Home / Blog / The evolution of … Machines

The evolution of … Machines

A rough sketch of machine evolution towards awareness. Credit: FDC SAS Initiative

The Symbiotic Autonomous Systems Initiative has completed its first WhitePaper (it will become available through the SAS website by the middle of November once the cleaning up is complete). It is an interesting document and in its concluding remarks it shows the possible, expected, evolution of machines towards awareness over the next decades (the horizon has been set to 2050 but quite a bit is happening today and a lot will be accomplished by the next decade).

 

Clearly it is difficult, may be even unreasonable, to make prediction over such a long span, however it is not about wild guessing, rather it is about looking at what technology offers today, where research efforts are around the world, what the market demands and the social drives that will make the evolution a reality.

IEEE is aware of most of the technological research efforts and this global visibility makes prediction in the area of symbiotic autonomous systems an exercise in rationality.

So, let’s take a look at this sketchy roadmap.

Machines have become smarter and smarter thanks to an ever increasing processing capability, access to large storage for local and remote data, sensors and communications. We have cars that have shown the ability to drive autonomously, although they are still rare and there are regulatory hurdles in the way (not to mention their affordability in terms of cost). The basic technology for self driving cars exists today, it is just not completely practical nor affordable. But it is just a matter of time, no longer of “possibility”.

This self driving cars are “context aware”, that is they “understand” in an operational sense what they need to do given the context around them. They can identify a person walking on the sidewalk and evaluate the probability that he may cross the road all of a sudden, as well as evaluate distance and velocity of an incoming car to evaluate the safety of overtaking the preceding car.

In the next decade this context awareness will become more and more generalised and, most important, affordable. Notice that it is not just cars. Robot vacuum cleaners have already some sort of understanding of their context and this understanding will grow to include something like: “uhm, there is a person watching a tv show so it is better to wait for cleaning not to disturb him, or the lunch is just finished so it may be a good time to vacuum the kitchen…

A significant contribution to the evolution towards context aware machines will come from military applications, as it happened in the past. So it is not difficult to forecast that machines will become context aware, wherever and whenever it makes sense.

We are also noticing, today, that a number of devices are interfacing directly with us, mostly in the medical space, getting information on our status and acting in consequence.  Insulin pumps are becoming smarter providing the exact dose by measuring the glucose directly in the body (smart contact lenses are available in the labs of Google and Samsung, and most likely in other research labs to detect the sugar level in the tears and communicate it to a chip that can take action delivering the required amount of insulin). In the next decade this devices are likely to become proactive, analysing the behaviour, guessing the expected one and injecting insulin as soon as it makes sense without waiting for reaching a thresholds. Bio interfaced machines will allow them to connect to nerve termination, to the metabolic system, to muscles, to our senses and even directly to the brain. Hence an evolution that we can expect is towards augmented machines, augmented through the information provided by a living being, including, of course, ourselves. Again we are seeing the first occurrences, although crude, of augmented machines in robots, like Baxter, that learn by watching people, or in sensors leveraging on living cells to detect specific molecules. Of course tools are “augmented” by people using them but in this case we are not talking about autonomous system. A hammer cannot do anything without a hand (and a brain behind the hand) operating it. A self driving car, on the contrary can operate autonomously but it can also benefit from a standing by driver. In the coming decade the situation where people can “lend” their brain to a machine to augment its intelligence will become more and more common.

In order to become “intelligent” a machine needs to pass a certain thresholds of complexity, similarly to living things. A bacteria is fully operational and in a way smart, but that smartness is the consequence of millions of evolution steps, of generations that finely tuned its response to the environment. To get a local intelligence you need to have much higher complexity. Not all machines will reach this thresholds but there will be some that would aggregate into complex systems and intelligence will result, emerge, out of the whole system. These machine swarms are becoming possible through a connectivity fabric that connects thousands, millions of them, like a anthill makes intelligence emerge out of thousands of ants, individually incapable of showing intelligence.

Both machine swarms and context aware machines will likely take a further step becoming machine aware. In a way complex living things are an example of this evolution. One can see our human body as a cell swarm, hundreds of billions of cells, connected to a context aware machine, the brain, that all together result in a being that is “aware”. Would these machines be “sentient”, in the sense of being aware that they are aware? Opinions differ and no stand has been taken by the SAS White Paper.

About Roberto Saracco

Roberto Saracco fell in love with technology and its implications long time ago. His background is in math and computer science. Until April 2017 he led the EIT Digital Italian Node and then was head of the Industrial Doctoral School of EIT Digital up to September 2018. Previously, up to December 2011 he was the Director of the Telecom Italia Future Centre in Venice, looking at the interplay of technology evolution, economics and society. At the turn of the century he led a World Bank-Infodev project to stimulate entrepreneurship in Latin America. He is a senior member of IEEE where he leads the Industry Advisory Board within the Future Directions Committee and co-chairs the Digital Reality Initiative. He teaches a Master course on Technology Forecasting and Market impact at the University of Trento. He has published over 100 papers in journals and magazines and 14 books.