Anders Sandberg, Researcher, science debater, futurist, transhumanist
I would like to greet the members of the Congress, congratulate them on having meeting that allows them to talk about the most wonderful, valuable and precious thing that there is - the future, because by their discussions and their ideas we, of course, going to be able to shape it. We cannot predict the future very well, because our brains are not good at making long-term predictions. However, we do predictions every day. That is actually what our frontal lobes are doing - doing simple predictions about mechanics and action. We make predictions about our lives. And we can tend making predictions & plans for the future, but it is much harder. However, there are also intelligent beings that can refine our methods. We can work together and become better at coming up with plans for the future. That involves envisioning of the future and that involves seeing what could go wrong and seeing what could be very beneficial. And then we can implement that and we can use a magnificently extended recognition that language allows us, the telecommunication allows us, and indeed the kind of collective intellect emerges from conferences that is why I am very happy that this Congress is taking place, and I am really looking forward to see, how it will guide the future.
I think that for the development of humanity over this century there are certain areas of technology, which are going to be very important. One important aspect is automation - the ability to construct systems, machines and software that can act alone, that can do intelligent processes of various kinds. Ideally, we will to have control over these ones, so that means that we need better ways of interfacing humans to these ones, and it might be through smart user interface system gadget, but it might also be through direct connections. This is very important because we want our will, our ideas and values to manifest in the world and we are using technology to do that. So, as technologies become more advanced, then we must make it better at understanding, what we want, and implement it, and that is going to require intelligent technology and stronger links to it.
But in the longer term, of course, we want to do something about our physical state. Our biological condition as humans is certainly not bad: it is nice to be a human, it is nice to be alive, but we could be healthier, we could live longer, and there are various ways of doing that. But in the end biotechnological means of extending human capacity are limited by the limitations of biology and I do think that we are going to enhance ourselves using artificial means. At first, there may be implants, but in the longer run by transferring our consciousness to machines, if that is technically possible, and I do think this is a very important developments that are going to affect our economy, our society and our view of the world over this century.
One of the most interesting possibilities, in my opinion, mainly because this is my old research area, is to simulate & emulate the human brain. So, right now we have competition in neuroscience, we set up simulations of the connections between neurons and the impulses moving between them, and then we are trying to adjust this models to fit the real biological data, both from how the brain is constructed, but also the results. So, right now we have fairly good memory models, we have an idea, we think, of how memories are stored in the brain. We have a much less understanding of high level conditions in the brain, like emotions and consciousness. We do not really know what experiments to test for it. Memory is very easy: we can see it from simulation of remember something, but it does not. But people are increasingly investigating in these more advanced states of the mind, and, I think, we are going to get enough data to stop constraints in the simulations. However, there is another way of emulating the brain, and that is actually to try to get all the little pieces right without understanding, why they are connected like they are. So, this approach, which I call “whole brain emulation”, would consist of taking a brain and then scanning it carefully, finding what is connected to what, all the chemicals and all the interactions going on a lower level, without any understanding of what consciousness or intelligence is, and just simulating that. Now according to some theories of mine that would be enough to actually produce real consciousness, real memory and real intelligence. Of course, some philosophers and scientists disagree about this, but this is a testable proposition, with enough technology we are going to be able to do it.
In order to do a full brain emulation we would need, first, scanning technology that can detect every synapse in the brain and scan them, that is not available right now, although, we can do this on a very small scales, there already exist various forms of microscopy that can scan small slices of brain tissue, and they can probably be scaled up. This looks like a fairly normal engineering, not too different from what we see in bioinformatics, where people have been scaling up from gene sequencing from enormously small sequences to today relatively cheap genome sequence of whole people. The second part is understanding this scans, turning, that, you know automatic fashion into a neuron network simulation that is probably the most challenging research area, but again we have progressing in computer vision and computations of new informatics that seem to be going there. Finally, we need enough computing power to do it. To run a human brain - at this computation and resolution would require an enormously powerful supercomputer. However, more slowing progress is possible in this regard. We are probably going to be there within 20 years. And it seems like, if we do not need all the biological data, if it turns out that we can simplify brain emulation, then we might already have enough computing power. The real problem is, of course, putting together this enormous research project. We have a lot of different research groups that need to talk to each other and realize that they are working on the same goal.
I know that Dalai-Lama apparently warning that it might be a possibility of reincarnating into computer, so there is absolutely no contradiction between many Eastern beliefs and the idea of transferring minds into machines or nanotechnological construct. I do not think that from a religious prospect it is very problematic. Coming myself from a very much empirical prospect, I think, well, the best what we may finding out, whether dualism is true, that there is a soul material that differs from the normal material, would be to try brain emulation, for example. If we scan the material parts of the brain and then try to emulate it into computer, if it absolutely does not work at all, then that might be an interesting evidence that there is something we missed. If it works very well, then either we did a good way in transferring soul material, or may be soul material does not exist. So, I think we are going to get interesting answers actually from experimental neuroscience and transhumanism in the near future, in the next few decades to many of these ancient questions.
I would certainly want to transfer my mind to an artificial body, if the technology existed that had demonstrated that it was safe enough. Some people have a problem with this because the idea of what makes them themselves is of a particular philosophical type. And this is something that is going to be different between different people. I regard myself as the equivalence class of all endless-like processes, which is a very abstract and slightly unusual prospective, so I have no problem with the idea of there existing copies of me, backup copies of me, variants of myself. Other people might disagree of that, but I certainly think that practical benefits of being able to live for ever, if I transmit myself digitally, I will be able to run on bodies which are not biological or enhanced biological and be able to backup copies in case, if something goes wrong, would be enormous. So, I think, that in the future I am hoping to be software.