Conversation with a yoghurt
E&T considers some possible scenarios of life on Earth in 2020 and beyond.
Convergence of biological and information technologies, enabled and empowered by nanotechnology and cognitive science, will create opportunities for future life. Increasing understanding of the biological domain is leading us into the realm of synthetic biology, where human ingenuity is added to the techniques given to us by nature to allow custom-designed life forms. In parallel, advances in cognitive science and artificial intelligence continuously cross-fertilise and will ultimately lead to the creation of intelligent and conscious beings, entirely based on electronic platforms, as well as to intricate and deep links to organic nervous systems, ultimately including our own brains. Thereafter, convergence of this field with the progress in the biological domain is inevitable. We will get hybrid life forms that exist partly in cyberspace and partly in the real world, opening up the possibility for a globally connected, self-managing artificial ecosystem. The far future prospects for humans include being able to communicate telepathically, to have vastly enhanced intelligence, even to live on electronically after our bodies die.
It is already possible to use ink-jet printers to print electronic circuits onto paper or plastic film, and to print onto other firm but non-flat surfaces such as fingernails. In due course, it will be possible to print electronics directly onto areas of the skin surface. And certainly possible to stick them on thin polymer films (with touch-sensitive computer displays printed on them) as well as smart membranes, communications devices, and a variety of health sensors. Using jets of compressed air, it should also become feasible to blast small capsules of electronics at varying depths into the skin surface - for security identification, computer interfacing, health and body adornment. Electronics deep in the skin could monitor blood chemistry and provide an excellent link between IT and the body's own nervous system. As well as monitoring health, sensations could be recorded and replayed by sampling nerve signals at their point of production and replaying them into exactly the same place. This could dramatically enhance computer games, immersive virtual environments and Web-based communications.
The latest processing chips include up to 16 processing cores, produced using electron beam lithography, and the number of cores is set to increase dramatically in future. By 2020, bottom-up assembly is likely to be used in the electronics assembly process, using synthetic proteins doing much of the actual assembly. Instead of hard wiring thousands or even millions of processors together, it will make good sense to suspend them in gel, using optical signals for interconnection. By then, self-organisation technology and evolutionary design will also have progressed, as will our understanding of cognitive processes and the ability to construct analog electronic neurons to tackle many of the tasks where digital processors are found wanting. Powerful analog processing will work in conjunction with digital. Pulling all these technologies together, it should be possible to create a processor gel with enormous digital and analog-neural processing capability, but relatively unstructured, capable of developing much of its own neural design by a learning process similar to that used in our own brains. Such a platform would lend itself well to the development of highly intelligent and conscious entities, complete with advanced sensory and emotional capability. Emotions are a very important component of human consciousness, and many people wrongly assume that computers cannot have emotions. In this processing gel, emotions could be implemented by simply bathing a region of the gel in a light beam, creating a neural bias so that inputs from other neurons will be processed slightly differently. If we reasonably assume that many different colours of light will be used for interconnects, then many various emotions will be possible and could be implemented with very high specificity and precision. Processing gel could present a highly flexible platform allowing one or many AIs to be supported depending on its volume, shape and total processing, sensory and storage capability. AIs could even timeshare in much the same way as digital processes timeshare conventional digital processors, sharing some regions such as sensory and processing capability and having other regions dedicated to themselves, delimiting their own 'minds' from that of their co-inhabitants. AIs could therefore have a range of individuality, either being totally separate and independent, or representing a small part of a collective mind. Human understanding of biology is progressing quickly. The discovery of DNA, the human genome project, the first assembly of a virus from off-the-shelf genes and the imminent announcement of the assembly of a simple bacterium are all ever closer milestones in a timeline that goes all the way to the capability to design and build life forms from scratch.
Long before we understand how to design a particular string of DNA bases to achieve particular organism characteristics, extensive libraries of gene/characteristic will be built simply by experimentation. Though we may be far from designing and building a whole ecosystem to specification, we will be able to build one from a large menu of possibilities much earlier. These would most likely live alongside natural and genetically modified organisms, but as humankind is wiping out natural species at an alarming rate, the ability to redesign, resurrect, upgrade or build replacements will be increasingly useful. Given the rate of parallel development of using synthetic biology in electronic circuit construction, it is possible, or even likely, that one of the first capabilities will be assembling and even supporting electronics within an organism's cells. Indeed, proteins in bacteria are likely to be used to assemble electronics, later being dissolved away to leave the circuits behind. As the field progresses, bacteria that can function perfectly well biologically with the electronic circuits intact (supporting them by providing the energy for their use, given a good food supply) could be designed. A 'smart yoghurt' would result, making it much cheaper to make computing capability, since these bacteria could reproduce naturally, making scalable processing that is bought by the litre. With the right blend of initial architecture, evolutionary design and self organisation, 100ml of smart yogurt could theoretically support roughly the same intelligence as the whole population of Europe. Before we connect a super-smart machine to a human to enhance our own IQ, we obviously need to ensure that we can do so safely. Regardless of how much computer simulations we use, at some point we will have to do try it on laboratory animals. However ethically challenging it might be, it is greatly amusing to think that not too far in the future, we will have to make some hamsters smarter than Einstein! What we do with them after that is anyone's guess, but they would certainly improve the quality of decisions in many boardrooms.
Emigrating to cyberspace
Many people have played games such as 'The Sims' (EA Games), a compelling virtual soap opera where a small community of simple AI people is designed and partially controlled by the user, who effectively acts as their god. There is already a strong similarity between games like this and socialisation sites such as Second Life, where avatars linked to real people socialise in another virtual world. Part of the population in these socialisation sites is also AI based. As AI develops quickly towards consciousness, with intelligence likely to approach human levels by 2020, it is clear that the potential for such games and socialisation sites will increase dramatically. Among the many obvious ethical issues, the potential exists for people, even children, to control entire communities of conscious AIs, with life and death power over them. Because of this and other related reasons, as AIs progress they will have to be given a degree of independence. Machine rights will be debated and eventually implemented. Smart AIs will one day be able to earn money by providing commercially valuable services and to buy robots, so that they can migrate into the real world. They might even do so as a society on a time share basis living mostly in cyberspace and just occasionally wandering into the real world. If AIs get rich enough and if real world regulation permits (and they may well be in a strong position to influence that), we might see a large population of AIs taking on robot form. As always, other technology progresses in parallel and at the same time as biotechnology is routinely used to mend and enhances our bodies, control ageing, and keep us fit and healthy, the simple nervous system links in active skin will eventually develop into brain implants. At first, these will have medical purposes such as controlling or alleviating Alzheimer's disease, but eventually it is certain that people will want upgrades to their memory, sensory capability or intelligence, or personality modification. Over time, as these upgrades inevitably account for an ever-increasing proportion of a person's overall capability, there will be a point where biological death will only result in a small loss of a person's functionality, rather like a stroke. Indeed, by the end of the century, it is likely that death will be thought of merely as an inconvenience, and no longer as a major career problem. It is not necessarily that we will be able to 'download' our minds, just that most of a person's mind will simply be in the add-ons rather than inside the brain itself. As more and more of a person's mind becomes resident in the electronics rather than brain tissue, it becomes increasingly 'normal' to experience and interact via networked capability, not via a keyboard as today, but directly as part of our mind. So we might one day routinely explore far away places without actually going there, simply by experiencing them via remote sensors, which would feel as if they were part of us, just as our eyes and ears are. And it is also possible that for some people almost all of their existence could be cyberspace-based. They could become largely inert in the real world, though very busy in cyberspace. They might even see their minds merge with those of other people or AIs, so that their own individuality eventually evaporates. And then, of course, the possibility that a person can effectively migrate from the real world into cyberspace, may be resurfacing in time-shared robots from time to time.
Word of caution
The technologies described will be extremely powerful and the question arises whether humans are ready for them. Without adequate international governance and policing, it will be possible to use some of these capabilities for weapons. Accidents in some of these fields could cause devastation, and some potential mishaps, or indeed misuses may become extinction-level events. It seems dangerous to achieve such technological capability in advance of the required level of human political and cultural development to make it safe. But this could be a problem that faces all civilisations as they develop. Indeed, if all civilisations wipe themselves out within a few hundred years of discovering radio it would certainly explain why we don't see any extraterrestrials!