) routinely removed sizeable sectors of cortex in conscious patients for the control of intractable epilepsy. By performing the surgery under local anesthesia only, the authors ensured that their patients remained conscious, cooperative, and capable of self-report throughout the operation. This allowed the neurosurgeons to electrically stimulate the exposed cortex while communicating with the patient, in order to locate functionally critical areas to be spared when removing epileptogenic tissue. They then proceeded to remove cortical tissue while continuing to communicate with the patient. They were impressed by the fact that the removal of sizeable sectors of cortex never interrupted the patient’s continuity of consciousness even while the tissue was being surgically removed.
Penfield and Jasper noted that a cortical removal even as radical as hemispherectomy does not deprive a patient of consciousness, but rather of certain forms of information, discriminative capacities, or abilities.

4. Hydranencephaly – no cortical tissue

Merker additionally discusses consciousness in victims of hydranencephaly because that pathology affords the opportunity to observe consciousness, or its absence, in humans who are equipped with a “reptilian brain,” but have no cortical tissue at all. Damasio has written, “These children, however, are anything but vegetative. On the contrary, they are awake and behaving. To a limited but by no means negligible extent, they can communicate with their caregivers and interact with the world. They are patently minded in a way that patients in vegetative state or akinetic mutism are not. Their misfortune provides a rare window into the sort of mind that can still be engendered when the cerebral cortex is absent.”

Damasio continues: “If a parallel could be drawn at all, once the motor defects are factored out, it would be between hydranencephalic children and newborn infants, in which a mind is clearly at work but where the core self is barely beginning to gather. This is in keeping with the fact that hydranencephalics may be first diagnosed months after birth, when parents note a failure to thrive and scans reveal a catastrophic absence of cortex. The reason behind the vague similarity is not too difficult to fathom: normal infants lack a fully myelinated cerebral cortex, which still awaits development. They already have a functional brain stem but only a partially functional cerebral cortex.” (These remarks imply that we are all “brainstem babies” in our first few months of life!)

My Conclusion

Everything I've read about these conditions is compatible with the hypothesis that consciousness, certainly the core consciousness as defined by Damasio, is generated by sub-cortical activity, most likely in the brainstem, but certainly in sub-cortical structures. In addition, I believe the evidence taken as a whole poses clear challenges to the hypothesis that consciousness is produced by the cerebral cortex.

Merker proposes that cortical structures are “a medium for the elaboration of conscious contents” and, in my opinion, all of the cortical consciousness proposals I've encountered could easily be interpreted as the elaborative processing he theorizes precedes conscious experience. Where are the studies and papers that provide, as with Merker's paper, similar compelling evolutionary, experimental and observational evidence – at the same level of specificity – that support of the various cortical consciousness hypotheses?

Gnz said...

One thing that we use to identify consciousness as opposed to automation would be having persistant irrational biases. I suspect we wouldn't recognise a super AI as objectively conscious unless it was the sort that was likely to be dangerous to us.

Gnz said...

To be fair to the super AI, I don't know how I could counter it if it argued it and it's interests were more important than us like we are more important than chimps. It would seem to be either objectively correct or... an extreme form of communist veganism philosophy holds.