=Paper= {{Paper |id=Vol-2287/paper34 |storemode=property |title=Strong and Weak AI: Deweyan Considerations |pdfUrl=https://ceur-ws.org/Vol-2287/paper34.pdf |volume=Vol-2287 |authors=Johnathan Charles Flowers |dblpUrl=https://dblp.org/rec/conf/aaaiss/Flowers19a }} ==Strong and Weak AI: Deweyan Considerations== https://ceur-ws.org/Vol-2287/paper34.pdf
        Strong and Weak AI: Deweyan Considerations

                                Johnathan Charles Flowers

                    Worcester State University, Worcester, Massachusetts

                              jflowers@worcester.edu



       Abstract. Work in artificial intelligence and machine consciousness is often dis-
       cussed using Searle’s (1980) distinction between Strong and Weak AI. Weak AI
       presents AI as a tool for solving problems, whereas Strong AI is the generation
       of an “actual” mind. This paper will reconsider the possibilities of Strong and
       Weak AI in the context of John Dewey’s naturalistic pragmatism to recast our
       understandings of the qualities of “weak” and “strong” AI, and ultimately present
       the two as in continuity with one another.


1      Weak AI and Strong AI.

The field of Artificial Intelligence (AI) research and Machine Consciousness (MC) re-
search has been dominated by computational and functional theories of mind since early
engagements with computer programs. Proponents of this theory have argued that the
relation between the brain and the body is identical in many respects to the relation
between the hardware and the software of AI and MC. As such, mental states, con-
sciousness included, become functional or computational states in view of their causal
relationships, thereby enabling the development of AI to act as an aid to explaining the
human mind Against this, Searle (1980) raises serious objections against the possibility
of AI, specifically the argument that “the implemented program, by itself, ins constitu-
tive of having a mind. The implemented program, by itself, guarantees mental life.”
(Searle 1997.)
   In raising this objection, Searle provides a thesis of “Strong AI,” which is the objec-
tive of contemporary AI and MC research; and “Weak AI,” which is an epiphenomenon
of ongoing research into AI and MC. On this distinction, Searle states:
          According to weak AI, the principal value of the computer in the study of the
          mind is that it gives us a very powerful tool. For example, it enables us to
          formulate and test hypotheses in a more rigorous and precise fashion. But ac-
          cording to strong AI, the computer is not merely a tool in the study of the
          mind: rather, the appropriately programmed computer really is a mind, in the
          sense that computers given the right programs can be literally said to under-
          stand and have other cognitive states. In strong AI, because the programmed
          computer has cognitive states, the programs are not mere tools that enable us
          to test psychological explanations; rather, the programs are themselves the ex-
          planations (Searle, 1980,p. 417)
2

   For Searle, Weak AI is the attempt at modelling the human mind in a similar way to
modelling weather conditions, climate change, or other natural phenomena. By exten-
sion, weak AI does not aim to reproduce, or produce a mind any more than a computer
model of an ongoing storm seeks to reproduce an actual storm. Indeed, “no one sup-
poses that a computer simulation of a storm will make us wet… Why on earth would
anyone in his right mind suppose a computer simulation of mental processes actually
had mental processes?” (Searle, 1980 p. 37-38). On this analogy, there is no storm pre-
sent in the model of the storm, just as there is no mind present in the model of the mind.
The mind-as-simulation is therefore useful for testing hypothesis and for application to
problems of human cognition, and not as evidence of a conscious machine.
   Strong AI, on the other hand, seeks to actually produce a mind, or an intelligence
which literally possesses and understands other mental states. This machine would be
said to have a mind, albeit one whose composition is grounded in programs running on
hardware, as opposed to emerging from the conditions of biology. More crucially,
Searle predicates the distinction between weak and strong AI on what has come to be
called phenomenal consciousness as distinct from functional consciousness. Loosely,
phenomenal consciousness refers to our first-person experience of the world through
our sense perceptions. Functional consciousness, on the other hand, refers to the ways
in which consciousness “helps us deal with novel or problematic situations for which
we have no automatized response.” (Franklin, 2003)
   Thus, we may understand the distinction between Strong and Weak AI as the dis-
tinction between a tool which can be applied to a situation, or serves to explain the
nature of human cognition, and the presence of a phenomenally aware cognition that
possesses and understands its own mental sates and subjective experiences. Moreover,
we may also treat Weak AI as operating purely in the realm of functional consciousness,
while Strong AI operates in the realm of phenomenal consciousness and thus can be
said to possess a mind.


2      Dewey’s Theory of Mind

Dewey’s theory of mind begins with the organism in the environment. An animate or-
ganism, as distinguished from an inanimate organism, is an organized pattern of behav-
ior that evidences some bias towards some states of equilibrium and not towards others.
In discussing the distinction between iron and an animate organism, Dewey states that
Iron, in interaction with water, “shows no bias in favor of remaining simple iron; it had
just as soon, so to speak, become iron-oxide.” (LW1 195) Should iron opt to remain
iron through modifying the conditions of its interaction with water, it would demon-
strate the basic qualities of an animate organism. Animate organisms, therefore, are
those organized patterns of behavior that evidence a selective bias in their interaction
with the environment.
    To further define the conditions that give rise to mind, Dewey developed the term
“psycho-physical.” A system or organized behavior becomes “psycho-physical” when
it engages in activity through its organized pattern of behavior to acquire from its envi-
ronment the means to satisfy its needs where the maintenance of its organized pattern
                                                                                         3

of behavior is concerned. Additionally, the animate organism, through its psycho-phys-
ical processes, uses the results of past activities to determine the satisfaction of future
needs from its environment through the manipulation of its current interactions with
that environment. Thus, “responses are not merely selective, but are discriminatory, on
behalf of some result rather than others. This discrimination is the essence of sensitiv-
ity.” (LW1 197) Sensitivity, for Dewey, is the basis of feeling, which itself is the aware-
ness of the useful and harmful elements of an environment as a culmination, or predic-
tor of future consequences. On this basis, for Dewey:
          Complex and active animals have, therefore, feelings which vary abundantly
          in quality, corresponding to distinctive directions and phases—initiating, me-
          diating, fulfilling or frustrating—of activities, bound up in distinctive connec-
          tions with environmental affairs. They have them, but they do not know they
          have them. Activity is psycho-physical, but not "mental," that is, not aware of
          meanings. As life is a character of events in a peculiar condition of organiza-
          tion, and "feeling" is a quality of life-forms marked by complexly mobile and
          discriminating responses, so "mind" is an added property assumed by a feeling
          creature, when it reaches that organized interaction with other living creatures
          which is language, communication. (LW1, 198)
   For Dewey, mind emerges when the psycho-physical processes that makeup the or-
ganized pattern of activity of an organism is implicated in a social context. A mind
emerges through communication with other minds, which enables the feelings engen-
dered through the psycho-physical processes to make sense as the immediate meaning
of things experienced directly. To this end, for Dewey, mind is not a special property
of the human organism, it emerges where ever there is organized communication such
that psycho-physical processes and sensations can be treated as the meanings of inter-
actions with an environment. These meanings are apprehended and discriminated
within the total context of the organism within its environment and its situation as the
means whereby an organism identifies the traits of objects.
   However, it must be noted that, for Dewey, mind emerges within situations and as
situations are “minded.” More specifically, mind is an event that emerges through the
bodily engagement with environing conditions and is, therefore, continuous with the
organism as a distinctive pattern of activity, which itself is continuous with the envi-
ronment, and not a distinct entity from the biological processes or organic conditions
that give rise to the mind, which enables the organism to articulate the different quali-
tative, consummatory, ways in which situations are minded. Thus, For Dewey mind
emerges through the implication of an organism’s phenomenal consciousness in a so-
cial context whereby it may communicate the meaning of interactions with the envi-
ronment, interactions which include but are not limited to the sense perceptions, for the
purposes of future engagement with the world.
4


3      Dewey’s Theory of Consciousness

Consciousness, for Dewey, is always within a situation. As such, mind is not a distinct
cognitive function of an organism, but a function continuous with consciousness as
situations are felt or minded through interaction with the environment. Moreover, as
consciousness is like mind, another phase of an organism’s experience within a situa-
tion, understanding consciousness requires understanding Dewey’s concept of the sit-
uation. Defined by Thomas Alexander:
          Situations are integrated and organized by a pervasive quality (in human ex-
          perience) or undergone immediacy that is not cognized but which makes cog-
          nition possible; it is the tacit, mutual involvement of conditions of undergoing.
          In human existence, it is our established, prereflective, qualitatively “had”
          world that give sense to specific actions, including inquiry, speech, thought,
          affection… situations have an indefinite “horizon” with a defining pervasive
          quality; they also have a “focus,” a vortex of transformation which manifests
          itself in human experience as the “tensive” or “problematic.” (Alexander
          2013)
The pervasive quality articulated above refers to the directly apprehended meaning of
the situation through tension between doing and undergoing within experience. It is,
for Dewey, what enables an organism to distinguish one event from another. In human
terms, the qualitative unity of a situation is what enables human organisms to distin-
guish between similar situations. This prereflective horizon forms the basis for our
“sense” of the world; in the “tensive” or “problematic,” it serves as the basis for con-
sciousness, which emerges out of a functional need to reconstruct the situation in the
midst of the tensive or the problematic elements of a situation. More clearly, to be con-
scious of a quality of a situation is to be conscious of the ways in which a situation
moves from a state of precarity, or disequilibrium, to stability, or equilibrium: it is to
be able to reconstruct the situation in terms of its dramatic alignment to better under-
stand the result of the situation for future activity.
   As the disruption of the equilibrium between the organism and its environment,
which initiates an outreach into the environment, is itself a problematic situation which
necessitates the initiation of a process of inquiry to determine how to best restructure
the relationships adopted between the organism and the environment, “Dewey regards
consciousness itself as having emerged from the tensive relationships organisms have
with their environments; consciousness was the focus in experience through which the
organism strove to reorganize or “reconstruct” the situation. Consciousness arises from
fulfilling a functional need; it is not a pure given.” (Alexander 1988) Functional, here,
should not be taken to mean simply the generation of objects of knowledge, rather, it
should be taken in the context of a movement towards equilibrium following the dis-
ruption of the equilibrium between an organism and its environment, either internal or
external. The “functional need” is therefore experimental and ongoing as well as di-
rected towards an end in view. Consciousness, therefore, is the phase in experience in
which the organism reorganizes experience into a qualitative whole following an en-
counter with the problematic.
                                                                                          5

   In keeping with the above, “consciousness itself is but the tensive nexus of a situa-
tion, arising originally as a means of helping the organism interact and organize its
interactions with the environment.” (Alexander 1988) Consciousness’ functional need,
as articulated above, therefore arises as a consequence of our ongoing interaction with
the environment: without consciousness, specifically of the relationship between our
actions and their consequences in experience, inquiry and rational thought would not
be possible. However, this is not to say that the “functional need” that gives rise to
consciousness is solely limited to the objects of rational thought and “rationality:” con-
sciousness is also of the pre-reflective, qualitative horizon, and incorporates this quali-
tative unity in its reconstruction of the situation as what gives the situation its
“aboutness.” Thus, to be conscious is to be conscious of something, to be conscious of
the interactions and relations an organism adopts in the world.


4      Dewey’s Theory of Intelligence

Intelligence, for Dewey, is the ability to see the actual in light of the possible. This is
also the definition that Dewey provides for imagination. Imagination and intellect arise
as part of an ongoing action within a situation. Specifically, according to Alexander:
           It arises in an ongoing activity already structured by the fundamental narrativ-
           ity of any act (that of having a beginning, middle, and end); it also arises in
           consciousness as a crisis of that activity, carrying within itself the contradic-
           tion between what is and what ought to be; i.e., between actuality and possi-
           bility, necessity and contingency. (Alexander 2015)
    Intelligence and imagination, therefore, arise in continuity with consciousness as
consciousness seeks to reconstruct a situation. The moment of imaginative arising, for
Dewey, contains within it the tensive or problematic structure of a situation, and is part
of the driving need for consciousness to reconstruct the situation. Imagination, in this
context, is the projected completion of action which enables us to perceive the actuality
of the situation in light of the possibilities of that situation in an experimental way. Put
another way, imagination and intellect afford the possible meanings of the outcome of
a situation, which serve to narrow the focus of consciousness as it reconstructs a situa-
tion.
    To this end, Imagination is continuous with consciousness, and serves to present to
consciousness the possibilities for multiple meanings of a situation. Meaning, here,
should be understood as the total effect of a give resolution of the situation on all of the
relations that make up the organized processes of an organism’s behavior. Through
imagination, we can predict the consequences of an action to be tried, a conclusion to
be reached, as we engage in action to reconstruct the situation. Put simply, imagination
is crucial to the ability for consciousness to reconstruct a situation such that actions can
be taken within the environment.
    Imagination, like consciousness, therefore occurs within a situation: the organism
does not exit a situation except through taking action (Alexander 2015). When a course
of action is decided upon and the organism disposed to activity within the situation, the
6

situation itself may reach a consummation and thereby become part of the “natural his-
tory” of the organism from which it draws to project future action. In this way, imagi-
nation, together with consciousness enables an organism to have an experience of its
world as meaningfully apprehended and not merely bare sensation.


5      Implications for Weak and Strong AI

From a Deweyan perspective, it is possible for a Weak AI to possess a mind if that AI
is provided the means whereby it can symbolize the feelings that it has in response to
interactions with the environment. In this context, “feeling” need not correspond with
a human correlate as indicated by Dewey’s commentary about animals: a “feeling” for
a Weak AI may be articulated as sensory inputs or changes in the ways it interacts with
a digital or physical environment. Digital environments are included on this view as,
for Dewey, the environment extends beyond the “natural” environment and into the
“worlds” that organisms interact with. For a Weak AI, such a world might be organized
sets of data, or defined by the limitations of its input apparatus. To this end, the Weak
AI might possess “feeling,” but it would not know it is having “feelings” until it was
implicated in a symbol system that enabled it to understand the feeling as the meaning
of an interaction with an environment.
   Moreover, the above requires a redefinition of what is meant by “need.” While
Dewey presents the examples of food, sex, and shelter as “needs” which are the result
of the organization of the psycho-physical processes; a Weak AI or a Strong AI might
consequently have distinct, but analogous “needs” depending upon its embodiment. As
a pat example, a Weak AI may characterize processing power, electricity, or even in-
formation as “needs” in similar ways that a plant might characterize sunlight, water,
and nutrient rich soil. Like the plant, an AI might initiate interactions with its environ-
ment to satisfy this need in order to maintain, or renew, its equilibrium with its envi-
ronment. On this basis, it is possible to hypothesize an animalistic Weak AI embodied
in a variety of functional forms that seeks to alter its relationship with the world in order
to satisfy a need. As an example, a Weak AI driven solar farm could interpret a func-
tional “need” to reposition its solar panels to maximize the collection of sunlight. In
this context, the AI would perceive the limited collection as a “need” to be fulfilled
through outreach in the world.
   In contrast, a Strong AI would not merely possess “mind,” but “consciousness,” and
“imagination.” To be clear, in making this claim, it is not the case that the mind, con-
sciousness, and imagination of a Strong AI, even one patterned on a human mind, would
respond in ways that parallel or are intelligible by humanity. A Strong AI, as a unique
organism, a creative response to nature which actualizes one possibility of nature,
would respond in ways that are the outcome of its natural history. As the natural history,
and embodiment of a Strong AI are fundamentally distinct from the human context,
recognizing a Strong AI as conscious could not be done simply on the basis of a human
analogue.
   To this end, a Strong AI would operate on the basis of not merely “feeling” but
imagination, mind, and consciousness. At ground level, the Strong and Weak AI both
                                                                                             7

would possess feelings grounded in their interactions with the world, however, the
Strong AI would be able to not only symbolize these feelings as the meanings of an
interaction with the world, but it would be able to creatively reconstruct situations to
preserve or expand a given equilibrium with the environment. The crucial distinction
here is that a Strong AI would know it had feelings, and thus would be conscious of its
situation; whereas the Weak AI would not know the meaning of the feelings experi-
enced in interaction with the environment. However, it must be restated: it may not be
the case that the Strong AI would symbolize its consciousness in ways that were intel-
ligible to humans.
   Strong AI, therefore, must be treated as an organism unique in organization. While
a Strong AI would be able to reconstruct situations in line with its perception of the
qualitative unity of that situation, the “sense” of the world that would enable it to or-
ganize its interactions with its environment beyond mere fulfillment of a need would
be fundamentally alien to human cognition. As such, a Strong AI would be able to
respond creatively to the disruption of the equilibrium between the AI and its environ-
ment in ways that we may be unable to conceive or predict given the distinct af-
fordances of the AI. This creative response would subsequently involve the imaginative
projection of the possible meanings of a situation as it seeks to select from the multiple
meanings immanent within a given situation, and thereby result in novel responses to a
situation, and not merely automatic responses.
   However, in presenting the possibility of Deweyan consciousness in Strong AI, it
must be made clear that such a consciousness should not be judged according to human
analogues. As an organism’s interaction with its environment is determined through its
embodiment and the organization of its environment, and the embodiment of Strong AI
either in a digital environment or some mode of chassis is fundamentally different than
human embodiment, it must be restated that any Strong AI that is conscious in a Dew-
eyan mode would be tantamount to an alien consciousness. It is possible, however un-
likely, that humans would be unable to recognize such a consciousness when it
emerged.

References

 1. Alexander, Thomas M. The Horizons of Feeling: John Dewey's Theory of art, Experience,
    and Nature. Albany: State University of New York Press, 1987.
 2. Alexander, Thomas M.. The Human Eros: Eco-ontology and the Aesthetics of Existence.
    New York: Fordham University Press, 2013.
 3. Dewey, John. The Later Works of John Dewey 1925–1953 Volume 1: 1925 Experience and
    nature. Carbondale: Southern Illinois University Press, 1981
 4. Franklin, S. (2003). IDA: A conscious artifact. In O. Holland (Ed.), Machine consciousness.
    Exeter: Imprint Academic.
 5. Searle, J. R. (1980). Minds, brains and programs. Behavioral and Brain Sciences, 3, 417–
    457.
 6. Searle, J. R. (1997). The mystery of consciousness. London: Granta Books.
 7. Searle, J. R. (2002). Why i am not a property dualist. Journal of Consciousness Studies,
    9(12), 57–64.