2

Some classification schemes don't list logic as a separate branch of philosophy. I assume they regard logic as a component of epistemology.

Of course, others regard logic as a separate branch. The following discussion touches on some of the differences between epistemology and logic:

Does epistemology = logic?

It occurred to me that computers use logic, but they aren't associated with "intelligence." Of course, artificial intelligence could change that. But, for the time being, most computers operate on logic, requiring human operators to input intelligence.

On the other hand, computers can also function as portable libraries, holding vast stores of knowledge.

Have any philosophers or doctrines examined the relationship between computers and epistemology? Or, to put it another way, what are the differences between computers and humans in the realm of epistemology?

  • 1
    Computers flip bits and humans supply the meaning. Isn't that Searle's point in the Chinese Room? When you say humans supply intelligence, that's tricky ... intelligence and meaning aren't quite the same thing. Maybe that's the question. – user4894 Dec 7 '19 at 7:51
  • 1
    The older sense of "logic" (used from Plato to Kant, Hegel and Husserl) was closer to what is now called "epistemology" than to what is now narrowly called "logic", see What are the differences between philosophies presupposing one Logic versus many logics? Similarly, in the traditional sense of "knowledge" (justified true belief), computers do not have any. The question is partly addressed by how it has to be changed to drop "belief", see Do machine learning algorithms have knowledge? – Conifold Dec 7 '19 at 9:31
  • It doesn't seem a difficult issue. Computers cannot know anything while humans can. If epistemology is the study of how we know things and what we know then it has nothing to do with computers. It is not logic that allows us to know things, but the ability to know things that comes with consciousness, or simply is consciousness. . – PeterJ Dec 7 '19 at 14:47
  • 1
    @PeterJ, my computer knows where I've been for the last 10 years. – Hierarchist Dec 7 '19 at 15:04
  • 1
    @gnasher729, pete is trying to equate shoes to computers to show that computers don't know where I've been. Of course, pete committing the false equivalence fallacy because shoes are not computers. – Hierarchist Dec 9 '19 at 15:20
2

You speak of AI as if it requires human-level intelligence to raise epistemological questions when the current technology level of robots and machine learning has already raised them. Some philosophers have already embraced seeing the current state of technology as having epistemological implications. In their article Epistemology and artificial intelligence, Wheeler and Pereira claim that metaphysical implications of existing narrow AI raise serious philosophical questions:

In this essay we advance the view that analytical epistemology and artificial intelligence are complementary disciplines. Both fields study epistemic relations, but whereas artificial intelligence approaches this subject from the perspective of understanding formal and computational properties of frameworks purporting to model some epistemic relation or other, traditional epistemology approaches the subject from the perspective of understanding the properties of epistemic relations in terms of their conceptual properties. We argue that these two practices should not be conducted in isolation.

The question of why some people might be dismissive of computers as mere bit-twiddlers may demonstrate presumptions about the state of computers capacity to simulate meaning and intelligence. It is presupposed that since computers don't understand symbols (which is even an open question given the arguble notion of "understanding"), that they simply aren't intelligent in their current form. Intuitively, since they can out-think chess masters and outperform diagnosticians such as medical doctors, it certainly begs the question of why we can be certain they can't eventually manifest any intelligence, commonsense, emotional, or otherwise. Certainly, eliminative materialists such as Dennett have raised objections that claims of the uniqueness of human intelligence based on notions of belief or desire may not even be meaningful.

Additionally, cognitive scientists such as Ray Jackendoff have proposed empirically testable hypotheses about the origins of reference, truth, and meaning that have confirmed that the failure of the symbolist camp in producing human-level AI stems from the fact that human representation and meaning itself generally non-symbolic. On page 423 of his Foundations of Language, he states:

.. it has been important... to abandon the idea that linguistic entities in the brain are symbols or representations. We have instead been able to treat them simply as structures built of discrete combinatorial units.

This, of course, validates the long-standing convictions of the connectionist camp.

The question of whether our computers know and are intelligent is open to exploration philosophically along the lines of the philosophical question of intentionality. This is a difficult issue because of varying and imprecise definitions of knowledge (is it justified true belief?), intelligence (what does it mean to say a person is intelligent?), and intentionality (don't computers have the ability to use non-symbolic means such as sensors to generate percepts to represent the state of affairs?). From the SEP article on intentionality:

In philosophy, intentionality is the power of minds and mental states to be about, to represent, or to stand for, things, properties and states of affairs. To say of an individual’s mental states that they have intentionality is to say that they are mental representations or that they have contents.

This question about intentionality is complicated by the fact that it is still not even clear scientifically how people possess intentionality.

Another set of philosophical questions revolve around the symbol grounding problem. It's hard to compare a computer systems' ability to have meaning or knowledge, when the same problem is controversial for people. To dismiss out of hand that computers have no capacity to have knowledge when one cannot explain how people have knowledge may be egocentrism. Is there really something special about human consciousness or intentionality that cannot be recreated by a functional equivalent? It's very much an open philosophical controversy in the philosophy of mind what constitutes consciousness and what varieties there might be. John Searle's claim to fame with his various Chinese Room arguments may be that he has been the most rebutted philosopher in the field. These are all recent extensions to the long-standing epistemological question regarding the certainty that even other people have minds.

Starting in the early days of electronic computers at Caltech in the late 1940's, into the mid and late 1950's at places like Dartmouth or the National Physical laboratory in the UK, there arose an optimism that computers would begin approaching human-level intelligence quickly. Ten years later Hubert Dreyfus working for RAND wrote Alchemy and Artificial Intelligence and deflated optimism that has lasted until only recently where a growing body of AI researchers have been arguing again that artificial general intelligence is indeed possible:

Some authorities emphasize a distinction between strong AI and applied AI (also called narrow AI or weak AI): the use of software to study or accomplish specific problem solving or reasoning tasks. Weak AI, in contrast to strong AI, does not attempt to perform the full range of human cognitive abilities... As of 2017, over forty organizations were doing research on AGI.

Certainly, many AI experts have accepted that computers like desktop PCs, even if capable of knowledge representation, generally lack understanding. Poole and Mackworth from their Artificial Intelligence: Foundations of Computational Agents say on page 179:

It is very important to understand that until we consider computers with perception and the ability to act in the world, the computer does not know the meaning of the symbols. It is the human that gives the symbols meaning. However... it can draw conclusions that are true in the world.

Note the phrase "until we consider perception and the ability to act in the world". Are you really so sure Boston Dynamics' Atlas doesn't manifest intentionality? Early in their text on page 10:

The science of AI could be described as "synthetic psychology," "experimental philosophy," or "computational epistemology"... Modern computers provide a way to construct the models about which philosophers have only been able to theorize.

The idea that current computer architectures are just dumb terminals, seems to parallel the heavily debunked Cartesian proposition that animals are just automatons because they don't have a soul. Descartes appears to be wrong on animals as well as the mind-body duality. If you're interested in the current state of philosophical inquiry into computers, do a little more digging at https://philpapers.org/browse/philosophy-of-artificial-intelligence!

The advances of AI and those of cognitive linguistics which attacks the computational origins of meaning combined raise serious questions about the feasibility of computers manifesting intentionality and epistemic capacity.

SEE ALSO

Do machine learning algorithms have knowledge (if not justified true beliefs)?

|improve this answer|||||
  • 1
    I've upvoted this because I think it's a serious attempt to address the question. It seems absolutely right that the field has considered alternative formulations of intentional content and how these theories might pose a connection between symbol manipulation and knowledge acquisition. It's perhaps trying to do too much in a short character space, but that's the nature of the question, I guess. – Paul Ross Dec 9 '19 at 15:25
  • I appreciate the feedback. The nature of the question seems to pose a dilemma by presuming that the current state of artificial intelligence isn't enough to contribute to epistemological deliberations when there's clearly a small, but active body of philosophers working to advance the field, and then tackle a metaphysicality of intentionality and intelligence itself. I see philosophical prejudice against inquiring after the general nature of the intentionality of computers primarily rooted in the lay definition of computers, which are often equated as "mere tools". Thanks for your upvote. – J D Dec 9 '19 at 16:05

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.