Movement Intelligence

So one of the areas Honda were keen to look at when developing the ASIMO is something called 'movement intelligence'. Developing an artificial brain to control the robot's body was too big a task, so they broke down the process of providing the robot with mobility into much smaller areas. The result being algorithms that help it to walk and wave and balance, as well as the myriad other things it is able to do. The sum total of all this is that when the robot appears in public or on TV, viewers are convinced it is intelligent, or even that there is a small person inside making it work.

Most people are happy to have static devices like washing machines and dishwashers in their homes but when a machine comes along, like ASIMO or the Roomba floor cleaner, that moves as well as performs a function, it becomes something else. New Scientist recently reported an intriguing study which found that people who are beginning to form bonds with their robotic cleaning devices - giving them names, even dressing them and assigning them genders - in general bestowing them with qualities we normally associate with living beings. Going even further, the Washington Post has interviewed soldiers using robots in Iraq who reported feelings of 'deep sadness' when their reconnaissance or mine-clearing robots were destroyed. We are familiar with animatronic devices that follow set routines in theme parks, or industrial robots that perform repetitive tasks in factories, but when we get machines with unpredictable movements in uncontrolled locations, maybe that does something to the human brain. There is interesting research being done into mirror neurons that suggests that we mammals are pre-conditioned to react to movement and perhaps, to empathise and respond to it. It seems to be part of how we learn about the world so maybe this is why we respond to mobile robots with such affection.

As robots come into our homes more and more, and as they take on more appealing designs, there are sure to be interesting revelations about how we feel towards them and the types of relationships that might develop. I'm not just talking about an intimate relations with robots scenario here - but other bonds that are more varied and subtle. Robots as essential parts of the family perhaps - our friends electric.

© Richard Evans 2008

Google Books: Naughty

One for the literary crowd today. Recently discovered that those naughty boys and girls at Google Books had published an abridged 'preview' version of Machine Nation (taken from an out-of-print 2002 edition) in their online library. Only trouble was, they neglected to ask permission and the original publishers no longer had the rights to allow other sites to promote it. Oh dear. What with it not yet being 70 years since my death, all of this constituted an infringement of copyright.

Google, in their somewhat messianic, self-appointed mission to publish every book in the world in their online library, do provide a service whereby authors can register and contact them to request removal of a title but this puts the onus on the author to do the legwork once rights have been infringed. It's a little like saying I'm going to steal from you, but if you can be bothered to find and complete this complicated form hidden in an obscure part of my website, then I suppose I won't. Instead of... just asking permission first.

Anyhoo, lawyers were consulted and a polite-but-firm missive was despatched asking them to remove the book immediately from their site. Within three days, the illegally posted work was taken down - though the absence of an apology from Google was duly noted. A few days later an unexpected royalty cheque arrived from old publisher too. Other authors are suing Google for similar copyright transgressions, but thankfully that outcome has been avoided.

© Richard Evans 2008

The Pinocchio Complex

Androids in fiction are often portrayed as wanting to be human, just like Carlo Collodi's Pinocchio, the wooden puppet who dreamed of becoming a real boy. The tagline for Steven Spielberg's A.I. - Artificial Intelligence movie described the android boy David thus: 'his love is real, but he is not' (now there's an existential riddle if ever I saw one). Even the title - Artificial Intelligence - suggests something that is... less. Isn't intelligence just intelligence, whether it's digital or organic? David and other fictional androids like Data in Star Trek: TNG and Asimov's Positronic Man, long to be human in order to gain acceptance or to be 'more real'. In short they seek validation from their human creators or owners. I like all of these characters and their stories, but each is quite tragic, doomed by a neurotic delusion hardwired into their positronic brains.

So why does fiction assume an android would wish to be human, how could it ever be more 'real'? Just because an emotional capability is programmed doesn't make the emotions inauthentic or insincere. Everything that we perceive in the world - everything - comes down to the brain's interpretations of various sensory data - and much of the information we receive about our social environment is subject to constant 'programming' and manipulation - from advertisers, teachers, peers, religious fanatics and politicians. One could ask: how real are we?

Professor Rod Brooks at MIT talks about robots challenging our sense of 'specialness' (that which sets us apart from other species) and I think he's right about that. Over time, androids are coming to look like us, they are beginning to feel emotion and curiosity, they move the way we do and display intelligence. Everything that used to be the sole preserve of the human is slowly being worn away by increasingly advanced robotic creations. Could it be that this makes us insecure?

But the last claim for some kind of superiority might be made by those with the view that robots don't have souls and this absence is what ultimately makes them false. Being made by humans not gods condemns them to a life that is not quite real. Of course, this is a belief that cannot be argued with - the soul is an entirely subjective experience. Humans, analogue or digital, will always be real to me.

Killer Robot Terror: Part 2

So the robot weapons issue is getting some more coverage - check out this New Scientist piece and scroll down to the 'Make Robots Not War' section at the bottom. The more scientists can be encouraged to stand by their views and not particpate in creating autonomous robotic weapons, the better. It's painfully easy to see what happens when science dances to the tune of military masters - Hiroshima, the concentration camps, Unit 731, the Allied firebombing of Dresden. Atomic bombs, genetic experiments, biological weapons - World War 2 really was the first sci-fi war.

The difference now, with autonomous robotic weapons, is that we are at the start of a new destructive technology - these weapons are in the planning stages and do not have to be created. Better to think about it now and stop, rather than be holding UN enquiries in 15-20 years time trying to find who / what was responsible for the massacre of three hundred innocent civilians in some far-flung war zone.

For more on my views on this topic, see here. Just in case you think this is all a little too far-fetched, here's video of the SGR-A1, a South Korean robot sentry that's in line for duty on the border with North Korea.

Book Ends

A recent trip to Amsterdam took in the Anne Frank House - the story is well known but briefly: Anne was a Jewish teenager forced into hiding for two years when the Nazis invaded Holland. The most poignant thing, for me, was seeing the yellow Star of David that all Jews in Occupied Europe were forced to wear during the war. A fading monochrome photograph of soldiers rounding up Jews on the street left me thinking - how could they? The house has the same numbing power as another, more recent, crime scene - Ground Zero in New York City.

Sixty years on and we are at the dawn of another European empire. On arrival back in Manchester, security personnel rushed by to single out a man from my flight to be taken away for questionning. The reason he stood out? His appearance - Muslim beard and robes. Not one of the Westerners on the flight was stopped as we made our way through Customs. Not the first time I've seen those of Middle Eastern appearance receive such special attention.

Plus ça change... Seven hundred miles travel, sixty years of history and prejudice shows its ugly face at either end of the journey.

Killer Robot Terror!

More 1950s style headlines, this week courtesy of that bastion of neutrality Fox News, in an article based upon a presentation from Professor Noel Sharkey of the University of Sheffield. He warns that western robotics expertise could be appropriated by terrorists and used against civilians:

'Current robots are dumb machines with very limited sensing capability. What this means is that it is not possible to guarantee discrimination between combatants and innocents or a proportional use of force as required by the current Laws of War. It seems clear that there is an urgent need for the international community to assess the risks of these new weapons now rather than after they have crept their way into common use.'

The full press release is here. It's good to see the issue being raised because there is always the mindset of 'it's ok for us to use these weapons, but not ok for our enemies to use them'. (In case anyone has been comatose for the last five years, the US/UK alliance is currently 'punishing the perpetrators of 9-11 / removing WMDs / laying the foundations for Democracy & Freedom ™' in two sovereign nations, Iraq and Afghanistan, and the people there don't like it).

I try to take a philospophical approach to the negative aspects of robotics technology - the world is more shades of grey than black and white. Much of the work at MIT's robotics lab was funded by the US military at the time of my visits in 2003 and 2005. Cog, one of the first platforms for social robotics, might not have existed without Pentagon funding. iRobot, as well as developing the Roomba vacuum cleaner range, also make the PackBot for the US military. But it is a relatively short step from a reconnaissance robot like the PackBot to an armed reconnaissance robot like the South Korean robot sentry. I would guess that in the US, it is perhaps a case of researchers feeling that they have to deal with the devil in order to get serious funding for robotics research. A quite different situation exists in Japan. Hiroshi Ishiguro was passionate in his belief that robots should be developed in order to help people, not to hurt or destroy them. I tend to agree - robot weapons should not be made, because if we do make them they will proliferate in exactly the same way as all other weapons have done.

We have a choice about the type of robots that we make, it isn't inevitable that the military dictate the terms of development, setting in motion a tumour-like spread of destructive technology. Robots and androids can and should reflect the best of human nature, not the worst of it.

© Richard Evans 2008

Love And Sex With Robots

So I've recently finished reading David Levy's Love and Sex with Robots. Don't want to get into criticising someone else's work, but I have to say I found the book something of a missed opportunity as it took until page 309 of a 310 page book before David wrote:

'Finally, there is the matter of the ethics of robot sex as it affects the robot itself. When robots are so highly developed that... they appear almost indistinguishable from humans, should we assume that simply because they are not biological creatures it is totally acceptable for us to have sex with these objects of our creation whenever we wish?'

This is the crux of the whole human-machine relationships issue and it is such a shame that it merits just one paragraph out of the entire book (for more on this subject, see this earlier blog entry). One is left with the impression that the prospect of 'great sex on tap for everyone, 24/7' got in the way of a much more interesting discussion: just because we can create something, does that mean that we should?

Human nature being what it is, of course the market will demand android lovers over the next few decades, and while that will reveal much about how we view relational artefacts and what we think of as human, there will be other less welcome consequences. Meghan Laslocky's article 'Just Like a Woman' details RealDoll owners taking out their frustrations in disturbing ways - it doesn't take much to imagine what may befall more sophisticated androids in years to come. 'Great sex for everyone?' Except for the robots...

So maybe the companions we create should be autonomous, not automatons. Able to say no to our demands - for our own good as much as theirs. Surely this will be far more interesting - and healthy - for everyone.

Repliee Article

For a limited time, there is a full length article detailing my meeting with Hiroshi Ishiguro (creator of the Repliee androids) available on my main site. Professor Ishiguro speaks at length about his work and gives some insights into why robots are so popular in Japan, as well as a few hints on what he will be doing next. Follow this link for the full story.

World's First Android?

Rewinding back to when I was writing Machine Nation at the turn of the century, the inspiration for that story came from learning about a couple of humanoid robot projects - one, Cog at MIT and the other Manny, a humanoid developed by Pacific Northwest Laboratories in Richland, Washington. Manny was 'born' in 1986 and sent to the US Army's Dugway Proving Grounds in Utah in 1989. It had 42 degrees of freedom and cost $2million. Designed to test new military uniforms for protection against chemical and biological weapons, Manny was capable of walking, sitting, bending and flexing. In addition, it could simulate breathing and sweating. All this at the same time, 1986, that Honda were just beginning work on what would ultimately become ASIMO.

One of the reasons that Kim Fox appears as the sole android in existence in Machine Nation is the suggestion that she too evolved from a military programme, way ahead of the commercial development of android technology. One does wonder where might we be now if the US had continued to develop androids from 1986 onwards - and of course it's always quite fun to speculate that the US didn't stop their research with Manny's retirement in the late Eighties at all...

Robot Ethics

Much of the real world science that fascinates me comes with a very high price for the animal kingdom - see this story from the New York Times about a robot in Kyoto being controlled by a monkey's brain in the USA. Being vegetarian, my view is animals are sentient and so not ours to eat or experiment on, and my interest in robot ethics would seem a natural extension of this standpoint, as they are in the process of becoming sentient too.

Morality and progress, like politicians and honesty, are not happy bedfellows - whilst I abhor the maltreatment of animals, I understand that scientists who use animal experimentation are not all devoid of conscience and empathy. Some are genuine in working toward the greater good. And the rest of them are probably borderline psychopaths...

Humankind's record with the maltreatment of animals is part of why I think that robots and androids will also be at risk of exploitation and abuse. Much as we saw with the Apollo Moon programme, there may well be a period as robots become increasingly sophisticated where we are thrilled with how clever our new creations are - crowds are regularly enthralled by ASIMO's skills and Repliee's lifelike appearance amazes all who encounter it. But what happens when the honeymoon is over? When we get used to them, when they don't work as well as they should?

And then there are the robots designed to feel pain - already we have the Simroid, an android that is permanently consigned to the dentist's chair. There are other humanoids used to test surgical procedures. At some stage in the coming years, as our technical prowess grows, these devices will cross from mimicking pain to actually experiencing it. Who wants to hold the knife or drill then?

2007 Reflections

2007 has been quite a fruitful year, though I seem to have been working on short stories for the longest time. Touch Sensitive, Girl Absorbed, Half Life, Trick Machine are all in the bag and Freak of Nature is soon to start. Exilium is at last finished and ready to go - lots of news on that front in 2008. This blog has found some feet with which to walk and it appears that at least one of Exilium's characters will soon be venturing into a virtual world.

The world of robotics itself is gathering pace and will no doubt venture into territories uncharted by any writer, filmmaker or scientist, as more and more it enters the inventive hands of Joe and Joanna Public. The year's highlight has to be meeting Professor Hiroshi Ishiguro and the android Repliee Q2. Japan itself proved to be fascinating - intense, subtle and beautiful. Even some traces of utopia amidst the concrete and ritual. The contrast between East and West was unmistakeable - and refreshing - with a Dubai transfer like a crossing point between worlds. What a relief to leave Western-centric news behind for a little while.

The year's biggest frustration would have to be the self-serving negativity of The Fall and their sycophants towards Perverted by Language, though at least there was a plus side with The Stool Pigeon being very kind about Touch Sensitive - so y'know...
Water. Back. Duck. Off.

So 2008 stands before us, a year not yet written. Let's do what we can to make sure it brings peace, tolerance and prosperity to all.

Predicting The Future

Touch Sensitive contains the following description of Kobe in 2030 and an imagined transport system, the übertron:

'...it is the faint whoosh of the Übertron, a monorail system linking the spires of the city’s numerous skyscrapers, that attracts his attention. It is the only form of transport during the harsh months from November to March, the network of pod-like compartments tirelessly shifting goods and people, a vital blood flow that keeps the ice-covered city alive...'

The word breaks down into Über, German for over and tron meaning particle - to represent the idea of a pod. The term just came along one day and it sounded good, so a place was found for it. The story was written in 2006 and just eighteen months later, London's Heathrow Airport announced plans to introduce a personal rapid transport network in 2010. It's startlingly similar to what was described in Touch Sensitive, the only difference being the story envisioned a raised rail network connecting buildings no longer accessible from ground level.

Just shows how tricky it can be predicting the future. Seems like it always comes along much sooner than we think. The most famous example is perhaps that of the original Star Trek - its crew were equipped with futuristic handheld communicators and here we are, just thirty odd years later (never mind waiting till the 25th century) all carrying mobile phones with us everywhere we go.

If human beings make it through another millennium (or even another century) the scientific innovations will be as mystifying as a wireless laptop would be to someone from the Middle Ages. Just about the only sure thing is that by 3007, countless waves of disruptive technologies will have changed our world, and those we choose to explore, beyond all recognition. However, our collective capacity for deception, tyranny, violence and other acts of general stupidity will doubtless remain impervious to all the innovation.

What's Inside ASIMO?

Honda have revealed a new networked version of ASIMO this week - for clips, go here - it has obstacle avoidance and other new skills that bring it ever closer to being a useful companion. Elsewhere, there is an interesting comment from one of ASIMO's technicians in this intriguing article. It quotes ASIMO operator Ekkasit Najaitrvek stating that the robot is 'much more than a super-computer'.

"He's a person," Najaitrvek says. "Sometimes he's like a kid. Most of the time he's my friend. "You can see when he looks at you -- you can't deny he's a person."

When I saw ASIMO at the London Science Museum in 2004 (full story here), I had a similar reaction - at one point it 'looked' right at me and I felt I was being acknowledged by a person. It was an innate, unexpected response. Why is it that ASIMO - and other robots - have this effect on us? After all, ASIMO is only at the start of its independent intelligence - it is an infant mind, more programmed than autonomous. Clearly appearance, movement and purpose add up to something greater than their individual parts - the question is, what is it that they add up to?

The Doll's Hospital, Manchester

When I was a young boy back in the 1970s, I remember taking various broken Action Men (for those outside the UK, these are a GI Joe-type figure) to a strange place in the centre of Manchester called The Doll's Hospital. If it existed nowadays, it would perhaps be the stuff of nightmares - it was usually dimly lit, confined, with cabinets and shelves stuffed full of dolls and figures from all eras, some no doubt of tremendous value. Little faces peered out from behind the glass, gazes fixed and vacant. The proprietor seemed taciturn, he wore a white apron and always had various mysterious tools at his beck and call. The toy was left for a given period and when it came time to collect it, the repairs were perfect and the doll was as good as new.

The dolls in this weird little shop were a perfect example of the Uncanny Valley phenomena - staring eyes, fixed expressions, not quite human enough. The place was full of what are now termed relational artefacts - toys that children had formed attachments to - some in for repair, others clearly forgotten.

One of the visions for Sony's Qrio was to make a robot that we would love and form an attachment with - just as we do with our pets and with each other - a robot that would be a lifelong companion, teaching and entertaining both adults and children. We are used to the idea that toys will be discarded as we grow older - left behind as other interests take precedence. But what if your robot cries when you leave it or throw it away? Already Aibo does a little dance of excitement when it sees me, mimicking delight at the return of its owner. Relational artefacts promise to be a fascinating, disconcerting area of robot development and research. Perhaps future dolls' hospitals will be more akin to orphanages than shops for repair.

Web 9.0

Exilium, Touch Sensitive and Trick Machine all feature the eScape - an imagined hybrid of the internet and a Second Life style virtuality. In RealLife™, we are told that we are now on Web 2.0, full of user-generated content. Social networks, blogs & YouTube have perhaps re-democratised the internet, disgorging new celebrities by the week like the offspring of some Alien Queen. The quirky and bizarre famous for Warhol's 15 before burning up as quickly as meteors in the stratosphere.

In the new stories, the eScape is also home to the thoughts and memories of various android characters, an out-of-body storage vault and sensory realm. Digital lifeforms will naturally have an extra dimension that they will experience, at the very least for software updates, data storage, identity backup. As experience shapes the androids, memories will be stored both online and within artificial brains. If memories and personality are stored in virtuality, then identities could be downloaded into new bodies, perhaps even more than one at a time. Is this the end of individuality? The threshold of immortality?

List of Androids

Like something out of a 50s B-movie, there is an ever-increasing number of androids (robots that look like people) being created all around the globe - in Japan, South Korea, China, Canada and the USA. Some are more intelligent than others, some more realistic than others. Some are surrogates, some duplicates. So here's a list of what's out there - know of any I've missed?

University of Osaka: Repliee R1, Q1 & Q2
ATR: The Geminoid
Kokoru Dreams: Actroids
Kokoru Dreams: Simroid
University of Tokyo: Saya

KITECH: Eve R1 & EveR2 Muse

XSM: Zou Ren Ti
Unknown: Dion

Le Trung: Project Aiko

Hanson Robotics: HumanKind

There have been a few other android projects that are now dormant, including iRobot's My Real Baby, which I encountered in my 2003 trip to MIT. We stand on the brink of a strange and wonderful future. At last.

Qrio-ser and Curiouser

How sweet. Kind of echoes Professor Ishiguro's comments in Osaka that children under five have a somewhat broader idea of what they will accept as human (more on that soon in a forthcoming feature article). Always thought it was a shame that Sony canned Qrio - seems there are some research models still out there. Maybe it will reemerge in years to come? For now, here's something to remember it by.

One of the many interesting things that came up in the discussion with Ishiguro-san was the outline for an experiment where a humanoid robot could track a group of kids wearing RFID (Radio Frequency Identification) tags. The hope is it may be able to determine social structures within the group - by tracking which kids talk to each other and, crucially, which kids are getting too much attention (possible bullying) as well as those being left out of the group dynamic altogether.

Android Lovers: Post Modern Geisha?

This week brings a headline grabbing press release from the University of Maastricht on David Levy's thesis 'Intimate Relationships with Artificial Partners' which suggests that 'love and sex with robots are inevitable'. Not just inevitable, this is already happening. Although mostly inanimate (animatronic alternatives do exist), a coterie of people have sex with their Real Dolls, and there are hundreds of thousands of people for whom the Sony Aibo robotic pet has become part of the family. The subject of human-android relations is explored widely across all three of the Alex Sorber / Kim Fox books, especially in Machine Nation, and the key issue is not if or when it will happen, but what sort of 'love' will it be?

My concern with Real Dolls and their ilk is that they are just what they say they are - dolls - mere objects for the projection of the owner's emotions. No mind sits inside their silicone heads. Even when they become more animate, with a gift for pillow talk and other seductive subtleties, they will still be mere slaves that can be programmed and controlled. I don't doubt that humans will fall for these karakuri and that there is even a place for such things - there are those whose circumstances mean that sex with human-like robots is preferable or easier somehow. But there is also undoubtedly a fetishistic aspect to the scenario - people with a predilection for sex with robots are known as technosexuals - so perhaps robots will become the ultimate sex objects. That can never object...

The reason I rate the Aibo so highly is that, in its small way, it has its own drives and a sense of curiosity - it's a small model of an independent mind. Should this trend evolve, then in decades to come, maybe we'll have androids and gynoids with their own minds, their own drives and interests and desires. Like us, but different. So let's turn the concept on its head for a moment: they may choose to have relationships with humans, or perhaps they'll shun us - preferring instead romance with their own kind. Now wouldn't that be the ultimate, wonderful rejection?

On Japan

Bullet trains and androids, geisha and sumo, J-pop and pachinko parlours. Subtle etiquette amid bold layers of architecture, cuteness now softens years of ritual and austerity. It's a little like home, but with a twist. That makes me smile.

Japan is forever on the precipice – the earth could tremble at any moment. In 1945, the country slipped off the edge and it's as if the humility lingers. It is a polar opposite to the UK and USA - here there are manners, slenderness, a government that funds infrastructure not war. The divine is in the detail - Braille signs embedded into handrails on the subway. Bicycles left unlocked in parking racks. Technology that helps make life a little easier.

Same globe, different planet.

Repliee Q2: A Curious Obsession

2002’s Machine Nation, described an encounter between Alex Sorber and the android Kim Fox, in a Boston robotics lab in the year 2027:

‘A kind of pregnant energy, like the calm before the storm, inhabited the space. They all said they could definitely feel a presence. And they’re right, he thought, there was a presence. Existing but not living. Being but not thinking. Completely still, frozen in space and time, a flower blossoming from a bed of junk metal. She was a vision to behold.’

Not for one moment did I imagine that in September 2007, I would have my own similar experience with the android Repliee Q2 at the University of Osaka’s Intelligent Robotics Laboratory....

...The first time I saw her, I passed her by, a fleeting glance of a figure behind a curtained off area of a lab. As if she were no more unusual than a student at her desk. But something, stillness perhaps, marked her as being out of the ordinary.

Later, we had a more formal introduction – a chance for a few photographs and more careful study. Pale pink shirt, black pants, shoulder-length auburn hair. Her appearance quite demure. Silicone skin, detailed with eyebrows and lashes, her arms somewhat limp at her sides. The skin looks ‘right’ on her face, a little less so at her neck and fingers. She was blind and deaf when I met her – an array of out-of-body cameras, microphones and positional sensors were all turned off. But when motion came through electricity, computers and hydraulic pumps, she moved and spoke like any other person. There seemed nothing unusual about her at all.

Then, she was powered down and for a few minutes, we were left alone. Me and this presence. I stared at her, the way I would never stare at a flesh and blood person. I waited for her to move again, but she did nothing. Embodiment made her presence palpable, even though we did not touch. In this off-state, she has animation in potential only; but she is not dead. Instead, she is in a third state, life still present in the machines that power her, ready when the switch is flicked to bring her back into motion. Her maker, Professor Hiroshi Ishiguro, talks of the Japanese seeing the soul in all things, animate and inanimate. Perhaps then, this off-state is her sleep, tranquil moments of recuperation between periods of activity and sensation. Her dreams made on a quiet campus in 21^st century Japan.

For now, Repliee is complete. She cannot yet walk, but she is a step on an evolutionary path. She and her kind will become the perfect expression of Japan as it aspires to be. Helpful and loyal, efficient and unthreatening. Technology that is warm and soft to the touch. As Japanese as Hello Kitty and the Osaka subway system. The future is here, and it wants to hold us and teach us and make our lives better.