Dear Mario Radovan

Perhaps you lack some appreciation for an "open" system. Achieve that appreciation and realize also that the AI robot can have different experiences from any human. (The exact content and in-effect the nature of experience is not what is programmed in AI; the AI robot has its own experiences, with only [effectively] qualitative parameters (LIKE those we ALL have), and otherwise just USING at-any-given-time limiting systems that simply are the same as a human.) Let me recommend for you: my Project, https://www.researchgate.net/project/Developing-a-Usable-Empirically-Based-Outline-of-Human-Behavior-for-FULL-Artificial-Intelligence-and-for-Psychology

AND especially http://mynichecomp.com/AImemory.txt and

http://mynichecomp.com/onmemory.txt

What it "boils down to" is that different experiences can yield different learning AND different DEVELOPMENT and thus yield very different results (likely in ways realistically bounded, but NOT controlled by the human makers). "Bounds" (boundaries) do not control or much limit experience and reactions, if done correctly. Certainly the possibilities exceed anything any of us can imagine, and if indeed possibilities are beyond what we can imagine, we cannot really make any clear pronouncements (like you did when you said: if a computer is better than Mozart, the programmer must have been better than Mozart too)(Oddly you seem not to realize that, by the same reasoning (according to you), you should say: any computer that can beat anyone at chess, must have had a programmer that can beat anyone at chess -- and we know this is not so.)

[ P.S. Philosophers over define and, thus, overly limit (creating limits that need not be there); they are simply putting themselves in a "box" AND certainly not a box we necessarily have or need for ourselves OR for what we create. ]

[From:

Messori, C. (2016) Intelligence vs Artificial Intelligence: The King is Naked, Open Access Library Journal, 3:e3115

https://www.scirp.org/Journal/PaperInformation.aspx?PaperID=72003 ]

60 years have passed since John McCarthy coined the term Artificial Intelligence and what we can say about the research in the AI field is that : is conducted by a range of scientists and technologists with varying perspectives, interests, and motivations. Scientists tend to be interested in understanding the underlying basis of intelligence and cognition, some with an emphasis on unraveling the mysteries of human thought and others examining intelligence more broadly. Engineering-oriented researchers, by contrast, are interested in building systems that behave intelligently. Some attempt to build systems using techniques analogous to those used by humans, whereas others apply a range of techniques adopted from fields such as information theory, electrical engineering, statistics, and pattern recognition. Those in the latter category often do not necessarily consider themselves AI researchers, but rather fall into a broader category of researchers interested in machine intelligence1.

Is there any kind of intelligence in AI?

In the best case, the performances of the so-called intelligent machines are and will remain a computational, artificial simulation of some abilities manifested by people affected by the savant syndrome2, or syndrome of the "idiot savant", where idiot is referred to a male (savant syndrome occurs approximately four times more frequently in males than females as does the incidence of autism) suffering from a series of more or less severe cognitive and mental disabilities, but that has one or more than one super-developed abilities, a jarring juxtaposition of ability and disability in the same person always combined with massive memory (typically eidetic memory). This is all that an intelligent machine can and will aspire to do: artificial simulation of the savant syndrome (ASSS). What the ASSS experts can do and will be able to do is to develop and get the most out of the super-memory (storing) and computational skills that qualify an intelligent machine, inventing new algorithmic systems, new software and new media devices for new applications (e.g., via emotion modelling the machine will be programmed and tuned on the affective history, memories and aspirations, disappointments and expectations, desires and frustrations of their owner, so to satisfy at best their wishes and demands).

Realize an intelligent machine equipped with consciousness? Ridiculous. What can be done, and that is already being done, is to acquire new persuasive and pervasive technological tools, including robotic and computer systems exploiting AI, to be used in the military and civilian field, as a means for expansion and control of the market dynamics, or as a deterrent for the resolution of conflicts related to it, both locally and globally, on a small and large scale.

1 Hughes, T. et Al. (1999) Developments in Artificial Intelligence. In: Funding a Revolution. Government Support for Computing Research, Computer Science and Telecommunications Board, National Research Council, Chapter 9, 198-225. http://www.nap.edu/read/6323/chapter/11#199

2 Savant syndrome is a rare but remarkable condition in which persons with some underlying developmental disability or other brain trauma or disorder have some extraordinary special abilities -“islands of genius”- which stand in stark contrast to overall limitations. Typically these special abilities are in music, art, mathematics or mechanical/visual-spatial areas. Whatever the special skill it is always combined with massive memory. Sometimes, but not always, the underlying disability is autism. Savant syndrome provides a unique window into the brain for the study not only of brain structure, savant skills and memory, but it also provides unique glimpses into intelligence, learning and creativity.

Brad,

"... the AI robot can have different experiences from any human."

In my view, the word "experience" does not apply to machines. They have no experience and they have no aims. People normally speak about machines as if they were sentient beings, which they are probably not. My question about AI was caused by one such highly anthropomorphic text which I read accidentally.

I do not criticize AI scientists (I made PhD in logic programming, and I teach computer networks). AI scientists have achieved excellent and very useful results; I only hold that the concept "artificial intelligence" is not appropriate because it creates an impression (in wide audience) that machines have (somehow) become more intelligent than people. They have beat the world chess champion! This is far more than I can do! In my texts, I speak about authentic intelligence (human) and about functional intelligence (machines). And I say that machines (lifeless entities) cannot have any authentic intelligence (because they know no aims), but in functional intelligence, they are getting far better than people are.

Dear Claudio Messori

You say: "Realize an intelligent machine equipped with consciousness? Ridiculous. What can be done, and that is already being done, ..."

I would say you simply do not have a reasonable functional definition of consciousness. I am biologically-phenomenologically oriented and HAVE NO PROBLEM HAVING A CLEAR, USEFUL CONCEPTUALIZATION OF CONSCIOUSNESS IN ANY GIVEN CIRCUMSTANCE (and this is all you need); my writings here on RG make that crystal clear and make it clear how such a definition suffices. Otherwise -- i.e. cross-circumstances -- consciousness will vary and be about impossible to usefully define. (The way many "philosophically-oriented" people like to think, seeking "general" important definitions, inevitably leads to CONFUSION.)

Biological intelligence IS good adaptive (open) functional intelligence. That is all.

The consciousness conceptualization I use is a consciousness that clearly an AI "robot" could have. To learn more check out my RG writings.

[ P.S. I have always had it as my "job" to better general psychology (esp. developmental and personality psychology). You may find it interesting to know, that seeking to do THAT, using the traditional (classic) definition of psychology (behavioral [pattern] response(s) to aspects of the environment) AND THAT IS ALL, _AND_ to do that with as much confidence in empirical bases as may well be warranted, you end up with the complete needed, intelligible, and functionally practical, biologically sensible, ultimate PROXIMATE concrete directly observable behavior patterns/environment foundations THAT ARE EXACTLY WHAT A PROGRAMMER WOULD NEED to "mechanize" (direct/control/orient/develop) a true AI machine. See: citations in my LAST POST in this tread. ]

Just as info.-processing theory led psychology, I will not be surprised to see AI lead new psychology, in the near future. The "right kind' of empirical" (true, real, concrete empirical bases anchoring everything) is what AI needs. Psychologists show less discipline in clearly relating to this (or requiring this) good empiricism.

[ It turns out, in psychology, beliefs (false, groundless, unproven pseudo-assumptions) out-weigh seeking the most empirical grounding that may well be possible; unless false 'assumptions' can be seen for what they are, and the consequences of them seen as skewing and limiting like they do, psychology is ridiculously "scattered", putting their faith in one ridiculous groundless model OR another -- creating the "Tower of Babel". This, starkly, is the situation, and NO ONE could show otherwise. ]

Here is a new saying for "today": IF YOU choose what to believe, that is believing and is NOT discovering and knowing and is, in fact, antithetical TO truly coming to KNOW. (Much related to this is the appreciation that you must do massive observations to confidently find the way to do the good, important inductive inference needed BEFORE any chance you have (and it may well be: piecemeal) to reasonably (and in a well-founded way) do any hypothetico-deductive reasoning.)

I should add that my theory of development and learning is the ONLY one to have everything clearly, reasonably founded and grounded , that is: key points in behavioral responding (key pivotal phenomenon) are ALWAYS directly founded and grounded, in directly observable behavior-patterns-and-environmental aspects that are PROXIMATE causes (necessarily present at all the first inceptions of the most-major qualitative shifts in behavior, especially those related to developing thought and higher thought); and, also there are clear signs of the associative learning that has existed before and now continues on after such shifts but now there is ALSO new associative learning due to the qualitative shifts (each of them, as they occur in ontogeny): behavior patterns and incipient understandings associated with the resultant behavioral products of key new perceptions -- again, these being the beginnings, i.e. the inception, of the development of "higher thought".

My theory is the only one to actually explain stages shifts and does so related to the key directly-observables, as indicated above. Other theories have unclear, overly "internal", quasi-explanations (wrongfully and needlessly/artificially lacking in external expressions and external referents *), resulting in "explanations" that are too unclear _and_ too indirect, at best. It is not hard to argue that other theories do not empirically explain the development of higher thought at all directly or in any reasonable empirical way .

The empiricism I provide shows major pivotal phenomenon of behavior and behavior change related to new key, important directly OBSERVED innately-guided responses to aspects of the then-present environment ("perceptual shifts"), and clearly, expressly involving PRESENT(at the key times) observable new behavior patterning for new responses and response change.

Thus my Ethogram Theory is the only one with some concrete present-at-the-KEY-times particulars FOR EVERYTHING. Only such concrete findings can be the basis for the programming of AI. The other developmental theories are too vague to "ground" (way too much being left to intuition, notions NOT close to directly testable hypotheses -- bad for science, as well as for AI).

* FOOTNOTE: These theorists never tried to look for certain empirical referents that would BE THERE, would be EXPECTED TO BE THERE, for any thorough-going empiricist who believes psychology (behavior patterns in response to environmental aspects) can explain everything ITSELF (in its evolving structure and functioning, and SHOWN (at KEY points) as directed by qualitatively new changes in OVERT behavior patterns) (i.e. believing there can be a science of BEHAVIOR per se, the way psychology has been classically defined).

For more on Ethogram Theory (and its justification and consequences) see the "book" and the major large papers under the Project, https://www.researchgate.net/project/Human-Ethology-and-Development-Ethogram-Theory .

For the catalog of all other related capacities and capabilities of the various Memories FOR AI (largely explained in relation to each other and indicating their relation to major shifts beginning cognitive developments), see: https://www.researchgate.net/project/Developing-a-Usable-Empirically-Based-Outline-of-Human-Behavior-for-FULL-Artificial-Intelligence-and-for-Psychology

I will be back on Monday and read the posts carefully.

This discussion is slightly too sophisticated for my humble abilities; anyway, my basic view looks like this.

Behaviour is not consciousness and it does not indicate that there is any consciousness behind the behaviour. The same holds for people and machines; but we consider ourselves conscious, and most of us hold that machines are not. Computers can behave like people, but this does not mean that they are aware of what they are doing. This is why I say that computers do not have authentic intelligence (they do not understand what they are doing), but only functional intelligence, which is produced and implemented by people. Functional intelligence can achieve impressive results, but it is not artificial; it is a product and an implementation of the human intelligence.

We do not know how/why consciousness emerges in human brain, and we do not know how to include the concept "consciousness" into scientific taxonomy (language) of forces, particles, etc. The knowledge about biology and physics of the brain does not tell us how a subjective state looks (feels) like. A physical (objective) theory that tells everything about the physical (or chemical) processes that take place in my brain, does still not manage to express my subjective feeling of tasting a lemon. (Einstein allegedly said that physics cannot express the taste of lemon, the "now", and similar subjective things.)

There is no indication that any computation could create ("induce") consciousness in/on any kind of lifeless hardware. Awareness and feelings are products of biological (or chemical) processes, not of computation, which is a syntactic (symbolic) process, not a biological (chemical or physical) process.

When we talk about artificial intelligence we are essentially talking about a huge range of approaches: symbolic AI at one end and connectionism at the other.

When it comes to symbolic AI perspective where the machines are closed, sensorless boxes it is true that they are no different than calculators or electric saw etc. However, when you add cameras and other types of sensors they start interacting with their environments.

Of course the underlying learning paradigms are still mostly based on either induction or deduction instead of the real gem of mind, abduction. But, I guess it would be fair to say that a robot walking around Mars on its own collecting some rocks and taking photos where it "sees" necessary can be regarded as a form of intelligent being.

"... a robot walking around Mars on its own collecting some rocks and taking photos where it "sees" necessary can be regarded as a form of intelligent being. "

Two problems:

(1) "... walking around Mars on its own ..."

Does a machine ever operate "on its own", or it operates in the way people programmed it to do.

(2) "... a form of intelligent being"

I would say "intelligent behaviour according to human criteria". Because lifeless machines have no criteria. There are no values (good & evil, intelligent & stupid, etc.) without feelings.

Artificial intelligence (AI) is an area of computer science that emphasizes the creation of intelligent machines that work and react like humans. Some of the activities computers with artificial intelligence are designed for include: Speech recognition

What I wanted to say is that "artificial intelligence" is not a good name. Because intelligence is not "artificial"; it is always human, but it may be implemented (made operational or functional) by means of machines.

http://www.leaderu.com/truth/2truth09.html

https://singularityhub.com/2018/09/19/thinking-like-a-human-what-it-means-to-give-ai-a-theory-of-mind/#sm.000u9bg6l18zmenwr3m2gg8abtxdq

https://www.bbvaopenmind.com/en/artificial-intelligence-hacks-the-human-brain/

https://universonline.nl/2018/03/26/artificial-intelligence-human-mind

Anyway, artificial intelligence is good when machines control other machines. But when AI wants to "help" me, it always produces a disaster! It makes me mad. I have been struggling for years to calm down a piece of "intelligent" tool which creates me all sort of problems. AI may soon transform people into psychotic idiots. It obviously has such effect on me.