Ladies and Gentlemen:
I understand that all of you are inventors, innovators, creators of new ideas. This is most attractive to me, for it is on just such an audience as this, that I can try out some possibly new ideas and get an immediate constructive criticism without the usual year or two of patient repetition while innovation is being rejected as ridiculous or crazy. Much invention and innovation is ridiculous or crazy and needs to be rejected, just as even the richest ore is mostly dross and needs to be refined. So let me offer you some ideas for repeating, improving, or translating into other areas of invention.
Perhaps I should begin by suggesting that the vapor pressure of inventors is very high, so that only by supercooling an environment for a week in advance can you freeze out this many inventors into one concentrated body.
I have been surprised by being asked some five times during the past two months to speak to widely different audiences on seemingly very different topics, but have found this interest in invention, innovation, “where do we go now?” a major consideration. Is it possible that we are at last realizing that the technological “slump” that our high technology industries are experiencing in competition with foreign industry is real, and of our own making, and that we must, and can, pull out of it?
A few years ago I was asked to organize a keynote symposium for an Airlie House meeting on Artificial Intelligence and Pattern Recognition. I chose the topic of “Recognizing Patterns in Pattern Recognition” and found that the group had never turned its pattern recognizing abilities introspectively onto itself, to establish a taxonomy of pattern recognizers, each group with its selective bag of tricks and unspoken but powerful taboos, fitting it nicely for one approach but leaving it ineffectual and vulnerable in other areas.
As imaginative people, we pride ourselves on devising new and ingenious solutions to varied and devious problems, but in fact each of us is bound by tradition and available strategies to a minute fraction of the insight that lies just in front of his nose.
I would like to propose that, instead of inventing each invention by sheer mental effort from the ground zero, we invent a theory and set of principles of inventions so that a new result can be cranked out to order by a reasonably clever person, aided by adequate experimental technical trial and error effort and a suitable family of computers.
You will note that I said a suitable family of computers and did not mean merely a suitably large and fast conventional binary computer. Perhaps the heaviest burden we carry as utility, and occasionally brilliant, inventors is the necessity of thinking in the very limited family of languages and hardware implementations available in present macro- and micro—computers. In this we have negligible associative memory abilities, wretched pattern recognizing ability, and only clumsy means for scaling and dimensioning human value judgment problems that are at the top of our need list.
Another time I would like to explore with you the possibility that the die
was cast for our present binary digital advocacy forms of law, religion, legislation
and digital computers by the rise of the pragmatic Roman empire from the more
continuum value and philosophically exploratory cultures that produced the brilliance
of Greece roughly four centuries earlier. I fully expect we will see the dawn
of value-judgement-oriented computers within the next decade or two to aid us
in fast insight and invention.
Not until I was asked to give the Sam Salkin lecture last fall did I ever put these principles to work on major concepts of invention, social responsibilities and education, organized to accommodate my views of all three. I realized then that our primary education is heavily biased toward beating into our youth a highly structured but guardedly liberal data base of “facts”, “dos”, “tin, no”s and some standard technical procedures, e.g. arithmetic, public and private game rules and appetites. Secondary and college education is dedicated largely to techniques, strategies and tactics. “How to” is prominent in Business, Law, Engineering, Science education, even if it does not break through into writing of clear and persuasive English prose, technical or non-technical. Only last week I was told by the head of one of our large engineering departments, with a little pride, I think, that “We do not teach our students modern engineering technology because that will be obsolete by the time it is wanted, but instead we stay with basic principles.” Ought we not to teach at least current technology, if we cannot properly teach anticipated new technology?
When do we get taught, by precept and apprenticeship, how to create new ideas and simplified ways of handling existing problems faster, more efficiently, more cost— effectively?
The key to this problem is the recognition of language creation as a problem-solving and invention technique. While we do have a fair share of new computer languages, developed to facilitate programming for a specific area of application and hardware-firmware capability, few of
us carry this to the point of really generating new efficient ad-hoc mathematics or computer implementation of such mathematics. I am especially interested in generation of new languages of manipulation, insight, and problem-solving based on biological precedent, be it macromolecular, cellular, organismic or even social-economic. I like the ring of the phrase “Figure of Thought” applied to such new linguistic concepts, in contrast to the familiar prosaic Figure of Speech.
Consider for yourself when you have last generated a useful, comfortable figure of thought 5 mathematical, technical or humanistic, good enough to be put into your everyday vocabulary of thinking, to be published, or to be stolen from you.
I shall leave out of this address completely the fourth layer of education that I think some small fraction of our advanced students should encounter, or at least recognize as having an existence. This is the inductive generation of unifying principles and the associated optimizing principles.
Now I fear that you will be starting to think that I am going to talk only of vague principles and generalities that have no practical application. I hope that is not true, but I must risk loss of your attention while I make one more system observation, then we can go on to my menu of inventive procedures and some feasibility runs that I have been making to test them.
The remaining bit of theory is the requirement for closing the classical Wiener-Shannon
information theory loop topologically. We must discover the complimentarily
of code and message, of heredity and environment, of hardware and software,
and need to discover that these are all efforts in different scientific dialects
to perfect just one figure of thought. Even the engineering concept of feed-forward
as related to feed-back is an effort to formulate this utility figure of thought.
We, as biological organisms, have a start-up package of hardware and primitive
but amazingly versatile aria aLurable software with which we proceed by trial
and error encounters with environment to build higher level languages in which
to conduct our communication, control, and, I would add, conscious perception
of our universe. Some of us build good chess software, build some/good common
strategy, some become inventors.
• But now to the menu of basic inventive/deliberate strategies. We can invent by deductively monitored/controlled inductive aberration. In simple terms, we can invent by making hypotheses that are deliberately just slightly illogical and then examining the implications deductively to see whether the slight aberration is a new valid principle.
• We can invent by linguistic translation. We state a problem in one a form and translate it into/nearly equivalent, already solved scenario.
• We can invent by linguistic reformulation. This is analogous to mathematical change of variable procedure and is quite different from the translation of topic 2.
• We can invent by matrix inversion, e.g. we study the problem from the answer end instead of the usual question end.
If we are serious, however, about developing invention and technology transfer to order, we must attend to changing our environment to an extent that will make this acceptable and even profitable. We must do some things that come hard to died-in-the-wool introverted inventors.
1) We must learn to advertise our wares repeatedly and simply, if necessary keeping up a theme with variations for apparent novelty, for years~
2) We must learn marketing techniques to turn ideas into systems that will automatically demand series of inventions to implement them.
3) We must learn about creating heroes and High Priests advocating community and individually beneficial innovations that will otherwise never see the light of day.
4) We must rearrange regulatory agencies to be optimizing rather then self-serving, adversary structures.
5) We must rediscover procedures analogous to those now remarkably successful in Japan and now rising in several west European nations in which government, industry and academia collaborate to solve problems instead of building strong walls of protective isolation between themselves.
6) We must reexamine our patent and trademark law and traditions to see whether they need modification to deal with the new realities of compute hardware-software and firmware and the primary intricacy of communication and control systems, especially as they become more intimately incorporated prosthetically into our individual and social bodies.
I compiled a list of questions in this area for a recent panel session of prominent corporation patent lawyers talking to a group of computer enthusiasts and engineers. Several of the questions simply have no reasonable answers in the current system.
Now having explored some of the principles of invention to make it less mysterious
and less expensive in intensive intellectual effort, let me turn to the almost
unlimited body ot innovation and feasibility tests that this approach opens
up, some of which we are examining in my own laboratory and other laboratories
where we can get the ideas introduced. Each example embodies at least one of
the stock principles. Naturally they tend to cluster around medical and health
problems or around control and computer electronics as these are our areas of
lesser ignorance.
1) Utilizing humans as relatively inexpensive components in computer systems to do the pattern discovery and recognition that they do so well in a multidimensional hyperspace of time changing data. Here we use a computer-generated hands-in space of 3, 5, or even 8 dimensions which the observer views through solid stateelectrodichroic glasses to inspect instantly an enormous volume of data for geometrically describable features. This device, really only an update of my old spatial resolver, now widely used, is enhanced by polychromatic display and local vector arrows in a three-dimensional geometric space. This device, practically without change except software, leads to a most interesting “frustrator” procedure for testing an operator's reaction and adaptive ability in a three dimensional pursuit procedure where the computer gauges operator skill and can automatically introduce first order or second order control dynamics to make the pursuit just possible or tantalizingly impossible.
2) We are exploring the practicality of a personally portable whole life medical history card like a passport that can accompany the patient and be under his control, to be serially updated but never deleted as the patient moves about during his life. This record, ergodically assembled from many patients over perhaps five years, could give us individually tailored medical diagnoses and prognoses and leave us with a great freedom to seek the best and most satisfying medical and health care adapted to our varying life styles. Obviously we could keep safe deposit copies of this record~ occasionally updated and avoid such disasters as that in Minneapolis last January 15th when thousands of patient records burned in a Medical Center fire.
3) We have been discovering how to phase-lock-loop people so as to get them into an episodal stationary, but not constant frequency pattern from which an epitome cardiac and. respiratory pattern can be extracted, much the same as we routinely lock onto a noisy or fluttery satellite signal. To our surprise we find that the patterned variations may oe telling us more about the inner control loop of the person than the noise freed cardiogram does about the heart. Perhaps we have opened up a most interesting bag of worms by realizing that this episodal biological time is really quite different from clock time and should really have its own identity. We have even found that this biological time variability leads easily into a valuable form of computer data compaction.
4) Recently we have realized that there is a market for several billion dollars worth of highly cost-effective microcomputer implemented instrumentation in the family participating medical testing, monitoring, and therapy Vield. A direct outgrowth of the do-it-yourself shift
across all domestic technologies as formally train/and centralized services price themselves out of the market, this one has around it many real and many false taboos and caveats. Surely we do not want to put dangerous drugs and instruments into totally unskilled hands, but with few week a miniature / certification course, most individuals who can get a driver's license could easily qualify to take a daily cardiogram, read a blood or urine sugar level or offer tender loving care with monitoring instrumentation to patients who would otherwise require extended, expensive hospital stays.
I offered recently to the advanced products group of one of our large Minnesota
high technology corporations a list of some forty of these home health care
devices that could be built inexpensively to about the quality level of ordinary
good television or hand held calculators. They surprised themselves and me by
ordering the merits of the devices in an unexpected sequence. They were very
favorably impressed with an instant alcohol level test, they liked having a
nervousness meter, they favored computer-aided relaxation, even biofeedback
exercises. A direct reading tremor index of muscular stability seemed attractive,
and to my complete surprise, a sperm concentration and mobility meter was found
attractive.
Obvious extension layers of this family participation medicine lead from the self-standing instruments dedicated to one kind of measurement to a generation of modular peripherals appended to a utility microcomputer control unit that most homes will soon need anyhow.
Further extension links this home micro by dial-up telephone and slow scan imagery to central medical facility backup of the home system by the professional medical expertise, often without the need of physical travel of subject, doctor, or medical technician.
This dial-up facility opens the way also to personal library work to become more expert technically in a health field and possibly to prepare for higher levels of certificate and deeper participation in important medical decisions.
A broad new line of research, triggered by the now familiar biofeedback training to control alpha wave generation, heart rate or other seemingly inaccessible function, is the bare possibility that even more exotic targets can be brought within access without drugs or surgery.
Even though biofeedback modification of homeodynamic control has been demonstrated beyond much doubt, its practical medical utilization is still very limited and it is not certain that its role will ever become widely useful. This cautious approach is probably wise. Suppose we were really able to teach the body to readjust the control algorithms on its hormonally governing glands where malfunction is not due to organic inadequacy but to improper control. This could just possibly be the start of a new drugless therapy of hormonal imbalance.
In conclusion I would like to raise one more issue of invention and innovation. How quickly and how intimately should we allow computer-like devices to be incorporated into our bodies in disease and in health? Please regard this as a very realistic issue of the next ten years, not some wild blue yonder speculation. Translate the case of the desk calculator to the intimately incorporated prosthetic computer. We have already installed surgically hundreds of thousands of electronic devices, pacemakers , pain alleviators, and now injectors of chronically needed drugs. Up to now we have avoided installation of test ports and 1-0 ports of convenience for rapid reliable electromagnetic communication with external instrumentation, incoming or outgoing. Would you like to have such enhanced transduction, are there new dangers inherent in these natural extensions of computer science? I suspect there are, but I hope we will face them actively, not by outlawing or evading the issue.
As you can see, I believe we can construct a theory of invention, a technology of invention and even a quantitative appraisal of community and individual value of invention, in contrast to the usual image of invention as an entirely spontaneous or totally unpredictable process.
If we can bring into realization through our inventions just one of the major
systems developments with directly perceivable individual and societal benefits,
then we will have a prototype for a most excitingly attractive, even though
slightly dangerous, next decade or two of technological progress.