Brain Organoids-It Begins with Playing Pong
And ENDS...?
Humans have domesticated organisms from every living kingdom for our purposes. Yeasts are used for fermentation, viruses for gene therapies, bacteria to produce antibiotics, plants and animals for all manner of food and fabric. We have even put human cells to work as bioreactors. But in none of these examples have we sought to exploit the cognitive capabilities of the biological entity until now.
Organoid Intelligence (OI) is a flourishing field of investigation. Skin or other cells from human donors are reprogrammed to a pluripotent stem cell state and coaxed down a developmental pathway that yields nerve and other brain cells. The result is a self-organized structure known as a brain organoid that is complete with functioning neurons, active synapses and glial cells.
These brain organoid structures have spontaneous electrical activity that resembles fetal brain activity, and some brain cultures can be trained to play games. The brain culture’s proficiency with playing the arcade game Pong led to claims of sentience, which was met with a robust rebuttal that was dismissed as an argument about semantics
Graphic from Open Access Journal Neuron
It would be natural to wonder what else might be going on in that culture dish-other than ambitions to win an arcade game? Might the thoughtful not wonder at cognitive processes in this odd playmate, just as we do with more intact companion organisms such as a dog or cat. No wonder then, the public reception -- if the public is aware at all--is sharply divided.
To keep this in perspective, polymer hydrogels with no signs of life “exhibit emergent memory” and can improve their Pong playing over time. Furthermore, the current state of brain organoid technology is limited. These organoids are small, on the order of a few millimeters in size, containing a few million cells in comparison to the hundreds of billions of cells in the intact brain, and they are imperfect in their recapitulation of brain anatomy. They are currently limited in growth because they rely on passive diffusion of nutrients and waste, leaving the central portions of the tissue prone to necrosis. However, this seems a temporary limitation since strategies for vascularizing the tissue and thereby eliminating this growth restriction are currently being developed.
Nevertheless, developers have expressed excitement over the potential use of human brain cells when embodied in an artificial world for rapid, energy-efficient computation. Of course, this takes the computer metaphor of the brain to a ridiculous extreme. Moreover, the offensive subtext that the human brain is nothing more than a greasy supercomputer is obvious. But what is clear is the intention to advance OI technology to enable the full exploitation of the computational capabilities of living human tissue emulating human brain activity.
To be clear, organoid research has extraordinary potential. Liver, kidney, gut – name your favorite viscera - all have the potential to serve as disease models that could speed drug testing while limiting the number of experimental animals sacrificed to the cause. But when we ask these cultures to be intelligent, essentially to think, moreover, to exploit uniquely human intelligence (otherwise why not use rodent or primate cells instead of human?), we have entered an entirely new realm. And this is very much the objective of many who study these brain organoids.
It would be hard to find subjects that have elicited more rancorous comments about irresponsible science writing than this. A robust review of the ethical concerns regarding brain organoids cautions against “sensationalism.” One commentary cites the “misleading implication that [organoids] might be experiencing the horror of being conscious while being completely cut off from the external world” and then places blame for “the ethical [problem on a lack of ] responsible and effective science communication.” Admonishments to “responsible communication” seem to appear in the literature rather more frequently than calls for responsible scientific conduct. A review of the media reception concluded there was “increasing polarization, coupled with misplaced and unfounded ethical concern,” although it’s unclear whether the authors were justified in characterizing the ethical concerns as “unfounded.” Examples of reports expressing ethical concern seem all the more poignant for their rarity.
The exhortations for good communication all point to the limited state of current technology, and therefore, deliberately or otherwise completely miss the point; the objectives of “Organoid Intelligence” must reasonably be the expression and exploitation of the products of intelligence by organoids.
Arguing that we should not be concerned about ethical issues because the current state of affairs has not reached its’ full potential is a facile dodge. What are we to do if every argument poo-pooing ethical concerns about OI starts with a recitation of the current limitations of the technology, dispatches ethical arguments about OI by asserting that it is impossible for organoids to be conscious but concludes with the word “yet”?
If organoids are to be intelligent or at least calculate as we want them to, what could possibly be the umwelt of these clumps of tissue, floating in their culture flasks? They lack sensory input from conventional sense organs such as eyes, or ears (although it’s been suggested that this problem could be conveniently addressed by co-culturing eye and brain organoids; see mention of assembloids below). Perhaps this is an unfair question since an umwelt is the manner in which an organism experiences the world and these cultures of tissue have no organs of sense, no memory, no familiarity with embodiment, have no experience, at least until we interact with them. However, this assumes that the absence of experience is not an experience itself. But is this correct? We know, for example, that silence is perceived as if it were sound. To quote the research about that absence of experience, “silence is truly perceived.”
Some have stated that it would be irresponsible for us to suggest that organoids exist as a “locked-in consciousness.” Locked-in syndrome (LIS), is a disorder which results from brain-stem injuries where patients are unable to move, speak or otherwise communicate. Such patients may be conscious and have full cognitive capacity but resemble a person in a comatose state. If organoids were to achieve some level of consciousness or even sentience their experience might have some similarities to LIS. However, this reflects only part of the dilemma that would be faced by such organoids, since unlike organoids those in locked-in states are embodied and retain their senses such as sight and hearing, a connection to an outside world not afforded to an organoid. So, irresponsible or not, the comparison is not complete.
A more terrifying approximation of a disembodied worldview was imagined by Dalton Trumbo in his novel, Johnny Got His Gun. The novel has been described as “a claustrophobic and nightmarish journey into the mind of a young man trapped in himself….enduring his existence as a nearly disembodied mind.” In it, the soldier Joe awakens in a hospital after being caught in an artillery explosion to discover he has no face, no eyes, no ears, no nose or tongue, and so has lost sight, hearing and the ability to speak. Moreover, he has lost his hands and arms stealing from him the exquisitely sensitive touch of the fingers and the ability to gesture. Finally, he lost his arms and legs making it impossible or move. But while this may be more limiting than even the locked-in syndrome, he can still register the touch of others and found the means to communicate by banging his head against the wall to send out messages in Morse code. Moreover, he has memories of a previous sensory existence, adding grief to the horror of isolation.
These are horrifying scenarios, and perhaps it was irresponsible for me to cite them, but I do so precisely because I believe they miss the mark. I believe the best approximation of the organoid existence is, in contrast, quite serene. This was the description of an individual brought to us by a medieval scholar. Just like brain organoids, this individual lived in a state completely devoid of sensory input and lacked any means of expression. Lacking any of the senses, neither the organoids nor this individual had had any experience or memory.
The individual I am referring to is The Floating Man, a thought experiment conjured up by the Islamic philosopher Avicenna (d. 1037 CE). Avicenna imagined a fully formed adult created by God floating in a void with nothing to hear, see, or smell. He floated with his arms extended away from his body such that he was touching nothing, not even his own body. One imagines the relaxing experience of a sensory isolation tank.
Being newly created, the Floating Man had no experience or memory. Indeed, the lack of any sensations left him unaware of the existence of his own body. Nevertheless, Avicenna asserted that the Floating Man would be aware of himself.
Brain organoids may be the closest we can come to a physical incarnation of Avicenna’s thought experiment. While for Avicenna, the Floating Man’s self-awareness in the absence of any knowledge of his own body proved that the self was incorporeal, a ghost in the machine, brain organoids might expose those very corporeal mechanisms of autonomous organization that underly experiences including that of self. According to a recent discourse by neuroscientist Kenneth Kosik on the subject of brain organoids and consciousness:
“In the case of the brain organoid, not even a historical memory of a body exists, [a]nd yet, an organoid’s neural activity resembles the patterns we associate with encoding experience …This pre-configured state, present in the absence of any experience, suggests a framework prepared to encode experience when it arises. We can then pose the question of whether a framework capable of encoding experience but devoid of experience is conscious. Can consciousness exist without content?”
In other words, the organoids are primed for embodiment but exist in an experiential vacuum. Their lack of experience may be an experience, as evinced by that neural activity resembling the encoding of experience. However, their true promise for organoid computing cannot be realized unless they are embodied in a simulated world. That embodiment could result in “phenomena of ethical relevance” to use one chillingly clinical phrase.
As should be expected, the technological limitations on organoid growth are being overcome. Human brain organoids can be transplanted to mouse brains where they become vascularized functionally integrated and responsive to sensory stimuli. By becoming vascularized, they overcome one of the limitations on their growth. Of course, the resulting mouse-human chimera presents yet another host of ethical quandaries. Similar to the integration of the human brain organoid into the mouse brain, different types of organoids, such as organoids of brain cortex and of sub-cortical regions can be cultured together and functionally integrate with one another into assembloids, assemblages of different organoids.
Another wave of the future is the embodiment of brain organoids in the internet. Methods for Internet-connected brain organoids have been reported as a STEM education science project. This work evokes Miguel Nicolelis’ Beyond Boundaries: The New Neuroscience of Connecting Brains with Machines. Nicolelis pioneered the notion of creating a “super-brain” by demonstrating that monkeys and mice could coordinate the activities of several animals through brain -machine interfaces to comprise a “computing device with multiple interconnected brains.” Those efforts have continued with the more recent report of a human “BrainNet” for the integration of brain activity among groups of people.
These efforts drive towards a future envisioned by Nicolelis where people merge their brain activity. The internet embodiment of brain organoids is a modest extension of these efforts. Such an interconnection of brains would be something that would convey incredibly intimate knowledge of those within that integrated network, but also result in a loss of privacy and perhaps autonomy. The consequences of such a shared psyche were explored by Jorge Luis Borges in his story Shakespeare’s Memory, wherein a Shakespeare scholar mysteriously comes into possession of the memories of the Bard. At first, overjoyed at the prospect of such an intimate knowledge of his favorite subject he quickly becomes terrified of becoming lost in the mindscape of another. Nevertheless, some view networked brains as the next level of human social interaction.
Neuroscientist Keneth Kosik stated that “being armed with knowledge of the underlying limitations of organoids is a responsibility of scientists.” One wonders, as those limitations fall away with ever-improving technology, what will be left for the scientists to arm themselves with, especially since they are the ones chopping away at the limitations. Each improvement in technology, each mix-and-match variation of assembloids brings with it greater risk of encountering a phenomenon “of ethical relevance.”
What then?