You Can't Not Believe Everything You Read (1993), Daniel T. Gilbert, Romin W. Tafarodi, Patrick S. Malone. It's a paper written by psychologists, but touches on a lot of philosophy and neurosciences in it.
Let's start by the abstract:
Can people comprehend assertions without believing them? Descartes suggested that people can and should, whereas Spinoza suggested that people should but cannot.
Descartes and Spinoza
Okay, let's check. Descartes wrote Principles of Philosophy in 1644, and it contains the following texts:The seeker after truth must once in his lifetime doubt everything that he can doubt... We have free will, enabling us to avoid error by refusing to assent to anything doubtful... our mind is better known than our body. The knowledge of our mind is not simply prior to and (1) more certain than the knowledge of our body, but is also (2) more evident.We see the basic ideas of Descartes:
- Doubt everything, except the existence of your own mind (I think therefore I am).
- Doubting and finding the truth are possible, because of free will.
- The body and the mind are different, and the knowledge from the mind is more certain and obvious than knowledge from the body.
In contrast, there's Ethics (Geometrically Demonstrated) (1677), Spinoza:
In the mind there is no absolute or free will; but the mind is determined to wish this or that by a cause, and this last by another cause, and so on to infinity... in no case do we strive for, wish for... because we deem it to be good, but on the other hand we deem a thing to be good, because we strive for it, wish for it...I confess having little idea of what Spinoza's philosophy is like, so I turned to a digest:
Baruch Spinoza (1632-1677)... study of Descartes and Hobbes led his philosophical views away from orthodox Jewish philosophy; subsequently, he was excommunicated... [he wrote] his philosophy by the geometrical method of proof... When Spinoza was offered a teaching position at Heidelberg, he wrote, “I do not know how to teach philosophy without becoming a disturber of the peace.”
... The Ethics, a book published posthumously from the fear of persecution from... pantheism... [roughly, "God is the universe"] Spinoza believed... the excellent life is the life of reason in the service of one’s own being. The soul seeks knowledge as a good; indeed, the soul’s highest good is knowledge of God... the mind and the body are, in reality, only one thing but can be thought of in two different ways... human beings have no free will, and the world cannot be evil.So what I got from this summary about Spinoza's beliefs:
- Body and soul are two ways of thinking about the same thing: the human thing.
- There is no free will, the soul, being the same as the body, is just as unfree as the body (as Descartes argued, bodies are just "automata", robots. Spinoza said that souls are just "spiritual automata").
- God is everything (pantheism).
- The mathematical style is the best way to write philosophy (in your face, Sartre!).
The digest proceeded with some excerpts from Spinoza's Ethincs, which I will excerpt even further (meta-excerpting... now, if someone were to survey my blog during a big blog surveying, it'd be meta-meta-excerpting... big flea, small flea, etc...)
Thus... a drunken man believes that he utters from the free decision of his mind words which, when he is sober, he would willingly have withheld:... believe that they speak from the free decision of their mind, when they are in reality unable to restrain their impulse to talk.Here, Spinoza is giving obvious examples where people claim they are doing something because they freely decided to do so, when in fact they aren't.
... men believe themselves to be free, simply because they are conscious of their actions, and unconscious of the causes whereby those actions are determined; ... the dictates of the mind are but another name for the appetites... Everyone shapes his actions according to his emotion, those who are assailed by conflicting emotions know not what they wish; those who are not attacked by any emotion are readily swayed this way or that... a mental decision and a bodily appetite... are one and the same thing.People think they are free to act because of an ignorance: they know they did something, but not why they did it, so they say "it's my free will!". But really, what the mind decides to do is decided by the emotions. What we decide to do is what we want.
... it is not within the free power of the mind to remember or forget a thing at will. Therefore the freedom of the mind must in any case be limited... we dream that from the free decision of our mind we do something, which we should not dare to do when awake... the decision of the mind, which is believed to be free, is not distinguishable from the imagination or memory...Here again, Spinoza points out how the "freedom of the mind" is limited and inconsistent.
... and is nothing more than the affirmation, which an idea, by virtue of being an idea, necessarily involves...This is the part that we are looking for!! Spinoza here is saying clearly that, to even comprehend an idea, one must actually affirm it, simply because that's how ideas work.
Actually, Spinoza wrote about this at length in Ethics, Book 2, Proposition 49:
In the Mind, there is no volition, or affirmation and negation, except that which the idea involves insofar as it is an idea.But I can't understand the proof. If you want to try, try reading this exposition.
Spinoza's philosophy, applied to society
I found this interesting article about Spinoza, that talks about his philosophy as understood in society.
Descartes divided the mind up into... intellect and will. The intellect gathers up data from the world and presents the mind with various potential beliefs that it might endorse; the will then chooses which of them to endorse.
This is how normal people think the mind works, isn't it? You learn and you read and you get a lot of guesses about how the world is, but which one you end up believing is chosen by your free will.
This raises the question of the ‘ethics of belief,’ the title of an essay by the mathematician William K. Clifford, in which he argued that ‘it is wrong... to believe anything upon insufficient evidence.’ If people choose what they believe, we can ask when it is morally right or wrong for them to entertain certain beliefs [because "ought implies can"].
Clifford is amazing, I want to learn his geometric algebra! It will have to wait for another day though. Today is philosophy day.
Jean-Paul Sartre went as far as to suggest that we are always to blame for our own suffering, since however much the evidence might suggest that our situation is a miserable one, it is always our choice to believe that it really is so.Sartre believed that humans are hopelessly free, and indeed, every second of suffering is in some sense one's own choice, because one can always suicide. When people pretend they are not free, they are acting in "bad faith": "...human beings, under pressure from social forces, adopt false values and disown their innate freedom, hence acting inauthentically."
But that's not what Spinoza thought, as our above reading of Spinoza's Ethics already showed us.
Will and intellect, he argued, are one and the same; as soon as the intellect gathers up its data, the mind is thereby made up on what to believe... We think of ourselves as free only because we fail to consciously note all the inputs that set the direction of our thought.The practical consequence is that
When people disagree with us, there is no point in imploring them to ‘face facts.’ The problem is not that they are refusing to face the facts; it is that they haven’t been exposed to the same facts as us, at least not in the same way.And as a final note on Spinoza, I noticed how optimistic he was
The man who has properly understood that everything follows from the necessity of the divine nature, and comes to a pass accordingly to the eternal laws and rules of nature, will in truth, discover nothing which is worthy of hatred, laughter, or contempt, nor will he pity any one, but, so far as human virtue is able, he will endeavor to do well, as we say, and to rejoice.It's almost like Nietzsche's love of fate (amor fati):
If we affirm one single moment, we thus affirm not only ourselves but all existence... in this single moment of affirmation all eternity was called good, redeemed, justified, and affirmed.
You Believe Everything You Read, and Spinoza was right
Now that we are sufficiently (over-)educated in the philosophy, we turn to the paper You Can't Not Believe Everything You Read:Three experiments support the hypothesis that comprehension includes an initial belief in the information comprehended. Subjects were exposed to false information... both load and time pressure caused subjects to believe the false information and to use it in making consequential decisions... In Spinozan terms, both manipulations prevented subjects from "unbelieving" the false information they automatically believed during comprehension.Spinoza was right. Just understanding what's being written makes one believe it. To un-believe it, then, requires a separate effort. If the effort is frustrated, then the belief would stay believed.
Philosophical review
The beginning of the article lays out the philosophical background the same way we have done. First, Descartes's philosophy and its status as the foundation of science and democracy is summarized.... the injunction that one must control one's beliefs is predicated on the assumption that one can control one's beliefs. In some sense, the whole of modern science is based on the Cartesian assumption that people do not have to believe everything they read.
This capacity for skepticism is the heart not only of the scientific method, but of modern democracy as well. The First Amendment's guarantee of freedom of speech is also grounded in Cartesian psychology. John Stuart Mill argued that people can achieve true beliefs only when their society allows all ideas–true or false–to be expressed, examined, and debated.
First people comprehend a message, and then later they may accept it. Understanding and believing are today taken to be the separate and sequential operations.Then Spinoza's view is summarized:
... [Spinoza] argued that understanding and believing are merely two words for the same mental operation. Spinoza suggested that people believe every assertion they understand but quickly "unbelieve" those assertions that are found to be at odds with other established facts... They may indeed change their minds after accepting the assertions they comprehend, but they cannot stop their minds from being changed by contact with those assertions.
Previous experiments
Then, the paper proceeded to summarize previous researches that support Spinoza's theory, and noted their deficiencies.A variety of evidence suggests that people have a tendency to believe what they should not (see Gilbert, 1991, for a review):
- repeated exposure to assertions for which there is no evidence increases the likelihood that people will believe those assertions.
- once such beliefs are formed, people have considerable difficulty undoing them
In one experiment, subjects were asked to read statements about a "foreign language", like "A monishna is a star",
- under some circumstances people will believe assertions that are explicitly labeled as false.
After reading each assertion, subjects were sometimes told that the assertion was true or that it was false. On some trials, subjects were interrupted just a few milliseconds after being told of the assertion's veracity. At the end of the experiment, subjects were asked to recall whether each assertion had been labeled as true or as false.And true to the Spinozan theory,
the interruption (a) would prevent subjects from unbelieving... cause subjects to report that false assertions were true, but (b) would not cause subjects to report that true assertions were false. This asymmetry did, in fact, emerge...But the authors noted that this is not good enough, since beliefs about "monishna" don't have irl importance. What's really important is if the Spinozan theory also applies for beliefs that has irl importance, and the authors designed the experiments precisely to show that!
The three experiments
... we asked subjects in Experiment 1 to play the role of a trial judge and to make sentencing decisions about an ostensibly real criminal defendant. Subjects were given some information about the defendant that was known to be false and were occasionally interrupted [by a digit-search task] while reading it. We predicted that interruption would cause subjects to continue to believe the false information they accepted on comprehension and that these beliefs would exert a profound influence on their sentencing of the defendant.Then, the report detailed Experiments 1 and 2. Experiment 1 is as described, and Experiment 2 is similar. They then proceeded to Experiment 3:
Experiments 1 and 2 provide support for the Spinozan hypothesis: When people are prevented from unbelieving the assertions they comprehend... they did not merely recall that such assertions were said to be true, but they actually behaved as though they believed the assertions.In Experiment 3, they used time pressure instead of interruptions, and found that it
affected memory for the veracity of a false statement in much the same way that interruption did.
Discussion and implications
Both philosophers realized that achieving true beliefs required that one subvert the natural inclinations of one's own mind; for Descartes this subversion was proactive [reject before believing], whereas for Spinoza it was retroactive [unbelieve after believing]... We have performed a half dozen experiments to examine this notion, and each has provided support for Spinoza's retroactive account rather than Descartes's proactive account... It suggests that belief is first, easy, and inexorable, and that doubt is retroactive, difficult, and only occasionally successful.Great, but what does it mean? A huge lot. If Descartes's theory underlies modern science and democracy, if it's wrong, we would have to rethink them.
Free speech might not be worth protecting as much, in this view:
Information changes people even when they do not wish to be changed... Freedom of speech is based on the old dualist notion that mind and body are separate things [but] as science continues to make the case that memories cause physical changes, the distinction between mental violence, which is protected by law, and physical violence, which is illegal, is harder to understand...
... if one accepts this suggestion, then restrictions on speech may seem quite reasonable... As such, those who advocate the regulation of expression in modern society might appeal to the Spinozan perspective as a scientific justification for censorship.The authors cautioned against this view, though, since there are other ways than censorship:
People, then, do have the potential for resisting false ideas, but this potential can only be realized when the person has (a) logical ability, (b) correct information, and (c) motivation and cognitive resources... acquisition of logical skills and true beliefs is primarily a function of education and is therefore under the control of society, whereas motivation and cognitive capacity are either fixed or under the control of the individual.So, instead of censorship, educating people about how to unbelieve falsities, might be a better way to go. The author further noted that Miller's "free market of ideas" without censorship could guard against falsities:
.When the marketplace is underregulated, the bad ideas that are present may be embraced by an individual whose wrong-headed beliefs may eventually be corrected by his or her fellows. However, when the marketplace is overregulated, the good ideas that are absent will never be encountered... The social control of belief may well be a domain in which misses [false negative error] have irreparable consequences that false alarms [false positive error] do not.
Elephants Don't Play Chess
Elephants Don't Play Chess (1990), Rodney A. Brooks.
This essay was written back in the second AI winter, when it seemed that AI research was crawling slowly, with disappointing results.
Artificial Intelligence research has foundered in a sea of incrementalism. No one is quite sure where to go save improving on earlier demonstrations of techniques in symbolic manipulation of ungrounded representations.Incrementalism: trying to improve a tiny bit on the previous one, like coral worms building upon their previous generations' dead bodies, hoping one day they'll reach the skies above.
It definitely didn't work well, and Brooks thought the problem is because the philosophy they used was wrong: "symbolic manipulations" of "ungrounded representations", like a mathematician who is so lost in their own mind that the real world has ceased to exist for them.
How to fix this mess and make some actually good AI? Grounding in reality.
The traditional approach has emphasized the abstract manipulation of symbols... [We] emphasize ongoing physical interaction with the environment... We show how this methodology has recently had significant successes on a par with the most successful classical efforts.The philosophical foundation of classical AI is like that of Descartes's dualism of the body and the mind:
The symbol system hypothesis states that intelligence operates on a system of symbols. The implicit idea is that perception and motor interfaces are sets of symbols on which the central intelligence system operates. Thus, the central system, or reasoning engine, operates in a domain independent way on the symbols. Their meanings are unimportant to the reasoner, but the coherence of the complete process emerges when an observer of the system knows the groundings of the symbols within his or her own experience.Basically, imagine that you are trying to teach the robot to walk. The classical approach through the "symbol system hypothesis" is to first mathematically formalize "walking", for example, by abstracting the logical relations between feet, ground, leg muscles... then write a logical engine that can start from some axioms, and calculate the correct proof of a conclusion, such that, when interpreted correctly by a physical engine, is in fact the steps of a robot that's walking.
And then, after this robotic brain is done, it remains to make a robotic body that interacts with the real world, observes the real world, translates the real world into the symbolic language that the brain speaks, then listens to its symbolic answer, and translates it back to real world motions.
Well, no. This method turned out pretty crappy. The second AI winter was largely caused by the disappointment. So what about a change of view? Instead of doing things symbolically, just do it the mundane way! In the world, not out of the world!
Enter Nouvelle AI, the AI that's definitely not out of the world, but got their feet on the ground, quite literally:
Nouvelle AI is based on the physical grounding hypothesis. This hypothesis states that to build a system that is intelligent it is necessary to have its representations grounded in the physical world... the world is its own best model. It is always exactly up to date. It always contains every detail there is to be known. The trick is to sense it appropriately and often enough.Instead of trying to keep a copy, a model, of the real world inside your head, just let the real world be its own model, and interact with it directly! Brilliant save of computational effort.
In fact, it might not only be a trick to save computational effort. Maybe a physical grounding is necessary for intelligence:
[Hans P. Moravec, Locomotion, vision and intelligence argued that] mobility, acute vision and the ability to carry out survival related tasks in a dynamic environment provide a necessary basis for the development of true intelligence.This point strongly reminds me of this TED talk, which was quite entertaining!
The paper then went on with the concrete details of how they made nouvelle AI robots that behaved intelligently in the real world, crawling and running and walking, with barely a "brain" to speak of, but depend on clever and fast reactions that depend on their physical construction.
Instead explaining the details with words, I'll just show you the impressive results: The walking Genghis robot, behaving quite life-like, relying only on "12 motors, 12 force sensors, 6 pyroelectric sensors, 1inclinometer, and 2 whiskers".
Conclusion
So what does it all mean? Why the heck did I mention Descartes, Spinoza, and AI, in the same post?
Well, I'd rather discuss more, but you can see the similarity yourself, can't you? For Descartes, there's a clear separation between mind and body, understanding and believing, but for Spinoza, things are mixed together, understanding is believing, the mind is a part of the body...
And the nouvelle AI approach is very Spinozan, intelligence is based on a grounding in reality, it cannot function without being based in reality.
I'd write more, but my hands are so tired, so until next time. Bye.
No comments:
Post a Comment