Julian Savulescu
@juliansavulescu
In Roald Dahl’s short story, William and Mary, William dies of cancer. But a novel procedure allows his brain, with one eye attached, to be kept functioning in a clear plastic vat. His wife convinces William’s neurosurgeon to allow her to take William (or rather his brain and eye) home with her.
When home, Mary places William in a prominent place in the sitting room from where he can survey all her actions. He had been a domineering and controlling husband. He forbade her to have a TV and to smoke. Now, Mary purchases a TV and takes up smoking, blowing smoke in the direction of William. She will punish him for his abuse and his brain may stay alive, utterly powerless, for up to 200 years.
This story was science fiction. But yesterday, the first step to creating the brain in a vat was reported in the US. Back in July 2013, scientists reported the first organ grown from stem cells: a liver. A kidney, heart and other organs have followed. The potential of these technologies to eventually provide replacement organs is also an opportunity to sweep away complex ethical issues: most obviously in avoiding the need for organ donation, but also in enhancing the ability to test drugs on lab grown organs before testing in humans- reducing the risk of harm to research participants, hopefully some day to a negligible amount.
Now, just 2 years later, the first brain has been grown in a laboratory. The organoid has been grown for 12 weeks, the equivalent of a 5 week old foetus.
Lead researcher Professor Rene Anand, from Ohio State University in the US,
said:
“It not only looks like the developing brain, its diverse cell types express nearly all genes like a brain.”
The nerve cells within the brain are functional, it has a spinal cord, and the beginnings of an eye
It has 99% of the genes of the foetal brain, and it may be able to develop fully. Anand explained:
“If we let it go to 16 or 20 weeks that might complete it, filling in that 1%
of missing genes. We don’t know yet.”
As with other organoids, there is enormous therapeutic potential in this new technology. The Press Association reported that Anand’s team have already produced models of brains with Alzheimer’s and Parkinson’s Diseases, and even autism in the hope of greater understanding, and eventually, a cure. Research into strokes, brain injuries, and even post traumatic stress disorder is also envisaged. There are enormous potential benefits.
The Guardian reports that the researchers do not consider this to have any ethical implications:
“The ethical concerns were non-existent, said Anand. “We don’t have any sensory stimuli entering the brain. This brain is not thinking in any way.”
The creation of the in vitro brain (IVB) is perhaps the most exciting thing to happen in neurology to date. There are no real treatments for most brain disease. This opens the door to a glimmer of light for people suffering from some of the worst diseases that can afflict a person. We are our minds, and our minds are the result of what goes on in our brains – everything that matters, all our hopes, experiences, abilities, personality, loves and desires are the result of what happens in our brains. For the first time, scientists have been able to create what until now only nature or God could do: make a human brain. With that power comes the possibility of treating the most intractable of diseases, but also augmenting human mental performance, and, perhaps, ultimately replacing the human as we know it.
The creation of an IVB, or in vitro brain, raises five challenges of ascending level of controversy.
The first two challenges are technical, though they do raise ethical issues.
The first challenge is to get this brain tissue to accurately mimic disease human brain tissue, such as the brain tissue from a patient with Alzheimer’s Disease, Parkinson’s disease or autism. This is the ostensible purpose of this research.
If you can get it to behave like diseased brain tissue, you can experiment on it, studying why disease occurs and what modifies it. All this can be done in vitro, or in a petri dish, rather than in a human being or an animal model, involving suffering of the animal.
Once these tissues models are perfected, batteries of promising drugs could be trialled as treatments for brain disease, like Alzheimer’s Disease, Parkinson’s Disease, multiple sclerosis, and so on.
The second, more far off application is regenerative medicine, the holy grail of medicine. When a part of the body is damaged, the ability of the tissue or organ to regenerate is limited. Thus, if you cut yourself, the normal skin is replaced by a scar that holds the skin together, but does not sweat, produce oil or otherwise function as skin. When someone has a heart attack or stroke, the damaged heart or brain is replaced by scar tissue that does not function, leaving the victim with heart failure or, in the case of a severe stroke, paralysed and speechless.
But this stem cell technology could be used to convert a skin cell from the patient into functioning brain tissue that could replace the damaged brain with functioning nerve cells. The challenge will be to get this tissue integrated and working like the original brain.
If that were possible, when a stroke wipes out a quarter of the brain, or it receives a traumatic injury, that damaged portion of the brain could be replaced. For this reason, the US military is showing an acute interest in this technology.
When such technologies transition from the bench to bedside, entering first in human trials, they will raise the usual ethical issues of reasonable risk, protection of participants, consent, exploitation and so on. But there is a vast research ethics apparatus poised to deal with these.
The third challenge is the application for human enhancement purposes, likely to be another topic of interest to the military. Since our abilities are the product of neuronal development and activity, augmenting brain function with IVB is not beyond the realms of possibility, especially if used at the same time as treating a brain disease or injury to render the person “better than well.” Human enhancement raises familiar concerns about authenticity, cheating, fairness, equality and so on which I have discussed at length .
The fourth challenge is that raised by the story of William and Mary. At present, the brain represents that of a five week fetus. At present, because the brain is derived from “ectoderm”, one of the 3 primitive cells types, and the blood vessels from “mesoderm”, the brain has no blood vessels. But it is only a technical obstacle to plumb the brain which will presumably be overcome. This would allow development to much later equivalent gestations. Scientists already suggested that they want to push the development to the equivalent of a 20 week fetus.
At about 20 weeks, the normal fetus becomes conscious. It is thus possible, in principle, that the IVB could have the potential for consciousness at such later stages of development. As Prof Anand has been at pains to point out, it has nearly all the hallmarks of a normal human brain.
The reality is we won’t know its potential until it is assessed. This raises a whole new gamut of challenges – how could you determine whether something has the potential for consciousness at a point in which communication is not possible? Should its potential be measured in a range of stimulatory environments, mimicking that of normal human development?
Of course, abortions are regularly performed up until 20 weeks and destruction or use of the IVB up to that point would raise no new issues over and above abortion. But if it were pushed on in development past 40 weeks, it is possible it would develop higher levels of consciousness or even the potential for self consciousness.
Reminiscent of Dahl’s brain in the vat, the IVB already has “the beginnings of” an eye. Presumably some social stimulation would be required for the development of higher cognitive abilities. But once we move past 20 weeks of equivalent human gestational development, there is potentially a being with moral standing.
At that point, the possibility of suffering and misery becomes real. Once that point is reached, it is important to treat the brain in vat as end in itself, not as a means. It would be necessary to assess its potential, its abilities, its level of consciousness and experience before it could be used for human ends, including treating disease. Such a brain, like William, might have the capacity to suffer and so have rights. We would not know until we assessed it.
Creating beings of uncertain moral status is truly playing God. It requires not that we assume it is not a person because it doesn’t look like us. Rather, it requires treating it with the maximum respect that might be afforded to a being of that potential level of moral status. This is a challenge that we have done very badly at facing in respect of non-human animals.
The last challenge is the most profound. The creation of an IVB raises the possibility of neural computing, or integrating human neurons into computers, a mixture of carbon and silicon computing. Given the huge capacity for development and learning of the human brain, together with the massive computational power of computing, this could usher in a quantum leap in computing which literally is unimaginable.
This could provide enormous cognitive resources for solving the world’s greatest problems. It also raises the possibility of self-conscious artificial computers. Such superintelligence raises the spectre of Terminator – the possibility of malevolent AI wiping out humanity, or Ultimate Harm.
The challenge will be to reap the enormous potential for benefiting human beings and promoting well-being, without creating runaway technology. One way to do this, for a while, is to limit the development of the IVB to 20-24 weeks gestation.
Interestingly, there may be a legislative lacuna for this kind of research. In the UK, since this brain in a vat was not derived from an embryo, but rather a skin cell, it may not be covered by existing legislation that prevents destructive embryo research after 14 days. In this way, iPS technology can bypass the constraints on ES cell technology. That is, it is possible that there are far fewer constraints on it than there are on embryo research. But since what matters is the brain, I do believe we should think carefully about such research. This does not mean jumping to prevent it. I believe we can consistently allow research up until the time normal abortion is performed, that is 24 weeks, but regard any research after that point as requiring new scrutiny.
Given the enormous potential of this research, we ought to find a way to conduct it ethically.
If this technology were to mature to the point where fully conscious persons are not born, but created as brains-in-vats, perhaps with digital connections of sensory input and communication output, it could greatly increase both reproductive rates and maximum population (assuming these brains require fewer resources and live in virtualized worlds).
At this stage, it would be much easier to create a high quality of life for them, since the research would allow enhancements that reduce suffering, and perhaps direct pleasure wireheading.
But the risks on the way to get there and the abuse potential seem far too high except for the most cold-hearted utilitarianism.
I agree about the problem with creating beings of uncertain moral status. I have argued nearly exactly the same in my work on the ethics of brain emulations in software: we better treat them with the same respect as the original biological system, just in case.
The problem here is that IVBs may be fundamentally different from in vivo grown brains: without sensory or body chemistry input they might develop very differently, and the overall patterning may also become alien if grown with artificial inputs. In many way cyborg minds based on neurocomputing are just as problematic as pure de novo AI in terms of making them behave safely; the crux of the AI control problem is that it is very hard to get a non-human system to understand human values well enough not to be risky if it is powerful. However, a IVB also has the same limited speed as a biological brain: the risk of it growing super-powerful surprisingly quickly is likely far below the risk from a purely software system (which can run faster on better hardware, optimize its own software, and perhaps extend itself with extra computational resources). That suggest that IVB artificial intelligence may be more correctable than “pure” artificial intelligence.
Fascinating throughout, but I was startled to read that a fetus achieves consciousness at 20 weeks. That claim runs contrary to much that I have read — unless by consciousness you mean low-level, input-output awareness rather than full-blown sentience. Can you explain and perhaps offer a citation or two?
Comments are closed.