Personal robots in homes and their integration in everyday life will be a major breakthrough of the 21st century. Yet, to realize this vision, important obstacles need to be overcome: these robots will have to act in unpredictable environments including homes and learn new skills while interacting with humans. Intelligent artifacts and robots are expected to operate in complex physical and social environments. The deployment of service and companion robots, however, requires that humans and robots can understand each other and can communicate.The goal of the Eu project WYSIWYD  is to be able to contribute to a qualitative change in human-robot interaction and cooperation (HRI)  and scientists are working towards advancing a robot’s ability to engage in communicative behaviours with humans.

By allowing robots to both understand their own actions and those of humans, will unable the interpretation and communication of the robot “understanding” into human compatible intentional terms. This is expressed as a language-like communication channel called “WYSIWYD Robotese” (WR). The WYSIWYD project will advance this critical communication channel following a biologically and psychologically grounded developmental perspective allowing the robot to acquire, retain and express WR dependent on its individual interaction history or “narrative”.

An integrated architecture to improve communication in HRI

To achieve transparency and communication in HRI a number of elements must be put in place: a well defined experimental paradigm, an integrated architecture for perception, cognition, action and intrinsic motivation that, among other things, provides the backbone for the acquisition of an autonomous communication structure, the WR-DAC architecture

WYSIWYD aims to contribute to a qualitative change in human-robot interaction (HRI) and cooperation, unlocking new capabilities and application areas together with enhanced safety, robustness and monitoring.

Project reviewed with excellence !

The Eu project WYSIWYD, coordinated by ICREA Prof. Paul Verschure director of the SPECS lab at UPF, has reached its 2nd year with a very positive report by the Eu project reviewers and has passed its 2nd review with excellent!

The yearly review meeting took place on the 19th of March and was hosted by the INSERM group in Lyon.  for more information on the project and more recent video see http://wysiwyd.upf.edu/

The 5th International Conference on Biomimetic and Biohybrid Systems will be held this year in beautiful Edinburgh, Scotland,18 -22 July. The three-day event, organised by the Convergent Science Network, will be hosted at a fantastic venue consistent with the spirit of the conference, the Dynamic Earth: a 5 stars visitor experience with incredible interactive technology to learn about natural events and much more….

The conference will offer amazing talks on a variety of topics related to the fields of biomimetics and bioybrid systems and technologies at the intersection of living and artificial systems. The program includes 5 plenary lectures from excellent experts in the field. The plenary lectures will be complemented by short talks on diverse topics such as robotics, active sensing, navigation, locomotion and others.

You can find out more about the plenary speakers HERE, the full conference programme will be published shortly!

The Living Machines conference will be preceded by a one-day satellite event, hosted by the University of Edinburgh Department of Informatics, and consisting of a series of research-oriented workshops. You can submit your workshops HERE.

We are looking forward to seeing you this year in Edinburgh!

Soon after liberation, camp survivors await their ration of potato soup. Bergen-Belsen, Germany, April 28, 1945.
— US Holocaust Memorial Museum

“Wir wissen nur dass wenn wir hier rauskommen, das wir alles dass wir hier erlebt haben in die Welt hinaus schreien müssen, anders kann man nicht leben”

“We only know that when we get out of here, we must shout out into the world about everything that we have experienced here. Otherwise one cannot live.”

These are the words of Charlotte Grunow recorded on April 20, 1945 by BBC reporter Patrick Gordon Walker.

Charlotte Grunow was arrested in Berlin in April 1943, transported to Auschwitz-Birkenau and moved with a large group of female prisoners in November 1944 to Bergen-Belsen. She was liberated on April 15, 1945 with about 55000 other prisoners, 10000 of which were dead and a further 15000 would die after the liberation from disease and starvation due to a deliberate SS policy of neglect. The gruesome reality the Charlotte Grunow of April 1945 wants us to know about was largely unknown and unimagined by the liberating countries and wasdescribed by the BBC reporter Richard Dimbleby as “the world of a nightmare”. This reality at the collapse of the Third Reich, could be found in over 42000 identifiedcollection, concentration and killing centers and sites, transports and death marchesacross the Europe created by the Nazis. But is Charlotte being heard, then after the liberation, now seventy years later and in the future?

We are facing a transition in the commemoration of the Holocaust. The authentic voices reporting on the horrendous crimes humans are capable of will soon fall silent. Just for Bergen Belsen, key witnesses such as ex-prisoners Gyorgy Denes and Arieh Koretz and liberators Maj. Leonard Berney and Captain Eric “Winkle” Brown have all died in the last year. How to deliver on the solemn pledge we have repeated for the last 70 years, that “we must never forget”? Have we succeeded to transform these testimonies into understanding, meaning or a society the victims hoped for? The answer unfortunately is “No”. For instance, although few systematic surveys exist a recent UK survey among 8000 high-school students showed that the majority only has a cursory understanding of the Holocaust. The same holds for the rest of Europe. Hence, at the end of the period of the witness, we face a memory crisis in terms of the conservation and presentation of the events and experiences at the heart of European history and identity.

We have developed a novel approach towards answering the memory crisis: the Future Memory project. At the start of Future Memory stands a personal experience when I visited the Bergen Belsen campsite where my grandfather Jan Verschure, a Dutch resistance fighter died: I found an empty landscape. The chirping birds provided a score to this peaceful and well-kept heath park that had integrated the elevated tops of the known mass-graves. Note that in 1945 birds avoided the place and the mass graves containing the remains of about 20000 victims are still not localized. However, behind this pastoral façade with no intrinsic footholds to assist in understanding and commemorating resides the ultimate “witness”: space itself. Future Memory aims at reclaiming this space in the service of the preservation of history and the shaping of collective memory now and in the future. Future Memorydigitally enhances space so that it becomes a medium though which historical sources and narratives can be discovered. The Future Memory project has started in 2010, in collaboration with the Bergen Belsen memorial site and was partially supported through the FET project CEEDS.

Future Memory builds historical learning on a twofold use of physical space. First, it acknowledges the fundamental role that space and action play in the formation of memory and experience. A scientific discovery worthy of a 2014 Nobel price. We have build on this link in our exhibitions and performances and the advanced neurorehabilitation technologies we have developed and deployed. Secondly, physical space is a permanent source for the authentication of historical knowledge: “this happened here”. Through the right use of technologies, spaces can be physically and virtually explored and discovered now and in the future, because they are laden with historical sources and reflections of the experiences of those who have been there. We have installed a number of integrated systems at the memorial site Bergen Belsen under the name “Here: Space of Memory” that implement these considerations. At the heart of this approach stands a 3D reconstruction of the former camp together with a database with geo-localized source material including diary fragments, images, drawings, video and audio clips. Visitors can access this physical/virtual space through an immersive virtual reality environment or by walking on the terrain itself using an augmented reality tablet App. By wandering among the reconstructed buildings, visitors explore historical sources in situ. Lastly, we have installed a sound installation that presents visitors with voices, including that of Charlotte Grunow, as they walk from the museum to the former campsite, creating a personal encounter with the fleeting past. The effectiveness of Future Memory can is evidenced through the associated educational program that is intensely used by visiting school classes and booked out for many months to come. After this important validation of the Future Memory approach, our goal is to digitally reconstruct, enhance and link together at least 100 sites across Europe, to show the system level organization of the murder machine created by the Nazi’s. We have started the Future Memory Foundation with the purpose to realize a neutral ground from which we can support this objective through both private and public support.

How close are we to our target? The UK holocaust education survey makes it painfully clear that at best there is a modest impact and we have to ask why the approaches followed over the last 70 years such as professionalizing commemoration, archiving and researching of historical sources, monumentalizing historical sites and offering museums has not translated into more societal impact? Possibly we still have not identified en effective way to link historical information to understanding. The Future Memory project builds a bridge between history, experience and meaning by advancing an integrated approach comprising science, technology, humanities and the arts, that not only investigates and presents “what happened here” but also how we can narrate this central chapter of European history to its citizens now and in the future as a source for continuous learning and reflection. This is a new and complementary approach to existing ones that can assist us in overcoming the memory crisis.

Today as the age of the witness is coming to a close, an enormous amount of work still needs to be done. We only have a few years left to conserve the living memoryof the sites of the Holocaust, while for some sites it is already too late. It is true that Europe has supported some important initiatives such as the EHRI network andEuropeana. But what has been done so far has not been enough, as the current state of Europe’s response to global humanitarian crises and rising anti-Semitism shows. There is a belief that enough is being done, but this is not supported by fact. Also in our case, despite the great interest that our project inspires, including at the level of the European Commission and their staff members, their requests for information have not translated into action, rather into “I have no time”. However, urgent action is required and a large-scale no holds barred European initiative must be undertaken, circumventing old habits and inertia in order to salvage the past to help us shape our European future.

Future Memory answers and propagates Charlotte’s Grunow’s rallying cry. We, the descendants of the victims, perpetrators, traitors, bystanders, survivors and resisters have an obligation to conserve the memory we risk to loose through the mortality of the survivor. To honor the victims, to safe guard and elaborate our European identity and to reflect on the darkest crevasses of the human soul, so that we may transcend them and find meaning and virtue in a deep understanding of who we have been, are and can become.

Get in touch with me @PaulVerschure

This blog was originally published in DAE blog on 21/03/2016

SO, the British Science Association has released a survey on the British public’s attitudes toward robotics and AI. Their headlines:

• 60% of people think that the use of robots or programmes equipped with artificial intelligence (AI) will lead to fewer jobs within ten years
• 36% of the public believe that the development of AI poses a threat to the long term survival of humanity.

Some other highlights:

• 46% oppose the idea of robots or AI being programmed with a personality

We would not trust robots to do some jobs…

• 53% would not trust robots to perform surgery
• 49% would not trust robots to drive public buses
• 62% would not trust trust robots to fly commercial aircraft

but would trust them to do others:

• 49% want robots to perform domestic tasks for the elderly or the disabled
• 48% want robots to fly unmanned search and rescue missions
• 45% want robots to fly unmanned military aircraft
• 70% want robots to monitor crops

There are also results showing some predictable divisions along the lines of gender (only 17% of women are optimistic about the development of robots, whereas 28% of men are) and age (of 18-24 year olds, 55% could see robots as domestic servants in their household, 28% could see having a robot as a co-worker, and 10% could even imagine a robot being a friend).

A reply has come from the UK-RAS Network (the ESPRC-funded organisation representing academic bodies working in robotics and autonomous systems) that explains while there is need to examine these issues and carefully plan our future, there’s really nothing to worry about. They cite a European Commission report that shows there is no evidence for automisation having a negative (or a positive) impact on levels of human employment, and point to genuine benefits of robots in the workplace, suggesting how robots ‘can help protect jobs by preventing manufacturing moving from the UK to other countries, and by creating new skilled jobs related to building and servicing these systems.’

The popular press also seems to have seized upon the issue of robots and AI replacing human labour – though a lot of this in recent weeks has been in response to other studies and speeches. The Daily Mail, however, can always be relied upon to strike fear into the heart of its readers, and they haven’t disappointed. Though their rather restrained headline on the BSA study seems innocent, ‘Do you fear AI taking over? A third of people believe computers will pose a threat to humanity and more fear they’ll steal jobs‘, the article (again) resuscitates Stephen Hawking’s and Elon Musk’s dire warnings about the future threat posed by AI. In case this wasn’t sufficiently terrifying – and it really isn’t – The Mail slaps up another one of THOSE TERMINATOR PICTURES to accompany the article (right), with the helpful caption that ‘There are mounting fears among the public about the threat posed by artificial intelligence.’ Well, honestly, I’m sure no one can imagine why.

(Sigh.) Some needs to sit down with The Daily Mail’s photo editor and have a nice, long, very slow, chat.

But what does this survey tell us? Simply, that there is still a problem with people’s perceptions of robotics and AI that must be addressed, and it seems that we are not even heading in the right direction. A Eurobarometer survey on the public’s attitudes to robotics conducted in late 2014 shows that 64% then had a generally positive view of robots (which, if added to the 36% in the BSA survey that believes robots and AI are a threat to the future of humanity, just about accounts for everyone). In that 2014 study, however, just 36% of respondents thought that a robot could do their job, and only 4% thought that a robot could fully replace them, so clearly this is area of heightened concern. A 2013 Sciencewise survey reported almost exactly the same general results: 67% held a generally positive view (though  this survey reports that 90% would be uncomfortable with the idea of children or elderly parents being cared for by a robot, so compared to the 49% that want robots to help take care of the disabled and elderly in the latest study there might be some progress there… or else people are just so desperate to deal with an increasingly ageing population that they’re perfectly happy to dispense with their elderly relatives by dumping them with psychotic, genocidal toasters.) However, a 2012 Eurobarometer report told us that  as many as 70% of Europeans were generally positive about robots.

These comparisons are very rough and cannot tell us much without more rigorous analyses (and the BSA hasn’t provided a link to the full survey). But it shows that there has been little movement in attitudes towards robotics, and in fact an increase in anxiety that robots will displace more humans in the workforce . Without more specific scrutiny, it’s hard to say what we’ve got here. It could well be the case that what we have is very unremarkable. But though it may be encouraging to see that a majority of Europeans are consistently generally positive in their perception of robots and AI, there is still a sizeable minority that could prove very disruptive to the development of future applications of robotics and AI, whose anxieties cannot – and should not – be ignored.

One way to alleviate a great deal of these concerns, particularly regarding the loss of jobs, is to explicitly undertake to address what is emerging as the vital question in the public imagination: what this increasing automisation means for our societies? Because it is not in any way inevitable that more working robots and AI means more poverty for unemployed humans. We get to choose what the consequences are of this mechanisation; and these decisions will be taken by human beings, not left to the whims of sentient robots, or even the indifference of disembodied market forces. If we decide to divide the advantages of such automisation more equally (for example, with the introduction of a Universal Basic Income), then it could be a very good thing indeed. (It is worth remembering that two thirds (or more) of us don’t like their jobs anyway, so more robots could mean less drudgery and freedom for a disaffected workforce.)

Again, without more scrutiny, it is difficult to judge what these numbers mean. It seems to suggest that the public are very ambivalent about the forthcoming developments in robotics and AI: if 46% oppose the idea of robots or AI being programmed with a personality, then it could mean that around 54% of people could be perfectly fine with emotionally engaged robots. If half of us don’t want robots driving public buses (49%, according to the BSA survey), half might be happy for the them to do so.

We might look at this study and say that we are ambivalent about robots and AI – that means, not ‘indifferent’ (as ambivalent is often, incorrectly, taken to mean now), but that we have mixed feelings. However ,this could be a terrible misreading of the numbers. What if people aren’t deeply ambivalent, but radically schizophrenic? If 50% are reporting that they are worried, the other 50% might not be; they might even be very enthusiastic about the possibilities.

Again, there is no evidence in this study to support this notion, necessarily. There is clearly a need for more research into the specific concerns  – and their sources – in order to properly address these issues, and to understand these anxieties more thoroughly (which will need a very different sort of study). However, the cultural record offers some some unique insights. Because what films, for example, show us is that we are not at all indifferent to robots and AI, or ambivalent. There is no middle ground: when it comes to robots and AI, we are deeply terrified OR wildly optimistic; we seem to be convinced that robots will either spell certain doom for the human race or our last, our greatest, hope for salvation from all of the terrible things that threaten us (including, inevitably, other robots and ourselves).

Let’s look again at the Terminator. (And why not? since so many seem unable to leave it alone we might as well make good use of it.) The first, 1984 Terminator, for many embodies what it is we fear about robots: the relentless, unstoppable, rational monster, the sole purpose of which is to destroy of human life. But already in the next film, Arnold Schwarzenegger is the Good Guy, posing as the only hope to save John Connor and our entire species, and subsequent instalments – including the aptly-named Terminator: Salvation and the latest Terminator: Genisys [sic] – build on this theme. In our cultural imaginations, robots are both to be feared and embraced, or are either genocidal psychopaths or benevolent messiahs.

Such diametrically opposed perceptions – such dread or aspiration – do not facilitate the sort of reasoned, rational debate that will be necessary to properly assess both the challenges and the opportunities that real robots and AI represent, outside the pages and reels of science fiction. And yet we are fed a steady diet of such vicissitudes. In my next post I’ll look at another example, when I finally get around to a full review of the latest Avengers offering, The Age of Ultron.

This post originally appeared in Dreaming Robots.

Researchers discover a new approach and tested approach towards neurorehabilitation through the combination of art and science, stubbornness, perseverance, teamwork and FET.

In May 2002 the exhibition projectAda: Intelligent Space opened its doors at the Swiss national exhibition Expo.02. As project leader of this initiative I had managed a team of about 25 scientists and technicians over a period of about 3 years to turn an idea on an interactive neurally controlled sentient space into a turnkey application. The basic idea was derived from the, so called, RoBoser project where a robot controlled by a system level model of the brain, called Distributed Adaptive Control (DAC), composed music in interaction with its environment. RoBoser was developed in 1998 together with the Brazilian composer and mathematician Jonatas Manzolli.

ADA: the Intelligent Space 2002

Upon the invitation of the Expo organization to submit a proposal the notion seemed obvious: if an interacting robot could compose music, why not have a whole space compose a complex audio-visual composition in interaction with its visitors telling a story on the future confluence of biology, technology and humanity? The idea was a hard sell to the institution where I was working, for a number of reasons some petty and political others more fundamental such as the tendency of scientists to look at phenomena in terms of a fragmented set of atomic elements rather than a connected system further compounded with a deep distrust of presenting results to a general audience which, so the cliché goes, would not understand anyway. Indeed, most colleagues saw me ruin what was to be a scientific career by pursuing a flaky dream. We can see these negative attitudes as an expression of what Ortega y Gasset  has dubbed the barbarism of specialization. Anyway, despite an extremely tortuous organizational trajectory and tribal skirmishes, almost magically and based on a prolonged and dedicated effort of the research team and support from its sponsors, Ada saw the light of day on May 15, 2002 and over a period of 5 months over 500K people visited this unique and unprecedented synthetic organism. The impact of this autonomous sentient space on its visitors gave rise to what now has become the Rehabilitation Gaming System.

Among the many visitors to Ada were also people with different kinds of physical and mental handicaps. Their engagement with Ada was astonishing because it showed that Ada was highly intuitive and inclusive through its strongly implicit audio-visual modes of interaction. It was clear to me that this could be turned into a quality of life enhancing technology but we had our hands full on running Ada’s phenotype, the machinery of the neuronal control system and conducting experiments so this idea had to wait. After the successful termination of the Expo.02 exhibition, where Ada had provided the most bang for the buck (or SFr) as compared to most other exhibits, my goal was to develop a new research activity based on the Ada experience towards quality of life technologies with a focus on stroke and Alzheimer’s disease. A new idea and research line brings with it the usual scramble for resources, expertise, (clinical) partners and subjects. Above all the challenge is to pose the right questions and to follow them wherever they lead you, in order to turn an intuition into an application. Here the DAC theory of mind and brain played a key role. DAC, was first proposed in 1992 and has evolved in a mature theory with many applications. Its synthetic validation methodology goes back to Giambattista Vico whose epistemological principle was that we only understand what we can build. Hence, DAC advanced as a function of the artefacts we have created with it and studied such as Ada, the mobile robots Khepera, ePuck or the humanoid robot iCub.

DAC-Robot-experiment

This illustrates the notion that a scientific theory should not only explain and predict but also provide the ability to control. Hence, by moving to the domain of neurorehabilitation we could test this third criterion even further: can we have an impact on brain repair in the clinic on the basis of the hypotheses on fundamental principles of brain organization as captured in DAC? One would expect that with widely publicized boisterous claims of building a brain by 2020 as long as the money is provided, we should be in a good position to assist people in recovering function after stroke or other neuropathologies. Unfortunately, nothing is further from the truth: we have made very little progress over the last 50 years. In addition, with the ageing population we are pressed for generating effective solutions as opposed to hiding behind big claims. Hence, there is a critical role to be played by new technologies in building a dignified society. However, once one has decided on the technology, decisions on content must be made and here technology and methods are silent. Rather one needs ideas and/or a theory. RGS developed from combining the idea of interactive media, in particular virtual reality, for neurorehabilitation with the DAC theory of mind and brain. This decision was a key step because it made choices on the content of treatment protocols non-arbitrary and every intervention became a well defined interaction with a user from which lessons could be immediately drawn. By now RGS incorporates about 20 specific DAC derived principles that range from the key role of sensori-motor contingencies in organizing cognition and action to the importance of goal-oriented and error-driven intervention.

RGS in use at Vall dHebron Hospital in Barcelona

RGS has advanced over the last 12 years via an extensive experimental agenda realized with dedicated partners in Barcelona. In order to support these experiments we have installed RGS therapy stations in associated hospitals, which are in continuous use. As a result, RGS has build up an unprecedented empirical track record having been tested in a wide range of conditions on over 500 patients at the acute and chronic stages of stroke including in at home settings. Building on these results we are now together with our clinical partners validating the generalization of RGS to other neuropathologies such as Parkinson’s disease, cerebral palsy, traumatic brain injury and spinal cord lesions and the initial analysis looks very encouraging.

Many of the users in our experiments have asked to be able to continue the RGS therapy. This demand combined with the clinical results that show that RGS is more effective than any other intervention available today, has lead to the creation of the spin-off company Eodyne.com together with the University Pompeu Fabra and the Catalan Institute of Advanced Studies. Eodyne’s goal is to make RGS available to as many people as possible for a minimum cost. Starting a spin-off was a very interesting experience for the research team, because it showed that our objectives of large-scale low cost distribution to the world population of 60M stroke victims of a science based product was something only very few of the companies, managers, VCs and consultants we had to deal with could comprehend. Hence, Eodyne has entered the market under its own steam and pursues an organic growth path. This experience, however, raises questions and new challenges for the domain of impact investing and ethical entrepreneurship in the local European context where the business culture is of a very different quality than in the USA. It might be another opportunity for Europe to build its own model of business creation as opposed to following what we believe others do.

RGS has developed as a prototypical example of a FET project: visionary, disruptive, multi-disciplinary, high-risk and science based. It has received indirect support via the FET PRESENCCIA (FP6) and CEEDS (FP7) projects and direct support via the AAL instrument in the, so called, RGS project (FP7). Currently the RGS work on neglect is supported as part of the ERC advanced grant cDAC. It should be mentioned that within the health calls of the EC, RGS never fared well because it was usually reported by the reviewers to be too scientific, the idea of basing technology on basic principles appeared difficult to understand. This, however, is not a reason to change the approach we have taken. RGS is successful because of its inclusive science based and technology oriented multi-disciplinarity that sees societal impact as the objective of science. Indeed, RGS was born in the interaction between science and art that started with RoBoser and Ada and has continued with Re(per)curso, the Brain Orchestra and many more installations and exhibitions build by SPECS. In addition, RGS sets an example for the development and deployment of advanced neurorehabilitation technologies: First, ground it on solid science; second, perform the clinical studies; third, introduce it into society. I emphasize this structure because the norm is rather to spin a nice story to investors without much of a scientific and/or clinical grounding, a reprise of the bubble economy of the “new new thing” model of innovation. This might work on the very short term for those at the receiving end of the money stream but as scientists it is our duty to have specific societal impact not to fool the public and our sponsors with promises of the elixir of life. Rather given the science based approach behind RGS provided by DAC, we can envision a form of deductive medicine where scientific theories find validation in the clinic and the clinic in turn provides key insights in advancing our theories of mind and brain. DAC has through RGS closed this validation loop showing that a virtuous cycle between basic and applied research can be created and above all science can deliver on its promise to contribute to the creation of a dignified and inclusive society. RGS is only the beginning.

this blog was originally published in DAE blog on 17/02/2016

Go game. (Credit: Nature / Google DeepMind)

So the media last week was absolutely full of the latest Sure Sign that the robocalypse is immanent: apparently, Google-backed DeepMind have now managed to create an AI so very sophisticated that it has beat human champions at the ancient Chinese board-game of Go. DeepMind’s AlphaGo has defeated the European champion, which marks another important development in the progress of AI research, trumping IBM DeepBlue’s victory over Gary Kasparov at chess back in 1997: Go is, apparently, a much more difficult game for humans – and, it was thought, for computers – to master, due to its complexity and the need for players to recognise complex patterns.

I expected, when setting off to write a note about this achievement, to find the usual sources in the popular press, with their characteristically subtle declarations, heralding that the End of the Human Race is Nigh!; however, thankfully, responses seem to be more sanguineand muted. The British tabloids have even avoided using that picture of Terminator that almost invariably accompanies their reports on new developments in AI and robotics.

So perhaps this is a sign that things are changing, and that the popular press are becoming more sensible, and more responsible, in their technology reporting. (Lets see how many weeks – or even days – we can go without this sort of thing before claiming victory, or even that we’ve turned a significant corner.)

But there is a lot interesting about DeepMind’s success, from a cultural perspective, even if it hasn’t stirred the usual panic about the robopocalypse. It made me recall a conversation I had atan EURobotics event in Bristol in November. We humans, it seems, like to think that we’re special. And maybe the possibility that robots or AI are a threat to that special status is another reason why we are so afraid of them. Maybe we fear another blow to our narcissism, like when that crazy astronomer Copernicus spoiled things by showing that the earth wasn’t the centre of the Universe, or that Victorian poo-pooer Darwin demonstrated that we merely evolved on this earth and weren’t not placed here at the behest of some Divine Creator. Maybe we don’t really fear that robots and AI will destroy all of humanity – well, maybe we fear that, too – but maybe part of what we fear is that robots and AI will destroy another one of those special places we reserve for ourselves as unique beings amidst creation.

And yet our scientists aren’t going to let us sit wrapped in the warmth of our unique being. They keep pushing ahead and developing more and more sophisticated AI that threatens our… specialness. So how do we, as a culture, respond to such a persistent challenge? Like any good politician, it seems we have decided to confront the inevitability of our failure by constantly changing the rules.

Choose your sporting metaphor: we ‘move the goalposts‘, we ‘raise the bar’.

Once upon a time, it was enough for we humans to think of ourselves as the rational animal, the sole species on earth endowed with the capacity for reason. As evidence for reason as the basis for a unique status for humanity crumbled – thanks both to proof that other animals were capable of sophisticated thought and the lack of proof that humans were, in fact, rational – we tried to shift those goalposts. We then transformed ourselves into the symbolic animal, the sole species on earth endowed with the capacity to manipulate signs and represent.

Then we learned that whales, dolphins and all sorts of animals were communicating with each other all the time, even if we weren’t listening. And that’s before we taught chimps how to use sign language (for which Charleton Heston will never thank us).

And then computers arrived to make things even worse. After some early experiments with hulking machines that struggled to add 2 + 2, computers soon progressed to leave us in their wake. Computers can clearly think more accurately, and faster, than any human being. And they can solve complex mathematical equations, demonstrating that they are pretty adept with symbols.

Ah, BUT…

Humans could find some solace in the comforting thought that computers were good and some things, yes, but they weren’t so smart. Not really. A computer would never beat a human being at chess, for example. Until in May 1997, when chess champion Gary Kasparov lost to IBM’s Deep Blue.  But that was always going to happen. A computer could never, we consoled ourselves, win at a game that required linguistic dexterity. Until 2011, when IBM’s Watson beat Ken Jennings and Brad Rutter at Jeopardy!, the hit US game show. And now, Google’s DeepMind as conquered all, winning the hardest game we can imagine….

So what is interesting about DeepMind’s victory is how human beings have responded – again – to the challenges of our self-conception posed by robots and AI. Because if we were under any illusion that we were special, alone among gods’ creations as a thinking animal, or a symbolising animal, or a playing animal, that status as been usurped by our own progeny, again and again, in that all-too familiar Greek-Frankenstein-Freudian way.

Animal rationabile had to give way to animal symbolicum, who in turn gave way to animal ludens… what’s left now for poor, biologically-limited humanity?

A glimpse of our answer to this latest provocation can be seen in Star Trek: The Next Generation:Lieutenant Commander Data is a self-aware android with cognitive abilities far beyond that of any human being. And yet, despite these tremendous capabilities, Data is always regarded – by himself and all the humans around him – as tragically, inevitably, inferior, as less than human. Despite the lessons in Shakespeare and sermons on human romantic ideals from his mentor, the ship’s captain Jean-Luc Picard, Data is doomed to be forever inferior to humans.

It seems that now AI can think and solve problems as well as humans, we’ve raised the bar again, changing the definition of ‘human’ to preserve our unique, privileged status.

We might now be animal permotionem – the emotional animal – except while that would be fine for distinguishing between us and robots, at least until we upload the elusive ‘consciousness.dat’ file (as in Neill Blomkamp’s recent film, Chappie)  this new moniker won’t help us remain distinct from the rest of the animals, because to be an emotional animal, to be a creature ruled by impulse and feeling, is.. to just be an animal, according to all of our previous definitions. (We’ve sort of painted ourselves into a corner with that one.)

We might find some refuge, then, following Gene Roddenberry’s  example, in the notion of humans as unique animal artis, the animals that create, or engage in artistic work.

(The clever among you will have realised some time ago that I’m no classical scholar and that my attempts to feign Latin fell apart some time ago. Artis  seems to imply something more akin to ‘skill’, which robots could arguably have already achieved; ars simply means ‘technique’ or ‘science’. Neither really captures what I’m trying to get at; suggestions are more than welcome below, please.)

The idea that human beings are defined by a particular creative impulse is not terribly new; attempts to redefine ‘the human’ along these lines have been evident since the latter half of the twentieth century. For example, if we flip back one hundred years ago, we might see Freud defining human beings (civilised human beings, of course, we should clarify) as uniquely able to follow rules. But by the late 1960s, Freud’s descendants, such as British psychoanalyst D. W. Winnicott, are arguing almost the exact opposite – that what makes us human is creativity, the ability to fully participate in our being in an engaged, productive way. (I will doubtless continue this thought in a later post, as psychoanalysis is a theoretical model very close to my heart.)

What’s a poor AI to do? It was once enough for an artificial intelligence to be sufficiently impressive, maybe even deemed ‘human’, if it could prove capable of reason, or symbolic representations, or win at chess, or Jeopardy!, or Go. Now, we expect nothing less than Laurence Olivier, Lord Byron and Jackson Pollack, all in one.

(How far away is AI under this measure? Is this any good? Or this? Maybe this?)

This reminds me of Chris Columbus’s 1999 film Bicentennial Man(based, of course, on a story by Isaac Asimov). Robin Williams’s Andrew Martin begins his… ‘life’, for lack of a better word… as a simple robot, who over the decades becomes more and more like a human – he becomes sentient, he demonstrates artistic skill, he learns to feel genuine emotion, etc.. At each stage, it seems, he hopes that he will be recognised as being at least on par with humans. No, he’s told at first, you’re not sentient. Then, when he’s sentient, he’s told he cannot feel. Then he’s told he cannot love. No achievement, it seems, is enough.
Even once he has achieved just about everything, and become like a human in every respect- or perhaps even ”superhuman’ – he is told that it is too much, that he has to be less than he is. In an almost a complete reversal of the Aristotelian notion of the thinking, superior animal, Andrew is told that he has to make mistakes. He is too perfect. He cannot be homo sapien – he needs to be homo errat – the man that screws up. To err is human, or perhaps in this case, to err defines the human. (Though artificial intelligence will not long be on to this as well, as suggested in another of Asimov’s stories.) It is not until Andrew is on his deathbed and is drawing his very last breaths that the Speaker of the World Congress declares, finally, that the world will recognise Andrew as a human. And perhaps this will be the final line; this is perhaps the one definition of human that will endure and see out every single challenge posed by robots and artificial intelligence, no matter the level of technological progress, and regardless of how far artificial life leaves human beings behind: we will be homo mortuum. The rational animal that can die. If Singularity enthusiasts and doomsayers alike are to be believed, this inevitable self-conception is not long off. Though perhaps humans’ greatest strength – the ability to adapt, and the talent to re-invent ourselves – might mean that there’s some life in the old species yet. Regardless, it will serve us very well to create a conception of both ourselves and of artificial life forms that try to demarcate the boundaries, and decide when these boundaries might be crossed, and what the implications will be for crossing that line.

This post originally appeared in Dreaming Robots.

According to Prof. Paul Verschure and his Distributed Adaptive Control theory of mind and brain DAC, when foraging and hoarding, animals behave according to 5 top-level objectives called: “how”, “why”, “what”, “where” and “when” or the so called H4W problem (Verschure, 2012). This form of complex behavior includes: to learn where and when to look for  resources, what to look for, where and when to return to the home base, how  to avoid obstacles and how to act in order to satisfy internal needs.

But how does the brain organization and underlying neural principles account for these complex behaviors?

This is the question that Verschure and his research team SPECS at UPF tried to answer in their latest paper published in Neural Network.

The SPECS’ team including lead author G. Maffei, D. Pata, E. Marcos, M. Sanchez and PFMJ. Verschure, have looked at fundamental learning paradigms of classical and operant conditioning and how these are realised by core brain systems such as the cerebellum, hippocampus and neocortex. Biologically constrained models of these systems have been elaborated and the authors have managed for the first time to bring these components together in the model called DAC-X gaining new insights in how the interaction between brain systems is coordinated.

The DAC-X model is not only unique because of its biological detail but also because of its ability to control a robot in real-time facilitating an understanding of the realistic dynamics of brains as they engage with the real world.

The DAC-X robot experiments focus on fundamental behavior o foraging and hoarding. The experiments done using mobile robots show that a naïve agent foraging in a new environment needs to acquire multiple kinds of knowledge, from sensory-motor associations to landmarks and goal oriented strategies. In particular, the agent learns over time to rely on local environmental cues to find useful resources and acquires the navigational planning skills that support efficient decision making, leading to an increase of behavioural efficiency, observed in terms of the overall cost-reward ratio.

The DAC-X model shows in detail how concurrently core brain systems contribute to this complex task: the hypothalamus defining the dominant needs, the cerebellum shaping specific action patterns to negotiate the environment, the hippocampus performing internal simulations of potential routes with the prefrontal cortex setting the specific behavioral goals. This work sheds light on the synergies between multiple learning systems in the brain and how these could lead to coherent behavioural outcomes observed in rodents, and other mammals. The objective of DAC-X and the whole DAC series of models is to render whole brain models that can assist us in understanding the brain in health and disease giving rise to novel control methods for robots and better diagnostics and interventions in the clinic.

The SPECS group has already made the first steps in this direction by the DAC grounded approach to stroke rehabilitation called the Rehabilitation Gaming System RGS.

see also  http://www.elperiodico.com/es/noticias/sociedad/investigadores-pompeu-fabra-desarrollan-robot-que-comporta-como-rata-4765003

Robotic Lobster by Prof. Josef Ayers at Northeastern University. Photography Jan Witting

Here are some examples of small biomimetic robots, inspired by sea creatures and insects, developed by scientists around the world

The RoboClam

RoboClam MIT

Inspired by the Atlantic razor clam, this small energy efficient robot, developed by Amos Winter at MIT can dig holes into the sand like a razor clam. This was possible since the researchers have understood the principle behind this clam’s ability  — localized fluidization — and were able to give a robotic digging clam similar abilities.  The RoboClam may be useful to monitor a biological situation under water or to bury anchors and terminate underwater mines. “And the study of the robot gives deeper insight into the important mechanics behind burrowing through localized fluidization” says Amos Winter. https://youtu.be/bztw9PUiRss

Row-bot

Row-bot with its mouth open. Hemma Philamore, Univ. Bristol/BRL

Inspired by the water beetle, at the Bristol Robotics Laboratory, a group of scientists have been developing a robot called Row-bot that can swim in remote locations by harvesting energy directly from the water using a microbial fuel cell as an artificial stomach.

“When it is hungry the Row-bot opens its soft robotic mouth and rows forward to fill its microbial fuel cell (MFC) stomach with nutrient-rich dirty water. It then closes its mouth and slowly digests the nutrients”. The Row-bot may be useful for environmental clean-up of contaminants in natural and man-made disasters.

3D-printed soft robotic tentacles

3D-printed robotic tentacle. Cornell University

Using an elastomer and a 3D printing technique, engineers at Cornell University have developed a method to re-create soft actuators. Using their new technique, a digital mask projection stereolithgraphy system, they have produced pairs of actuators that mimic the function of octopus tentacles.

As reported in a paper published in the journal Bioinspiration & Biomimetics, the researchers believe that “this nascent printing process for soft actuators is a promising route to sophisticated, biomimetic systems” https://youtu.be/BZ5W7LyyKL0

The RoboBee

RoboBee. Wyss Institute

This very small flying robot, inspired by the biology of a bee, was initially developed by researchers from the Wyss Institute at Harvard University in 2004. The RoboBee, designed at Robert J Wood’s lab, is a micro-robot, smaller than a fingernail, that flies and hovers like an insect, flapping its transparent wings 120 times per second. The research effort around the RoboBee project is believed to “foster novel methods for designing and building an electronic surrogate nervous system able to deftly sense and adapt to changing environments; and advance work on the construction of small-scale flying mechanical devices”. Scientist anticipate that these devices may have an impact in advancing fields ranging from entomology and developmental biology to amorphous computing and electrical engineering. http://wyss.harvard.edu/viewpage/428/

The Tabbot

Tabbot. by Ingo Rechenberg

The robot Tabbot has the looks of a cartwheeling desert-dwelling spider and it is named after tabacha, which means spider in the local Berber language in northern Africa. According to its developer, engineer Ingo Rechenberg “…such a means of locomotion would be an advantage in a device meant to navigate the rough surface condition on Mars“. Rechenberg, who teaches biomimetics at the Technical University of Berlin, believes that this kind of tumbling robots can be used in agriculture as well as on the ocean floor. https://youtu.be/OHo32JrkDRk For more biomimetic robots see our previous blogs and  Biomimetic robots at Robot SafariEU in London and Biomimetics: Where’s it at?

Large-scale simulations of the brain in silico, sometimes using robotics, can be useful, but they are only meaningful if built upon a solid understanding of brain regions. “We need to know the specific interactions between brain regions and we need know the control signals involved. We need to know how the brain functions as a whole”, comments ICREA Prof, Paul Verschure from UPF Barcelona, with Prof. John Lisman from Brandeis University. This conversation was followed by the idea to organize a workshop, “The Convergent Science of Mind and Brain” to broaden the discussion with experts in the field. The goal of this small workshop was to bring together visionary scientists in the hope of getting a handle on the function of the brain from a system level perspective. How can we improve what we know already about the brain and its function?

In the beautiful and scientifically rich environment of Woods Hole, and with the promise of lots of fresh fish, desserts from “Pie in the Sky”, good wine and good conversation, 12 scientists from different Universities around the globe have gathered  together to discuss in a novel format that combined topic-oriented discussion with specific case studies of computational models.

These case studies were anchored in the expertise of the participants and the main objective of their analysis was to use them to predict fundamental organizational principles of the brain.  In parallel, a computational robot based system level model of the brain was advanced starting from an available system. This new format was  designed to facilitate the transformation of ideas into realized computation, possibly validated by the behaviour of robots.

The event was organized by Paul Verschure (UPF and ICREA), John Lisman (Brandeis Univ.), and Anna Mura (UPF, Barcelona). Sponsored by CSN II [FP7-ICT-601167]

for more information see http://csnetwork.eu/activities/woodshole2015

Just last week, La Pedrera, Barcelona, has hosted the Living Machines 2015, the 4^th International conference on biomimetics and biohybrid systems.

Running from the 28^th – 31^st of July, Living Machines 2015 is sponsored by the Convergent Science Network and feature plenary talks by internationally-renowned researchers in roboticists and workshops examining the intersection of living and artificial systems. There were also poster spotlights and poster sessions, and robot and media demonstrations. A full programme of the events can be found here.

Biomimetic systems are technologies that draw their inspiration from biological systems; these can be used to improve artificial systems and offer solutions to technological and engineering, and can also be used to explore in greater depth natural systems themselves. Biohybridity refers to the merging of living and artificial systems to create new entities, and are used, for example, in robotics, materials, computing, brain-machine interfaces (e.g. neural implants), artificial organs and body parts.

Plenary speakers at this year’s conference include:

• Roger Quinn: Director of the Centre for Biologically Inspired Robotics Research at Case Western Reserve University in Cleveland, Ohio. Professor Quinn will talk on ‘Animals as models for robot mobility and autonomy:  Crawling, walking, running, climbing, and flying’
• Barbara Mazzolai: Director of the Centre for Micro-BioRobotics (CMBR) of the Istituto Italiano di Tecnologia (IIT) of Genoa, Italy, and Deputy Director for Supervision and Organization of IIT Centres Network. Professor Mazzolai will be giving a talk entitled ‘From plants and animals to robots: movement, sensing and control’
• Ryad Benosman: Professor at the University Pierre and Marie Curie, Paris, France, leading the Natural Computation and Neuromorphic Vision Laboratory, Vision Institute, Paris. Professor Benosman will be giving a talk entitled ‘Neuromorphic Event-based time oriented vision: A framework to unify computational and biological vision.’
• Robert Richardson: Director of the Institute of Design, Robotics and Optimisation at the University of Leeds. Professor Richardson will be talking about using robots for safety and security, surgical technologies for health and well-being, and rehabilitation and prosthetics.
• José Halloy: Professor of Physics at Université Paris Diderot. Professor Halloy will be speaking about collective intelligence in natural and artificial systems.

And workshops included discussion on topics such as:

• The robot self
• Nature-inspired manufacturing
• Bio-inspired design

Living Machines is one of the foremost conferences on robotics in the world, and is not to be missed. If you could not attend, the proceedings are already available here – do explore and have a look at some of the terrific ideas and developments being discussed. (Proceedings from previous years’ conferences can be found here.)

For all inquiries contact info.csnetwork@upf.edu