Preprints & Reprints
Back to Preprints & Reprints > Publications & Opinion > Homepage
DON'T SILICOMORPHISE ME
Peter Cochrane
Prologue
A member of the Artificial Life community recently pleaded that we should not anthropomorphise machines because they might not like it. Well, I feel as if I am being Silicomorphised by technology - and I don't like it! Even before the start of the electronics revolution in 1915 we have been bending people into technology. Well, it really is time to start bending technology into people - it is supposed to help and not hinder us. But our technology has been limited, and only recently have we been able to humanise machines. If information access equates to power, then usability becomes a moral issue, and the have and have-nots are all at risk. The question is - can we rise to the challenge - do we know enough about humans to realise better environments?
Riches
For aeons we have had a richness of male, female, race, colour, creed, art, science, a multitude of human skills, paper libraries, language and communication. But our species is in stasis, coming to the end of the road! Our wet ware (brain) limitations now sees us more incomplete than ever before. Partial people dominate with just a minuscule education in a minuscule portion of an increasingly vast spectrum of knowledge and experience. Take away our machines and we would be severely limited - relatively speaking we can remember little more than we need to survive. No human knows all the processes necessary to produce the machine on which this is being typed for example. But the machines do! Soon another richness will appear as intelligent machines emerge and create new mind sets. What we must ensure is that we have the interfaces to interact fully with them - and that they have a mind to help and assist us.
HAL
All spoken languages are extremely complex with the simplest sentence containing a wealth of information, conveying mood, emotion, personality and gender of the speaker. Not surprisingly, getting a machine to generate and interact with such subtle information is a major challenge. To date human-machine I/Os have been technology limited and dominated by the switch, button, keyboard and screen. Our natural mode would see us talking to the VHS, washing machine and computer. The inconvenience of typing, remembering obscure commands and navigating GUIs may soon look quaint as a modest degree of speech I/O can now afford reasonable control. Whilst true conversational interfaces are some way off, the integration of speech recognition, synthesis, natural language processing, and search engines allows new paradigms.
Although the public imagination has been fired by HAL (2001) and the ever-helpful computer on the SS Enterprise, the goal of producing natural language interfaces between humans and machines is some way off. The difficulty arises due to the great variability in speech characteristics. When a word or phrase is uttered by different speakers, or even when repeated by the same speaker, the variability is large in terms of the precision world of machines. Beyond the acoustic, ambiguities occur at higher levels of linguistic abstraction. The phrase 'it's easy to recognise speech' could be interpreted as 'it's easy to wreck a nice beach' at a conference of surfers! To make matters worse people do not always say what they mean, or mean what they say!
Looking Down
If I gave you an arrowhead and ask you to bind it onto a shaft, it would be unlikely that you would hold the shaft at eye level and work looking straight ahead. Most likely you would sit or crouch, look down, and work in a stooped manner. Similarly if I ask you to write a note. For reasons locked into our past, we are ~20% more efficient when we read and write looking down on a sheet of paper than when we look straight ahead. So why we do have vertical computer screens when we can improve our read, edit and compose abilities by ~15% by just looking down? Probably the TV paradigm and the keyboard! We are conditioned to vertical screens for entertainment and information display, and the keyboard just gets in the way and makes it difficult to see over the top.
Spreading out a complete work over a desktop and looking down at the entire entity, is a powerful means of enhancing composition and understanding. On the other hand, the screen is wonderfully flexible and a different kind of workplace, not just a metaphor of the desk top. If the two are going to merge and realise the advantages of both, then it is necessary for both to be active. It is also necessary to get rid of the keyboard and mouse. Imagine for a moment an active desktop with the definition and contrast ratio of paper, but with the flexibility and intelligence of a computer. For such a work space the electronic desktop would no longer be a metaphor, but a step ahead. So computers might see large, active, high definition, desktop screens, voice interaction and natural hand manipulation of objects. Combine this with a vertical screen for video-conferencing and telepresence and a new environment is born which resonates with us.
VR
We are very pictographic animals, able to absorb animated, 3D colour images at a phenomenal rate. It is unnatural for us to read, write, and interpret spreadsheets - we are soon overloaded and overcome by information in such formats. Moreover, many of us have a natural inability to mentally translate 2D drawings into a 3D models. And yet, if information is presented in VR, we are all inherently able to absorb and understand the equivalent of a twenty volume Encyclopaedia in about 15 seconds.
Designing all but the simplest products and artefacts on paper has had its day. VR affords greater understanding, facilitates simulation and testing, and realises savings. Producing anything from a mobile phone through to an airline passenger terminal can see savings >30% through the use of VR. Visualising the final design in virtual operation is a vital step in getting it right first time.
Medical applications are developing rapidly with body fly-throughs, operations and simulations. But in this area, along with many others, it is the mixing of the real and virtual worlds where the greatest advantage probably lies. Combining telepresence and VR allows surgeons the benefit of a real world augmented by computer simulations. In recent trials surgeons have been able to stand 'one inside the other' at a distance to experience new surgical techniques for the first time, or receive reassurance during a first solo operation. This technology is equally applicable to the repair of computers, jet engines or heating plants. It offers a new and alternative approach to education.
VR is a medium for experience and we can now step inside the atom, fly a proton and experience fission, rather than just gaze at complex equations. We can dissect a virtual frog and operate a virtual microscope. For the first time we can see and feel the binding energies in the alignment process of a long chain molecule whilst simultaneously viewing the equations and associated graphical information.
Icons
Just as the PC has made the transition to the iconic world, it is clear that such a world has a limited future. For >15 years icons have provided a rapid means of navigating computer environments. But they are flat, static, and only convey limited information. Moreover, they now clutter the desk top to the point where they hide from view. This is compounded by the layering of pages and applications. It is often impossible to simultaneously see your work and your icons. Is this then the end of the line for icons?
Perhaps icons could have meaningful shapes, and be dynamic and interactive, changing colour and form with use and status. With a modest amount of artificial intelligence they could mimic human traits to seek our attention and help us, or hide away until a more appropriate time. Giving them personality, contextual reactions, a voice, and perhaps facial expressions, would increase their ability to communicate with us.
The next obvious feature would be to put them into a 3D rather than 2D world so we could realise freedom of position and movement. We could then be guided in decision making and information navigation in new and novel ways. All of this is possible on the PC platform today - so perhaps the iconic world has not had it's day, perhaps it is only just beginning. Realising the full potential might ultimately mean reaching out to touch and feel substance. This would be another dimension that would parallel shaking hands as icons could then guide us, and be guided by us, in a virtual data world.
5 or 50
Exposure to computers is occurring at an earlier and earlier stage of life. The simplest machines are now ~?60, and include voice synthesis, and a good deal of interactivity. For many children the IT education and experience build-up starts before they are 1 year old. It is acquired in the same way that we learn to walk, throw, and eat. It is not something special, or taught as a subject, but subsumed like language and becomes intuitive. Not surprising then by age 5 children feel comfortable with technology and by 10 years can be skilled and capable. What chance then the 50 year olds who feel inhibited by technology, have a pride barrier of having to admit that they do not know and lack the time to learn? Who knows! Perhaps they can look forward to a sunset period where responsibility is handed on earlier and they will have time to play too.
Serendipity
The medieval librarian was the guardian and regulator of information, the contents list, index, filing system and retrieval mechanism. He alone decided who saw what, when and where. There was no open information access, it was regulated and controlled. No serendipity! For >200 years we have enjoyed increasing levels of serendipity with the decimalisation of paper libraries and the librarian's transition from guardian to assistant and information agent. Serendipity is now free; merely walk through rows of shelves and chance upon that book, newspaper or obscure journal. Contrast this with the seemingly infinite world of Internet. Here we have almost 100% serendipity - an overriding lack of order - no sign posts or eye catching indicators Actually finding what you want is now a challenge. In this environment information seems to come in two dominant classes: that of no interest, and the distracting, but of no direct benefit.
The CD ROM has almost zero serendipity. So well organised, sterile and deep is this medium that drilling down to the information you require can involve >5 clicks. A total lack of visibility of information either side of the mine shaft you dig is now a limiter. You soon find yourself totally disorientated and lost. With no frame of reference to help - you resort to Control Quit and start again!
Perhaps somewhere between the Internet and CD lies the ideal with the right percentage of serendipity allowing us to optimise our creativity and work rate. The question is, how is it to be realised? Serendipity by design is a challenge, for in the past it was realised by accident. Perhaps we have to wait for machines to spontaneously create this serendipitous environment for us with artificial agents learning about our habits and interests through direct observation. They could then take on the role of a personal librarian roaming global data banks on our behalf. We could then spend more time with people discussing problems and formulating our views to create new forms of serendipity that have so far escaped us. Either way we face a major challenge as mankind's knowledge now doubling in a period of <2 years. This is especially so with the creation of increasing levels of short-term information that just acts as a fog to us spotting what we are really looking for. We will need help through increasing levels of AI to cope with a world where information half lives fall <6 months. We have to hope that the machines can help us, or face an increasingly sterile and less creative world.
Chaos
Just 20 years ago all telephones were on the end of a wire with users making ~3 calls per day at unrelated times. True there were busy hours, with meal times and tea breaks seeing a distinct lack of calls, but by and large calls were random events. This changed with the arrival of TV phone in programme. Someone singing a song on TV could result in half a million people telephoning London to vote for their local hero within 15 minutes. A new world of network chaos was born! With the mobile telephone a new phase erupted. Traffic jams, train and plane cancellations all trigger correlated activity - everyone calls home or office within a few minutes. So cellular systems become overloaded as thousands demand connection at the same time. A transition has thus occurred, from a random world of distributed events, to a highly localised and correlated world of activity triggered by anything causing us to act in unison.
Whilst this might seem trivial and easy to repair, consider the prospect of NCs. When 5 or 10 of us meet our low cost NCs will be plugged into the same line or server. At critical times during our discussion we will wish to access information or download. This will be correlated activity with a vengeance and on a large scale difficult to contemplate. Many people equate chaos to randomness, but they are very different. Chaotic systems exhibit patterns that can be near cyclic and difficult for us to perceive. Random systems on the other hand are totally unpredictable. Curiously without computers we would know little about chaos, and yet they may turn out to be the ultimate generators of network chaos on a scale humans cannot be able to match.
Socks
If I walked into your office and asked to use your pen, telephone, or copier, chances are you would agree and think nothing of it. But if I asked if I could use your PC you would probably react as if I had asked to borrow your socks. Most see PCs as a really personal item - an extension of us - like our clothes and jewellery. Someone entering our world of data is akin to the prospect of being burgled. Will we react in the same way when we have NCs with limited personal and local information storage and processing capability? Perhaps not! At that point the computer will have assumed a similar level of depersonalisation as the telephone, radio and TV. Access to those bits we don't want other people to reach will be easy to control and limit. And our applets will be a communal facility anyway, like a bus or a taxi. Moreover, it will be a world where we access our personal information from any convenient terminal or location.
Wearables
Beyond the PC and NC we can expect computers and communicators we wear with voice I/O, optional head mounted visual systems. Ultimately the technology may become invisible as it is embedded into the fabric of buildings and vehicles, clothing and other personal items that think. We can also anticipate that it will become humanised and able to develop personas to suit us and our specific needs. At this point the socks syndrome will probably reappear as the technology becomes more than a piece of technology and more a part of us.
Delay
Only 15 years ago using computers and making telephone calls was a bit of an ordeal. Processing speeds meant waiting for seconds or minutes for screen fill or print out. Concatenated electromechanical switching delays meant you could wait over 30 seconds after dialling the last digit before you heard any ring or engaged tone? How different today! Now we get irritated if we do not hear ring tone immediately we press the last digit on the key pad, or when we have to wait seconds for a PC application to load. The generic problem is having to wait for a period that is too short to do anything else, but long enough to break our concentration. Delays of a fraction of a second disrupt our mental agility and interactive creativity to an alarming degree. But today we have an abundance of bandwidth, storage capacity and processing power, with optical fibre and power PCs. Moreover, technology promises higher levels of circuit density and clock speed at minimal cost. We may thus be approaching the realisation of a dream; to access everything, everywhere, anytime, within 3 clicks of a mouse and have screen fill and interaction within a second. For us to enjoy natural, and effective communication with people and machines, in real or virtual worlds, the need is for sensory delays of <100ms.
Just watch children interact with machines and it is apparent that they have an insatiable desire for instant gratification - the shortest response time and best graphics. Looking at professionals you see the same phenomena - a desire to be able to do more, faster. All the technology required is available; we only have to adopt the right mind-set and implement solutions. In the meantime, I suspect our progress will be frustrated by the delays of systems, computers, software and interfaces configured for the past.
Chickens
In some quarters there appears to be a fear that as technology advances, we will regress. Yet with the invention of the plough we are able to create more food, with the axe with were able to cut more wood and build, and with a bow and arrow we became very effective hunters. Today, very few of us have to use any of these instruments of production and killing - we have moved on. In the same way as the printing press negated the need for thousands of quill pen guiding monks, IT is empowering people, realising new skills and allowing them to do new things. In fact they can now do, experience, and achieve far more than any previous generation.
In business, medicine, production and education IT has a major role to play. It is no longer feasible to organise the logistics of a nation or the planet on the basis of the written word by hand on a piece of paper. Reverting to that process would see millions of casualties - people would simply starve for want of communication. Nor is it possible to go back to the days of Aristotle and word of mouth teaching. Thinking how a chicken works can be a lot of fun and very satisfying, but in reality taking a chicken apart, or better, building a chicken is far more productive. And so it is with interfaces and humans - there is only play! In many areas of technology and interaction playing is the only route to true understanding. Interfacing with machines demands humanised I/Os so everyone can access their power and potential. A more natural interface is now a prerequisite for future progress.
About The Author, Further Reading and More
http://cochrane.org.uk
Food for Thought
We may think ourselves very clever and at the top of the evolutionary tree, but the reality is we face some major challenges with the creation of AI and AL systems. Our progress is inextricably linked to the ability of the machines we create. To ignore their potential and not use them wisely would be foolish. To give them interfaces we find unnatural and difficult is equally foolish. The real challenge is for us to get them to serve us and our needs in the most efficient way. But we probably understand machines better than we understand ourselves!
The technology is getting exponentially more powerful and flexible, and it can now be progressively humanised. Soon computers will equal our mental capacity to store and process information. So increasingly, we are the major challenge, and not the machines.
This appeared on my screen one Sunday evening and I have never been able to track down the original source. To the author I extend my thanks and beg his or her indulgence for my editing down from the original.
Subject: Captains Log, Stardate 9986.104 , "Sentient Meat"
"There's no doubt about it. We picked several from different parts of the planet, took them aboard and probed them all the way through. They're completely meat."
"That's impossible. What about the radio signals? The messages to the stars."
"They use the radio waves to talk, but the signals don't come from them. The signals come from machines."
"So who made the machines? That's who we want to contact."
"They made the machines. That's what I'm trying to tell you. Meat made machines."
"That's ridiculous. How can meat make a machine? You're asking me to believe in sentient meat."
"I'm not asking you, I'm telling you. These creatures are the only sentient race in the sector and they're made out of meat."
"Maybe they're are a carbon-based intelligence that goes through a meat stage."
"Nope. They're born meat and they die meat. We studied them for several of their life spans, which didn't take too long".
" Okay, maybe they're only part meat. You know, a meat head with an electron plasma brain inside."
"Nope. We thought of that, they're meat all the way through."
"No brain?"
"Oh, there is a brain all right. It's just that the brain is made out of meat!"
"So... what does the thinking?"
"You're not understanding, are you? The brain does the thinking. The meat."
"Thinking meat! You're asking me to believe in thinking meat!"
"Yes, thinking meat! Conscious meat! Loving meat. Dreaming meat. The meat is the whole deal! Are you getting the picture?"
"Omigod. You're serious then. They're made out of meat."
"Finally, Yes. They are indeed made out meat. And they've been trying to get in touch with us for almost a hundred of their years."
"So what does the meat have in mind?"
"First it wants to talk to us. Then I imagine it wants to explore the universe, contact other sentients, swap ideas and information. The usual."
"We're supposed to talk to meat?"