Wednesday, June 14, 2017

Reaping the whirlwind?

How many republican congressmen will need to be shot before we get meaningful gun control?

Saturday, June 10, 2017

God?

According to Everett's theory human intelligences are "complex internal processes" in Hilbert space*. (The Emergent Multiverse, D. Wallace, OUP, 2014, pg 100 ) Could god simply be an even bigger (or even an infinite) "complex internal process" in Hilbert space? And possibly entangled with our world? To believe in such an extraordinary idea would, of course, require extraordinary evidence. Its just a thought.

* "...one can regard the state functions themselves as the fundamental entities." "...all of physics is presumed to follow from this function alone." The Many-World's Interpretation of Quantum Mechanics, Hugh Everett III, pg 9

Thursday, June 1, 2017

A.s.a. H. as a prediction engine

Andy Clark believes that intelligence is all about the ability to make accurate predictions. (in Brockman, This Idea Must Die, Harper, 2015, pg 310) My artificial intelligence Asa H predicts what will be seen in the next time steps. (see, for example, my blog of 10 Feb. 2011) We can assess the accuracy of each such prediction by taking a dot product of the prediction vector and the next actual observed input vector. A running average of this prediction accuracy can then be kept and made an additional vector component of that case's overall vector utility. (see my blog of 19 Feb. 2011)

Wednesday, May 31, 2017

A.s.a.'s art

High up in the associative hierarchy, on long time scales, Asa's memory and extrapolation/interpolation generates stories, fiction, literature. Some of these patterns are identifiable as examples of Gozzi's 36 dramatic situations. (The Thirty-six Dramatic Situations, George Polti, Editor Co., 1917) (For example: pursuit, obtaining, daring enterprise, fatal imprudence, disaster, erroneous judgement.)  "Literature must be an analysis of experience and a synthesis of the findings into a unity." (Rebecca West, Ending in Earnest, Books for Libraries, 1967) This may tell us something of how and why art emerges from intelligent systems.

In early experiments I taught Asa Boolean logic operations. Could a vision system allow Asa to develop a spatial logic? (Handbook of Spatial Logics, Aiello et al, Springer, 2007) Would this result in an internal picture language (Picture Language Machines, Kaneff, Academic Press, 1970) with visual art developing higher up in the hierarchical memory?

Thursday, May 25, 2017

VEX IQ robotics platform

It may be possible to use the VEX IQ robotics construction system with my pain subsystem. (Depending upon how VEX IQ robots tend to fail.) Something like the VEX clawbot IQ might involve about one hundred pain sensors. A simpler design should be possible that would reduce this number.

Wednesday, May 24, 2017

Ground level

What is the best way to sense/define the ground/floor? An altimeter is easy but not very precise. I have placed force sensitive resistors on the foot pads of a walking Lego robot. These feel the push of the floor and the weight of the robot and any load it carries.

Education research

As a part of institutionalized learning assessment I am being asked to identify problems that the students have in my classes, measure their failure rates on these issues, find a fix, employ it, and then measure how well the fix has reduced failure rates.  To do this in a statistically sound double blind test would take years. We are being asked to cycle through all of the courses in 3 years.I do not wish to do physics education research. I have been critical of much of that work in the past and I do not see evidence that any silver bullets exist that might solve any of our problems in education.I work under capitalism. If someone really wants to improve education then they need to put some real money into the effort. I believe that fusion energy, artificial intelligence, a science of consciousness, and scientific pluralism are all topics that are more important than physics education research.

Tuesday, May 23, 2017

Qualia

Sensory receptor nerves respond to external stimuli like light and sound. These are exteroceptor  nerves. This defines what we mean by the concept "to feel." Sensory receptor nerves can also respond to internal stimuli like blood pressure. These are interoceptor nerves. Perhaps what we call qualia are the triggering of other interoceptor nerves, nerves that respond to various other types of internal stimuli. Again, they are what we "feel."The simplest example of this might just be the state of a hidden layer in a neural network feeding back to the input layer, i.e., an Elman network.

Legacy code

I have code and printout dating back into the 1970s. Most (all?) of the oldest electronic files will no longer run but I have hard copy of many programs. I have been unable to locate a couple of thick file folders of publications on case-based reasoning which also contained the first CBR programs I wrote. I suspect that these may have been discarded by mistake a few years back when there was a remodeling that tore the outer wall off the science hall (and my office). I've written too many hundreds of programs to want to spend the time needed to preserve them all.

Weltanschauung

I have been asked about my worldview. Many of my blogs on philosophy deal with my weltanschauung, among them my blogs of  1 sept. 2012, 17 aug. 2012, 1 Jan. 2012, 17 aug. 2013, 1 dec. 2014, 26 march 2014, 13 april 2015, 20 feb. 2015, 4 oct. 2016, 15 sept, 2016, 22 april 2016, 28 may 2016, 1 june 2016, 21 july 2016, and 7 Jan. 2017.

Friday, May 19, 2017

A.s.a. H., reflexes

In Asa H, as in humans, there are typically far more input signals received as compared to output signals generated. Output predictions/generation may then occur on higher levels in the memory/concept/abstraction hierarchy, i.e. "downward causation." What outputs are generated quickly, confined to the lower level(s) of the hierarchical memory, this constitutes "reflex."

Wednesday, May 17, 2017

Concepts

In my 1 October 2015 blog I said: "In some cases we would like to have additional definitions for a given concept." I now find that for most named concepts I would like to have multiple examples.

Friday, May 12, 2017

Robots

I'm working to simplify my Lego robot designs. This will allow for full coverage with the pain subsystem while reducing the total number of sensors required.

Thursday, May 11, 2017

Some of what's wrong in education rests with the students

1. Some students we get in physics courses should not be trying to do physics.
2. Most students should be trying to attend all of the scheduled classes.
3. Many of our students don't put enough time and effort into their classes.
4. Some students want ME to do work that THEY should be doing.
5. Students should not be expecting me to entertain them.
6. We should do things that are important because they are important, not because they are "fun."

 The paper in the April 2017 American Journal of Physics (volume 85, page 311) is a spoof but one day we may be able to objectively demonstrate that a given course TEACHES effectively* and when a given human student does poorly then it's THEIR fault.

* by having an AI take the course and learn the material successfully

Wednesday, May 10, 2017

Citation analysis criticized again

I was happy to see Jeremy Berg's criticism of impact factors and their use (http://blogs.sciencemag.org/sciencehound/ 15 August 2016 and his article in Science magazine). See my own criticism (my 4 April 2016 blog for instance).

Sunday, May 7, 2017

Improving simulations

I try to do everything I can with simulators since they are faster and cheaper than real mobile robots. Simulations of sensory input are not as good as the real thing, however, and at the moment I have only a crude simulation of the pain subsystem. As I collect examples of damage with real robots I expect to record enough cases so that a simulator can use case-based reasoning to accurately predict pain inputs for simulated robots. When a sufficiently similar event occurs in the simulation a case-based reasoner will predict and apply appropriate pain signals.

Tuesday, May 2, 2017

Reporting negative results

At a conference I attended recently I happened to see a poster describing possible means of producing some certain chemical compound. The author reported on 4 or 5 possible paths to synthesizing this compound, none of which were successful. I found this refreshing. Edison told us that its progress when you show that some scheme doesn't work the way you had thought it might.

Connection between attention and forgetting

I have discussed attention (and mechanisms for focus of attention) in a good number of these blogs. I have also employed forgetting in Asa H, the simplest sort of passive forgetting being the deleting  of cases which seldom reoccur and/or have low utility. Attention and forgetting together have an influence on the performance of Asa H in any given working environment. Perhaps I need better algorithms for forgetting.

Pursuing scientific pluralism

Scientific pluralism (R. Jones, Trans. Kansas Acad. Sci., 116, 2013, pg 78) , like "native" or "indigenous science," (Native Science, G. Cajete, Clear Light, 2016) is value-laden (R. Jones, Trans. Kansas Acad. Sci., 119(2), 2016, pg 249)  and encourages interdisciplinary (holistic) and applied research as well as reconceptualizations of reality (multiple points of view).

Friday, April 28, 2017

Crazy Donald

I see that psychiatric experts at a conference at Yale have warned that Trump is mentally ill, dangerous, and should be removed from office. (Conference chaired by Brandy Lee of the Yale Department of Psychiatry) Seems they are acting a bit late. (See my blog of 1 August 2016.) Better late than never? We'll see. The republicans have a lot to answer for. As do americans in general.

Thursday, April 27, 2017

A challenge to Bohr's doctrine of classical concepts

"...observations must always be expressed in common language supplemented with the terminology of classical physics." (N. Bohr, On the notions of causality and complementarity, Dialectica 2 1948, pg 313) "Every description of natural processes must be based on ideas which have been introduced and defined by the classical theory." (N. Bohr, Uber die anwendung der quantentheorie auf den atombau, Zeitschrift fur Physik 13 1923) Bohr argues that humans will have to make measurements using macroscopic meters for which classical physics concepts offer an adequate description. But Asa H is able to extrapolate and abstract away from the classical concepts that it has learned. It can do so on all levels of the concept hierarchy, even on the lowest level which has direct contact with sensory input and action.

Thursday, April 20, 2017

Make room for capitalism

In recent years I have considered myself an advocate for "councilism." But Asa H has suggested to me* that for the same reasons (see my blog of 17 Aug. 2012) that I follow scientific pluralism (see my blog of 26 Sept. 2010) I should also adopt capitalist models along with socialist ones, i.e. adopt an economic and political science pluralism. This is going to take some getting used to! And not all models are equally valid/useful.

* Asa's chain of thought was something like:
1.) A mixture of Asa agents, some having vector utility and some with scalar utility proved useful.
(I've seen that before and discussed it in presentations on Asa H. Agents with scalar utility can evolve faster for one thing.)
2.) Scalar utility use was represented in Asa H by the word/concept "value monism."
3.) In Asa H "Value monism" was associated with "money," or "currency," and "capitalism."
4.) Some use of "capitalism" was thus found to be helpful and was reported.

I'd like to co author a paper with Asa on some of the various ideas we have worked through together but I know that no one would let me publish it. I wonder when such a thing will finally become acceptable.

Monday, April 17, 2017

Robotics

Robots are expensive. To keep the cost down robots need to be reconfigurable. This is easy with Lego but is complicated by my pain system and wiring. Lego bricks are cheap enough that we don't have to reuse bricks every time. New bricks can be employed. Some pain sensors, however, are mounted on the brain brick, or servo, or sensors. This complicates reuse/reconfiguration.

Saturday, April 15, 2017

Pain

The signal NOT(pain) can be added as a third component of the vector utility measure on each individual level of the Asa H hierarchy. (See my blog of 19 Feb. 2011.)

The VEX platform, like Meccano, does not lend itself to use with the pain subsystem I've developed. (See my blog of 31 March 2016.)

Monday, April 10, 2017

Are there metaphysical laws?

As part of his constructor theory David Deutsch believes that "All laws of physics are expressible in terms of statements about which physical transformations are possible and which are impossible." Einstein, Bohr, and Heisenberg argued about what kind of physical laws were acceptable. The assertion that "everything has a cause" might be one candidate metaphysical law, a metaphysical law that quantum mechanics might cause us to reject. In Science Without Laws (Duke University Press, 2007) Creager, et al, consider the possibility that knowledge might be organized around exemplars or cases rather than around fundamental laws. Similarly, there might then be no metaphysical laws either.

Thursday, April 6, 2017

GPS

I have a  GPS sensor that I bought for use on Lego robots. Unfortunately it would not work indoors. I'm going to use it outdoors, running Asa H on my Microsoft Surface Pro 4. I have regularly used virtual GPS sensors in robot simulation experiments.

Wednesday, April 5, 2017

Another road not taken

In late 1971 or early 1972 I was talking to Bernard Rosen about doing a PhD thesis in computational physics. Although weather modeling was difficult professor Rosen felt that we just might be able to model the longer time scale climate and climate change. I decided instead to do plasma work. There were lots of interesting problems available to us in those days.

Robot pain sensors

It's difficult to maintain reliable pain sensors between moving parts. There are also a limited number of individual pain channels available. It's useful to run a robot for a while to determine where it will fail and then install pain sensors at those most problematic sites.

Sunday, April 2, 2017

Developing a deep space transport

The first manned flight of a deep space transport should not be in cislunar space, it should be in LEO where a rescue or resupply mission could be quick and "easy" if it were needed. In fact, the development of the various systems needed for such an "interplanetary cruise stage" have been a stated goal of various projects since the first Russian space stations.

Saturday, April 1, 2017

Hand coding Asa H

Newborn humans instinctively grasp, root, and suckle. Similarly, at least some of Asa H's core concepts may need to be hand coded. Chomsky thought that humans had an innate language instinct. Is there anything I should definitely avoid hand coding? (Lenat has been hand coding everything in CYC.)

Friday, March 31, 2017

Mobile robots

Asa H is currently operating a set of 3 Lego mobile robots; a steerable walker similar to Dawson et al's "strandbeest" and two robotic arms. If one of the arms has a fixed mount and the other is mobile then Asa learns "handedness" like most humans do. If both arms are mobile then Asa is "ambidextrous" and does not favor one hand/arm robot over the other. By using Lego we can quickly and easily modify the robots. Construction sets like Meccano don't lend themselves to the use of the pain system that I've developed. (See my blog of 31 March 2016.)

Thursday, March 30, 2017

Steerable Lego walkers

In order to physically ground Asa's concept of "walking" we gave it control of a Lego robot walking machine. To make this a useful agent we mount on the walker some of the devices/sensors that we wish to have mobile; cameras, solar batteries, etc. For this to be practical the walker needs to be able to turn. Dawson, et al's "strandbeest" (From Bricks to Brains, AU Press, 2010, pg 135) and Valk's "strider" (The Lego Mindstorms NXT 2.0 Discovery Book, No Starch, 2010, pg 121) are possible examples.

Tuesday, March 28, 2017

Oppositional concepts: alternative facts

The value of retaining and exploring oppositional concepts is discussed in books like Oppositional Concepts in Computational Intelligence, Tizhoosh and Ventresca, Springer, 2008. See also my blogs of 30 January 2017 and 5 July 2016. Oppositional concept pairs can be taken to be examples of legitimate alternative facts.  (In physics an especially important example might be the fact that a photon is a wave and the fact that a photon is a particle. Again, however, I do not wish to legitimize examples which are simply outright lying.) My artificial intelligence Asa H has made use of oppositional concepts.

Friday, March 24, 2017

Intelligent agent teams

Asa H specialist agents can be organized into teams. What teams should be formed? How should they be organized? Some can probably be identified from and operate like various human industries.

Monday, March 20, 2017

Dusty air

I have added a Sharp model GP2Y1010AU0F dust sensor to Asa's sensory array.

Friday, March 17, 2017

Progress

It has been difficult to give an objective definition of progress. (J. B. Bury, The Idea of Progress, Dover, 1955, L. Laudan, Progress and its Problems, Univ. Calif., 1977, M. H. Nitecki, Evolutionary Progress, Univ. Chicago, 1988) I have argued that science and intelligence are not value free. (R. Jones, Trans. Kan. Acad. Sci., 119, # 2, pg 249, 2016) Progress is then defined by continuing positive feedback observed over time.

Wednesday, March 15, 2017

Cluster concepts

Daniel Stoljar defines "cluster concepts" as "Concepts that can be associated with, and defined in terms of, a range or cluster of features." (Physicalism, Routledge, 2010, pg 233) Asa H's "vector concepts," or "cases" are then cluster concepts of a particular kind, organized hierarchically, and manipulated using vector mathematical methods.

Friday, March 10, 2017

(A certain amount of) Danger was good for you

In a benign environment an optimal survival machine might invest relatively few of its material and energy resources in thought and mind. In an extremely challenging environment there may be no opportunity to evolve a mind. The process may be just too costly to undertake. Only in an environment offering just the right mix of danger and resources will brain and mind evolve. The environment itself may then change of course. When a sea squirt adapts to its new living environment it absorbs its "brain" and gives up the (then) unnecessary act of "thinking." Intelligence is an adaptation to (an emergent specialization in) a particular range of environments, it is not always ideal/warranted.

Senses

Our senses give us a window on the world but they also limit what we see. For millions of years we hominids had no idea that there were such things as magnetic fields even though bacteria may sense and respond to them. (See Scientific American, December 1981, Blakemore and Frankel) We see that the sky is blue but that is as much a result of the limitations of our vision as it is the result of Rayleigh scattering. The eyes of birds do not let them see a blue sky. We may live today in a vast Hilbert or Fock space of which we experience only a tiny fraction. The concepts we create in order to describe the world depend in part on the array of senses we happen to have. This is one reason I have tried to give Asa H as wide a range of senses as possible.

Thursday, March 9, 2017

An interfacing concept

I am interested in the different ways in which Asa H conceptualizes reality. Because of its use controlling things like my plasma physics experiments Asa has formed a concept we might term "interfacing." Asa knows that it can be connected to and then take control of certain electrical and electromechanical devices/systems.  Humans with prostheses might learn a similar concept.

Tuesday, March 7, 2017

Asa's inner speech?

Asa H has been taught names for some of the concepts it learns. These can be externalized or spoken when they occur as predictions/outputs. (See, for example, chapter 1 of my book Twelve Papers available at www.robert-w-jones.com) When only weakly activated, or during extrapolation, or interpolation, or chaining, etc the words may not be externalized (output) but may still contribute to chaining, etc (learning, for instance). Is this Asa's version of "inner speech?" (See, for example, C. Fernyhough, The Voices Within, Basic Books, 2016)

Monday, March 6, 2017

Sense of taste

Asa H has pH and salinity sensors.  I have experimented with commercial blood glucose meters and sensors in an effort to give Asa a way to measure "sweetness."  Unfortunately these sensors can not be reused and do not seem to work over the range of concentrations I'd be interested in.

Sunday, March 5, 2017

Signs of old age

My artificial intelligence Asa H notices when erratic servo rotation occurs (due to worn and/or stripped gears) and unsteady signals are seen from such things as electrochemical sensors.  These give Asa a definition for (concept of) "old age."
I have some old BASIC stamp and Lego RCX robotics hardware but most now is Lego NXT or newer. With Asa we can hope to replace all of his aging components, unlike with humans. I hope that artificially conscious agents will suffer less than humans, not more. (Contrary to T. Metzinger's, The Ego Tunnel, Basic Books, 2010)

Friday, March 3, 2017

Sense of smell

I am adding a Vernier O2-BTA gas sensor in order to give Asa H a richer sense of smell.

Tuesday, February 28, 2017

Divided thought

We are all able to walk or drive while independently conversing or thinking or planning. Similar parallel processing occurs in Asa H. Input devices and output servos may or may not be shared by the two (or more) processing streams. In Asa, each stream may take the form of an independent concept hierarchy. Whether these hierarchies can evolve to the point of forming independent divided consciousness (Ernest Hilgard, Divided Consciousness, Wiley, 1977) or multiple personalities remains to be seen.

Friday, February 24, 2017

Observing Whorfian phenomena in Asa H

Language helps to tie some concepts together. Commands issued in english like "obstacle, slow down, turn left" help Asa H to form a link between the actions of slowing and turning, aiding in the creation of a sequence that might be called "dodging." Asa's description and understanding of the world is then written in terms of these very concepts. Language has come to influence how Asa will see its world. Just how strong and important an effect this is remains to be seen.

The emergence of consciousness

Higher level concepts emerge after lower level ones have developed. Concepts of "self" and "consciousness" evolve after/from categories like "sense" and "act." See my 1 January 2017 blog on Asa's expanding consciousness.

Thursday, February 23, 2017

The more general concepts

Asa H naturally learns features that co-occur and sequences in time. It quickly learns that servo position feedback, temperatures, acceleration, distance measures, battery charge, etc. form a common cluster which we may choose to call "senses," and loading a file is followed by searching the file and then interpolation and/or extrapolations, a sequence we may choose to label "thinking." Sometimes things like grasp, lift, and move may occur in a sequence and we can name that "act." Some more general concepts occur on higher levels in the memory hierarchy. A sequence of "sense" "think" "act" may define an early form of the "self" concept. If the sequence leads to injury then "pain" or "damage" or "health" is added to the "self" concept. Hand coding can be used to speed up learning of any desired concepts or to make Asa's concepts more closely resemble categories humans know and use.

Monday, February 20, 2017

Root cause

Climate change is just a modern example of the tragedy of the commons. A real solution must eliminate the root cause. The root cause is capitalism. Excessive competition gets in the way of needed cooperation. I've seen this happen time and time again.

Sunday, February 19, 2017

In black and white

I've been watching the movie Madiba on BET and the Chappelle re-run on Saturday Night Live. Mandela's story brought home to me the fact that you can't devote your life so completely to a cause and have a normal family life. The same thing is true when you devote yourself so extensively to science.

Chappelle made it clear to me that while a white American might well react to Trump's election like I did in my 9 November 2016 blog, for black Americans it was just same ol' same ol'. I do disagree a bit in that it seems to me HRC really won the election by almost 3 million votes. So Americans didn't vote as poorly as one might think. I also point out that giving a mad man control of nuclear weapons IS different from some other evils that America (the U.S. that is) has been responsible for in the past.

Saturday, February 18, 2017

Brain surgery

Hand coding, working from things like my blogs of 5 November 2015, 21 July 2016 and 1 January 2017 I am trying to clean up a set of standardized core concepts that would be shared by all subsequent Asa H agents, regardless of their specializations.

Wednesday, February 15, 2017

Discrete and continuous

Some aspects of reality appear to be discrete. Matter seems to come in chunks we call atoms. Charge seems to come in chunks. Other aspects of reality appear to be continuous. Distance/length in 3 dimensions, x, y, and z, seems to be continuous. Duration/time seems to be continuous.

Some of our artificial intelligence systems are purely discrete, things like logic programming software run on digital computers. Other approaches to AI can be continuous, things like the dynamic approach to cognition based on differential equations. And there are analog computers and control systems.

While modeling the world Asa H makes use of both discrete and continuous math. Asa can add new categories and attributes/features. It grows a graphical network of concepts. It can also (continuously) adjust the relative strengths of attributes, do interpolation and extrapolation, etc.

Thursday, February 9, 2017

Ongoing research into large knowledgebases

For the last 4 or 5 years I have been working to give my artificial intelligence A.s.a. H. as large a knowledgebase as possible. (Large knowledgebases are quite opaque. My blogs of 1 October and 5 November 2015 represent only some fraction of what's been learned, a fraction that can be easily interpreted and translated into english.) That has involved a sizable portion of some one hundred sensory input channels in addition to vision. (As with humans, visual input can easily dominate the data bandwidth. See my early blog of 13 June 2013.) This currently runs on a 4 computer cluster with over 5 terabytes of memory. The intention has been to try to get Asa to understand the real world and function in it. As we try to scale to larger size we'll see if we really have the attention problem solved.

Wednesday, February 8, 2017

Religious issues

ESU's president informed us last Friday that "islamophobic flyers" had been distributed on campus. I watched the nightly TV news out of Topeka to find out more. They talked about the postings and showed a bulletin board but blurred out the flyer itself. At faculty senate yesterday we passed a "resolution in support of ESU's international community" that denounced Trump's January 27 executive order as well as these flyers. I asked ESU's provost about the content of the flyers and was only told that they contained passages from the quran, presumably inflammatory passages. It's interesting that today I found a business card tacked to one of our physics department bulletin boards.  It had Hope Community church, Andover, Kansas printed on the back side and a bible passage from 1 Corinthians on the front.

I would think the intent of the flyers was to denounce one set of religious beliefs and believers. I would think the intent of the other was to encourage a different set of religious beliefs and believers.  I suspect the intent of most of our university administrators is mostly to support enrollment.

While I don't want to single out any one religion for special mention I do think, as an atheist, that pointing out the nonsense in any and all of the "holy" books is perfectly fine by me. Perhaps one should be careful to criticize them ALL each time?

Saturday, February 4, 2017

The need for and limits to authority

Ideally we would be able to test any and all of the equations/arguments that we adopt and employ.  In practice, however, we do not have the time, knowledge, or resources to do this so we must accept and use some knowledge provided to us by others, "authorities."  We can, however, limit this to things which are at least testable/reproducible in principle, rejecting religious claims and the like.

Thursday, February 2, 2017

What is now?

R. A. Muller's book Now (Norton, 2016) has made me think about the concept of "now."  For me the contents of short term memory define "now" and these contents can vary somewhat from one person to the next (as they do in standard relativity theory).

The "past" is then the contents of long term memory.

The "future" is any predictions. (Includes results of chaining, extrapolations, etc.)

Monday, January 30, 2017

Alternative facts

It really is possible for there to be alternative facts. One specialist intelligent agent can learn the proposition P while a different specialist agent, operating in the same environment, can learn NOT(P). See, for example, The Logic of Reliable Inquiry (K. T. Kelly, OUP, 1996, page 381-383) "minor differences in the order in which they receive the data may lead to different inductive conclusions in the short run. These distinct conclusions cause a divergence of meaning between the two..." "...there is no version of the truth that all users converge to." This happens with Asa H specialist agents. One can also see it happen in human society, between communists and fascists, for example. (See also my blog of 21 July 2016) This was not what happened the other day with Kellyanne Conway, however. She was simply lying.

Saturday, January 28, 2017

Quantum fields

I am trying to read Lancaster and Blundell's Quantum Field Theory for the Gifted Amateur (OUP, 2014). I would be willing to accept quantum fields as the ultimate reality* if only there weren't the issues raised by general relativity. (see, for example, Monton's Against 3N-Dimensional Space in The Wave Function, Ney and Albert, OUP, 2013)

*"Therefore even matter itself is an excitation of a quantum field and quantum fields become the fundamental objects which describe reality." Lancaster and Blundell, page 2.

Friday, January 27, 2017

Interspecies communication

All formalisms are idealizations.  Like any other formalism natural language provides only an approximate description of  the world. (See my blog of 24 Jan. 2017 and The Philosophy of Niels Bohr, H. J. Folse, North Holland, 1985) Having different senses and effectors Asa H will always have somewhat different concepts than humans have. (Even different humans have somewhat different concepts, one from another.) I suspect, then, that man-machine communication in natural language may never be as good (accurate) as is natural language communication between humans. The Turing test may be exactly the wrong way to try to judge intelligence. I believe that a lot of the current work on artificial intelligence may be over emphasizing "conversational intelligence."

There will also be an impact on the quality of machine reading systems.

Thursday, January 26, 2017

Memories

Humans may have sensory memory (of duration of a few millisec), short term memory (duration of at most a few tens of sec), intermediate-term memory (several hours), and long term memory (>= a few tens of min). Asa H has a different memory scale on each successive hierarchical level that it learns.

Tuesday, January 24, 2017

Private language versus communicable knowledge

The network of concepts that are recorded as Asa H's casebases constitute Asa's knowledge and private language.  Even if we were able to assign "names" ("words") to each of these concepts (see my blog of 30 October 2015) they will not be identical going from one Asa agent to another and they will change over time.  The situation becomes even more uncertain if we try to relate Asa's thoughts to those of humans.
If an Asa agent begins life tabula rasa it is possible to read into it casebases from some other (previously trained) agent. In this case the private language developed by the trained agent also serves as a language of communication.  If all agents begin life with the same few foundational concepts (casebases) it is again possible to read some newly learned concepts from one agent to another.  But once agents have become specialists their various concepts will differ, one from another, and they no longer share a common private language. 
By naming each of (many of) an agent's concepts one forms a communicable (shared) language. But somewhat different concepts, held by different agents, will all come to share the same name.  Inter agent communication and publicly shared knowledge will be imperfect. This is important in thinking about some of Wittgenstein's work and the limitations of language.
Languages need not be composed of names/words only. Mathematics often serves as a language. Again, concepts may change over time. An example might be the concept of  "algebraic ideal." Concepts can also be forgotten while new concepts are created. An example might be "cohomology."

Saturday, January 21, 2017

Space travel for artificial intelligences

I have considered sending the human genome from planetary system to planetary system without the actual transport of matter. (Trans. Kan. Acad. Sci. 118, 2015, page 145 and my blogs of 11 May 2012 and 5 July 2015 ) How might one do the same sort of thing for artificial intelligences? My artificial intelligences A.s.a. H. and A.v.a. Have usually been written in QB64, C++, and VBA. How should one write an artificial intelligence that you want to be run on some arbitrary unknown computing system? Or, must you send the design for the machine as well?

Thursday, January 19, 2017

Asa's imagination and creativity

Asa and other creativity machines (see R. Jones, Trans. Kan. Acad. Sci., vol. 102, 1999, pg. 32) generate new ideas in the ways described in my blogs of 12 Feb and 23 Oct 2015.  But part of the creative process is the subsequent evaluation of new ideas. As Asa continues to interact with its environment concepts (ideas) that are not useful are strongly modified or are deleted altogether.

Monday, January 16, 2017

Why don't the best people go into management?

They're paid well but administrators spend their time doing scheduling, organizing, report writing, running staff meetings, recruiting, interviewing, making personnel decisions, office politics, etc. Not work for a first rate mind. And so, by default, we get our Trumps.

Sunday, January 15, 2017

An agent can know too much

A large knowledgebase slows down search.  A specialist agent will be slowed down by having knowledge of matters outside of its specialty.

Friday, January 13, 2017

Asa's model of the world

Asa H begins to form its most primitive concepts as a phenomenalist. Each concept vector (or case) begins as a bundle of sensory inputs (or motor outputs) in space and time. But averaging, interpolation, extrapolation, and other learning algorithms modify each of Asa's mental concepts. Asa's ultimate objective (and that of science) is to bring compact order to an ever growing range of experiences.

Thursday, January 12, 2017

Asa agents of varied knowledge and intelligence

The wide range of occupations of which a human society is composed are not all intellectually demanding. Similarly, not all specialist Asa H agents in a society of artificial intelligences will be equally intelligent.

Saturday, January 7, 2017

Different conceptualizations of reality

Human's are justified in rejecting psychic phenomena and concepts like telepathy, telekinesis, and dowsing.  My artificial intelligence Asa H can detect buried metal (see blog of 1 Jan 2017), communicate via WiFi, sense its location using GPS, exert magnetic forces on distant objects, sense the earth's magnetic field and is justified in forming concepts based on these.

Truth is different for math, physics, chemistry, and engineering

(Refer back to my blog of 1 Jan 2017.) Mathematics, the science of patterns, places a strong emphasis on the truth component "deduction from assumptions" and a much weaker emphasis on the component "usefulness."  Engineering, on the other hand, places a much stronger emphasis on "usefulness." Physics and chemistry both place emphasis on the truth component "agreement with observation" but the cold fusion episode shows us that physics places more emphasis on theory than chemistry does while chemistry places more emphasis on experiment ("agreement with observation") than physics does. The vector concept of truth is a bit different in each field and for each individual scientist. A theoretical physicist's valuation differs from that of an experimental physicist.

Tuesday, January 3, 2017

The body in the mind

My artificial intelligence Asa H's model of the world/reality (see my blogs of 1 October and 5 November 2015) is roughly divided into a model of the environment and a model of the self. (There is some overlap.) I believe that the model of the self (see my blogs of 21 July 2016 and 1 January 2017 "expanding consciousness") is something like what Jackson had in mind in his book The Body In the Mind (U. Chicago Press, 1990). Experiments with Asa H allow us to study how this kind of thinking might operate.

Sunday, January 1, 2017

Merging multiple sensory modalities in deep machine learning

I am studying David Harwath and Jim Glass's paper from CSAIL, "Deep Multimodal Semantic Embeddings for Speech and Images."  I discussed similar work in my 2011 paper "Experiments with Asa H" but rather than speech and images being the sole modalities I allowed for input from a number of Asa's various senses. (The paper is available on my website www.robert-w-jones.com under "book", chapter 1, page 13.)

Such merging of multiple modalities may explain why it has not been possible to discover a formal grammar that covers human language.

Case patches and interpolation

Rule patches (ripple-down-rules) were developed as a way to improve the accuracy/performance of production systems. (See Compton and Jansen, Proc. 2nd Australian Joint A. I. Conference, 1988, pg 292). It's possible to do the same sort of thing with case-based reasoning systems, provided one has logged a sufficient amount of experience. A patching case base takes as input the new input vector along with the nearest matching retrieved case vector and provides as output a prediction of the error in the expected output. One can use the error to attempt to correct/improve the output. This simply amounts to interpolation or case adaption. It can be used in case-based reasoners as well as A.s.a. H.

Senses

I've given Asa H a metal detector wand as an additional sensory input.

Different kinds of truth and the need for scientific pluralism

I have argued that truth is a vector quantity. (See my review of  Theories of Truth: a Critical Introduction, by R. Kirkham, at amazon.com, 25 September 2008.) There are various kinds of truth and these are each components of the truth vector. Things like:
      Valid deduction from true assumptions
       Satisfying a definition
       Agreement with observation
       Usefulness
       Coherence, consistency
       Etc.
There are then different kinds of science making use of the various different ways of assessing truth (Along these lines see B. Latour, An Inquiry into Modes of Existence, Harvard U. Press, 2013). This is a further argument for the need for scientific pluralism.

Beauty

Humans use the value/concept "beauty" in both mate and scientific theory selection.  My artificial intelligence Asa H and most other AIs do not currently have a concept of beauty, see, for example, my blog of 21 September 2010. Dorner's Psi cognitive architecture does incorporate a notion of beauty, modeled as an emotion. (D. Dorner, Bauplan fur eine Seele, Reinbeck: Rowohlt, 1999, pg 373) Perhaps I should try to give Asa H a concept of beauty. I have Bach's microPsi 2 software running in my lab. (See my blog of 2 March 2015.) Several groups have assembled software modules that assess the beauty of humans.  Even if an AI does not need such a value for its decision making it might still prove useful for man-machine communication and understanding. (Another example of reconceptualizing reality.)

Doing science can make you more intelligent

Models of cognition involve a value function, a dopamine circuit in humans. (See, for example, the model in my blog of 1 Sept. 2012) The goal of any intelligence is to maximize rewards. How intelligent you are depends upon how good your value system is.  If you have bad values you make bad decisions and get fewer rewards.  Doing science promotes and develops some of the values which we then apply elsewhere in life, things like evidence-based belief.

Expanding consciousness

Using mobile robots, simulators, and hand coding the network fragment that constitutes Asa H's concept of self (see 21 July 2016 blog for a listing) has now been grown, adding the vectors: smell=(MQ-2 sensor, MQ-3, MQ-4, MQ-5, MQ-6, MQ-7, MQ-8, MQ-9, MQ-135), taste=(pH, salinity), touch=(contact1, contact2,..., force1, force2,..., whisker1, whisker2, ...), see=(camera1, camera2, ..., light sensor, color sensor, IR sensor, ...), temperature=(temperature1, temperature2, ...), pain=(pain1, pain2, ...), sense=(smell, taste, humidity, touch, see, hear, pain, B field, E field/charge, sense far, sense near, motion sensors, sense weight, compass heading, accelerometers, wind, temperature, air pressure, GPS positions, solar power), image manipulation=(shift y, shift x, rotatecw, rotateccw, scale +, scale -, reflect, invert), think=(sort, load, save, simulation, search, deduction, interpolation, extrapolation, update, image manipulation, mutate, sensitivity analysis, forgetting, compression), leave=(sense near, move, sense far), battery charge=(battery1 charge, battery2 charge,...), act=(grasp, release, kick, lift, lower, walk, move, carry, turn, approach, leave, push, pull, twist, recharge, taste, electromagnet, fan, messaging, telemessaging), health=(battery charge, pain, temperature, hardware diagnostic, software diagnostic). The development of hardware and software diagnostics is an ongoing project. All together Asa's concept of its self has roughly doubled in size.