Understanding "Talking Lions" and Other "Black Boxes"

If a lion could speak, we could not understand him. -- Wittgenstein[1] Artificial intelligence is everywhere. But before scientists trust it, they first need to understand how machines learn. -- D Castelvecchi, Can We Open the Black Box of AI? [2] "With artificial intelligence, we are summoning the demon." -- Elon Musk (2014) Reported in CNN Tech , Oct 28th.
	Faust with Homunculus

Black Boxes Abound. Less so, our understandings of them. The "black boxes" are not only lions, or AI devices, but also humans, and even more. If we generalize, keeping in consideration our perceptual limitations, any source of emissions that appear to us to be more than random, from any object or even an "empty" focus of attention, is, initially, a black box.

Many commentators have remarked that Wittgenstein makes an incautious jump from speaking to understanding in the quote cited above. I will generalize their concerns: how do we know that a (what appears to us to be) non-random emission of, say, sound, is speech? Or, if the transmission is not sonic to humans, language? Animals can learn to mimic human emissions. Even non-English recordings, played backwards, can sound like English speech.

To many a monolingual English speaker, the following phrases sound like pieces of English nursery rhymes colored with a non-English accent:

1. French - Un petit d'un petit s'étonnent aux Halles (Humpty Dumpty sat on a wall.)
2. German - Oh wer, oh wer ist Mai Lido doch Gong? (Oh where, oh where is my little dog gone?)
3. Spanish (Caribbean) -- Grima! Sí, comí. Te excusé que rifa. (Christmas is coming, the goose is getting fat.) [3]

Black Boxes are "Understandings." How does Wittgenstein even judge what the lion is doing? What assumptions is he making? Even if the lion's emissions sound like an English sentence, what is relevant in the context? What is the lion doing? Suppose, for example, our friend, Harry, is standing before a lion's cage in which a cave has been constructed as a den. The lion, staring at Harry, emits the sound sequence, /2aym+ gowing+3 hówm1/ [4]

It sounds like the lion just said, "I'm going home." But did it "say" that? Suppose the lion then lies down and rolls over on its back. It starts to snore. Are we still inclined to think that the lion talked? (If so, did he talk to Harry? Did the lion inform Harry it was "going home?" And, thus, was the lion telling a lie?) Can Harry believe his eyes and ears? Has he jumped to conclusions? (See Artificial Intelligence Weirdness. Need categorizing relate to visual cues?)

The basic problem is how we distinguish illusions, visual, as well as audial or other, from realities. Even AI's have this problem.(See Autonomous Car Collides with Bus: an illusion of abstractions?)

And What Are Understandings? Understandings, when articulated in language, are narratives (or, if highly structured, a program) of collection or connection. We may judge them incorrect, false, or incomplete but they are still understandings -- we might say, "misunderstandings."

People may not be at all articulate as to what they understand so we may have to observe how they proceed from their present conditions toward those outcomes that we judge them to be pursuing. If we observe persistent failure to achieve a goal we might reasonably judge that their understanding of how to achieve that goal is deficient.

There is a large number of likely misunderstandings that persist in any population because they need not or cannot often be put to any test whose outcomes enjoy broad consensus as to their pertinence. (So what if kids believe in the Tooth Fairy!) These questionable understandings are often characterized as (empirically) "non-disconfirmable" beliefs. Examples are:

a. The mind is co-extensive with the body, or:
b. The mind is not coextensive with the body.
c. The universe is a hologram; or
d. The universe is not a hologram.

It would appear that any rationales, or chain of rationales, that contain misunderstandings would thereby be severely weakened. (Many pundits who presume to speak for Science share such misunderstandings with those who presume to speak for Religion. See Pseudo-Science: the reasonable constraints of Empiricism.)

The Fractal(?) Generation of Rationales. Understanding in general is the ability to produce chains of behavior or their narratives which link a confronted situation of interest to a goal to be achieved. But, understandings alone are often narrow and give may give no indication of the interests or abilities of the persons who understand and may be expected to act. Thus, understandings may be just part of what we're looking for.

Rationales can be elaborated from understandings. Rationales often bring up systemic concerns about, say, indicator validity (Cue), actor interests (Concern), and actor abilities (Control) which underly attempts at interventions based on narrower technical understandings.

Understandings can be chained together to produce rationales for action, and rationales themselves chained to produce broader understandings. Thus we may go from merely speculative understanding of how to cross a river, to a rationale for expanding commerce across that river, to an understanding of how political influence can be brought about reiteratively by using market rather than, say, armed forces. Understandings developed this way can be connected together, in chains or trees, etc., to articulate, say, foreign economic development policies. (See The Fractalization of Social Enterprise)


Why Can We Understand Human Black Boxes but find It harder to Understand AI Black Boxes?
It is because many people working in AI feel restricted to technical issues of understanding the how's of economically pertinant AI functioning. Despite the the persistent anthropomorphizing of AI output -- and despite the suspicions that the ultimate goal of AI research is to create some kind of homunculus -- issues of rationale, especially in the contexts of human development and learning are discounted as off target. [5]

Autonomous AI stimulates the same kind of misgivings that one might feel about extra-terrestrials. We are not sure we can predict, much less control what they might do with us. It is a non-sequitur to believe that high computational ability equates to altruism. (See METI. Here We Are! Come Eat Us! Our Children Are Especially Tasty!)

We have built-in, so to speak, Cue-identifying abilities inherited through evolution, both physical, mental and social, that are sensitive to the norms of the environments of our development. We are Concerned to share a world with motile, sometimes dangerous, beings that we can hear, feel, smell, taste as well as see. And we know from personal experience what fear, hunger, danger, hate and social attraction are. [6]

We measure the autonomy we are willing to extend to our ancient non-human friends and enemies because we can guess well what drives them, and how to accommodate them to our society. And, not the least important, we can, with more or less success, Control and defend ourselves against them.

To pursue the issues of understanding and rationale, see Intervention. Helping, interfering or just being useless?

Cordially, EGR

(P.S. Check out this interesting interview with Gary Marcus called Making AI More Human.)

NOTES
[1] L Wittgenstein 1958b. Philosophical Investigations. ed. G.E.M. Anscombe and R.Rhees. tr. G.E.M. Anscombe, 2nd edition. Oxford: Blackwell. page 223. See comments by Simon van Rysewyk at Wittgenstein Light

[2] D Castelvecchi Can We Open the Black Box of AI? Nature magazine, October 5, 2016

[3] Sources for pseudo-English concoctions: French - Mots d'Heures: Gousses, Rames, The d'Antin Manuscript. Penguin,1980; German - J Hulme Mörder, Guss Reims. The Gustave Leberwurst Manuscript, Clarkson N Potter, 1981; Spanish -- My confabulation, EGR

[4] IPA, digits indicate tone levels. Cf. H A Gleason, An Introduction to Descriptive Linguistics (New York: Holt, Rinehart and Winston, 1955)

[5] See EG Rozycki Behavior in Measurability and Educational Concerns 

[6] See D Gross, Why Artificial Intelligence Needs Some Emotional Intelligence

Gun Fun or Safe Citizens? Must We Make Trade-Offs?

It has been said that democracy is the worst form of government except all the others that have been tried. -- Churchill

For the fanatic with a hammer, every problem is a nail. -- Adage

Fun vs Risk. Like dodgeball, horseshoes, darts, and archery, shooting guns can provide a great deal of enjoyment. But like all of the activities mentioned there is often more than a small element of risk involved in them.

But it is only by learning to carefully balance benefits pursued against costs risked that we develop important skills and moral strengths. We do not live in a Garden of Eden. So it is that guns, whose unskilled handling is particularly risky, still provide us an educational tool for pursuing, maintaining and defending important qualities of life.

Preference ≠ Morality. Unfortunately, many people tend to mistake their personal preferences for moral principles. This has happened on the many sides of the arguments for gun control. Particularly where kids are involved, the emotional din raised by presumed adversaries tends to drown out any careful examination of the issues involved. Disputants postulate cabals who would threaten our peace and freedom by imposing their own favorite hammer to solve anything that looks like a problem.

So it is that we read of “gun guys” who are sure that anything that looks like a problem can be dealt with a well-placed bullet. Standing up to them are “educators” who sit on their high horses and accept no trade-offs for their fervent belief that it is only more “education,” schooling and higher degrees that will pacify the savage breast. Despite this name-calling and caricature, we can, nonetheless, start somewhere to seek common understanding.

Two Principles for Decision. There are two general principles that, I suspect, would be acceptable to almost all participants in the “debate.” The first is Fair Balance of Benefits and Costs. This principle can be formulated in this way: In proportion to the severity of the costs risked, those who benefit more should assume greater risks.

For example, private gun ownership could be restricted to members of private gun clubs who must buy insurance to pay for damages caused by their members. Clubs would incur liabilities for uninsured damages. This likely high expense would probably make it worth the members’ trouble to screen applicants for membership and to regulate their activities. Maybe in this way the “monsters” who some imagine are the only ones who misuse firearms might be separated out from legitimate users, even if such monsters have already been born into club-member families. (See, for further example, Six things Americans should know about mass shootings)

The second principle is this: Nothing is a means to an end if it’s ineffective to that end. That is, things that don’t or wouldn’t work to achieve a goal, cannot be reasonably argued to be required.

For example, one of the weaker arguments for unrestricted private gun ownership is that it provides a counterforce to government (federal, state, city?) use of armed forces to deny us our Constitutional liberties. A similar nightmare was given expression both in the 1984 film Red Dawn. and its 2012 remake, involving foreign rather than local would-be conquerors. The 1984 teenagers manage to put up a significant resistance with locally cached weapons against Russian invaders and their Cuban and Nicaraguan henchmen. So, clearly, AK-47’s, rocket launchers, even, are needed by us common folk, by us Washington D.C. outsiders, to keep government tyranny in check.

Does anyone read military history anymore? Better yet, has anyone watched the Military Channel lately and seen how Iraq’s Red Guard was shredded and Saddam Hussein’s government toppled by U.S. military might? In two different wars? Which American weekend survivalist warriors -- even if they had access to weapons much more powerful than AK-47’s (and were teenagers) -- could hope to stand up against U.S. special forces, combat drones and spy satellites?

U.S. Armed Services members take an oath to protect the Constitution; not, as in Nazi Germany, to obey the whims of some Supreme Leader. Besides, in a government such as ours with so many competing organizational, political and functional units, it is highly unlikely that any one group could secure the obedience of a sufficient number of weapons-controlling bodies at all governmental levels to bring the Red Dawn nightmare into reality.

Better to organize politically to pursue and maintain what you value, than pretend to prepare for a conflict, for a Spartan military existence, for which few people, even gun enthusiasts, seem inclined to trade away the comforts of job, family and daily life. That Americans are not Spartans is shown by the fact that there is no military draft which might reasonably spread the costs of national defense to those who most benefit from them. However, most people are content to go about their private lives and let the few Spartans among us, together with a large number of newly immigrated, impoverished or adventuresome young people, serve in our own “all professional” military services.

Being effective politically means learning to trade-off cherished habits to foster political stability and functional consensus. Some of these tradeoffs follow (note links):

1. Stop sneering at politicians. It’s a difficult job. No one is a worse person for making compromises.

2. Reduce name-calling: it produces pointless stereotyping and reification. Practice the self-control you preach to your kids.

3. Realize (and act accordingly) that your (or your group’s) special hammer is one among many. Your perception of problems may not be anyone else’s. Negotiate turn-taking. Learn to tolerate and be civil to those people you do not like. Indeed, don’t think you have to like people just because you tolerate them.

4. In solving problems try to deal with concrete issues. Avoid arguing about vague theoretical, especially incendiary, statements, e.g.” Is abortion really murder?”

5. Don’t expect to settle an issue once and for all. Situations change. Priorities change. People change. Consequently, policies may need to change. Expect negotiation to be a repeated process.

6. Finally, don’t expect even the most enthusiastically celebrated consensus to persist as people come to believe they misunderstood what they were agreeing to; or, how to achieve the agreed-to goal.

Vigilance and political participation are the price of liberty and justice. The lazy, the distracted and the complacent risk losing both of these.

For references and to examine related issues further, visit the hyperlinks given in the text and also see Trading Off "Sacred" Values.


Cordially
--- EGR

Masks: Words that Hide Agendas

edited 4/3/20

Vote for me and if the legislature doesn’t pass budgets on time, they won’t get paid. -- Political candidate in Pennsylvania, Fall 2012

Despite years and years of nationalistic indoctrination, Americans tend to coagulate into a loose confederacy of “tribes.” Consequently we tend to be “agreement junkies,” especially when we are in the company of people we believe to be of “our own kind.” It’s just not nice to be contrary, or even, to show lack of enthusiasm.

In such a culture, simple slogans, like toadstools, spring up to mask substantial false consensus as to our individual positions -- e.g. whom are willing to vote for, what we value, whom we disagree with. Experienced politicians recognize this lack of depth in our professions of allegiance and thus steer away from explicit and specific discussions of their proposals for change.

Technical knowledge is not very widespread among the voting populace. Nonetheless, scientific sounding words, such as, entitlement, efficiency, cost, benefit, rent, profit, and the like are frequently invoked in political rhetoric.

But even more common-sounding words like harm, threat, risk, safety, protection, are sufficiently vague to mislead voters to imagine that they will be getting benefits which the future office-holders, in fact, can escape delivering through clever reinterpretation of their “promises.” (For example, does the would-be Pennsylvania legislator cited in the epigraph really think he will be able to control paycheck distribution for legislative members?)

The dynamics of consensus confusion revolve around issues of value, priority and probability. In a controversy, these issues are finessed by unexamined concessions about what is valuable or cheap (value), what is important or expendable (priority), and what is imminent or remote (probability).

For example, efficiency is generally seen to depend upon minimizing costs in pursuit of a goal. But if the goal is not accepted universally to be of value, or of the same priority in competition with other possible goals, estimations of efficiency will vary substantially. Universal college education is such a goal.

Consider risk as another example. Unless a situation is universally viewed as both threatening and imminent, many will discount what others consider disaster. Global warming seems to be such a situation.

During times of political interest, we find vague or confused references to “liberals” or “conservatives,” “moderates” or “extremists,” emphasized by media “spokespersons” much more intent on attracting attention -- ya gotta sell advertising -- than communicating accurate information. (See Believer vs Atheist, Conservative vs Liberal, and other distracting frauds.)

The assumptions people make, seldom stated publicly, support their own agendas -- not necessarily ours. And any but the simplest word is open to misunderstanding if we are mistaken about the assumptions its users are making. (See, for example, items specified in a Controversy Analysis Worksheet.)


Also, important in everyday life are certain assumptions that even more deeply underlie concerns for conflicts of individual vs individual, or, even, group vs group. These are certain “social assumptions,” principles so profoundly embedded in our culture that in many situations we overlook them entirely; although, upon consideration, we would find them far from indubitable.

These principles are

The Principle of Command: To command is to control, i.e. presidents can control national employment and foreign events;

The Principle of Accepted Value: What we value, everyone (should) value, i.e. no sensible, sane person could disagree with us; and

The Principle of Objectivity: Facts are facts, and, e.g., they are determinable without error by human agencies.

How do such principles come to control our thought processes? The quick answer is:

a. we rely on such Social Assumptions to support not only our personal agendas, but our conceptions of Knowledge, of Authority and of Obligation;

b. consequently they support practices we take to be important, essential to our personal, social, corporate, political, and moral life.

Unfortunately, quickness does not guarantee accuracy.

For references and to examine these issues further, see Questionable Assumptions in Social Decision Making


Cordially
--- EGR

Raising Test Averages: is this good?

Put down six and carry two. ... Gee, but this is hard to do.
You can think and think and think, till your brains are numb.
I don't care what teacher says,I can't do that sum.
-- Victor Herbert & Kenny Baker

Suppose we find that over time more and more kids in a school have test scores above average. Is this good? Not necessarily, even though most people would jump to the conclusion that achievement was going up.

Suppose we find find that over time more and more kids in a school have test scores below average. Is this bad? Not necessarily, even though most people would jump to the conclusion that achievement was going down.

Find out how a common misunderstanding (even by so-called “authorities”) makes schools look bad. -- EGR

To see the details go to Every Child Above Average
Cordially,
-- EGR

Critical Assessment: are educators trained to do it?

edited 5/25/20

Many of the parents who allow their children hours of TV and who read little, if anything, besides the newspapers, contribute nonetheless vociferously to that chorus that decries the schools for failing to turn out readers. This irony is seldom lost on educators.

But even among educators, “criticism” is generally taken to mean “rebuke,” or “expression of dislike” lined up along partisan, usually political, differences. The ancient honorable notion that criticism can be non-antagonistic critique, an objective comparison to some commonly recognized standard, is, more often than not, forgotten. (See The Art of Constructive Criticism)

The deepest, most pathetic irony in education is this: a conference of educational leaders will more often than not turn into a celebration of non-communication, a values-clarification workshop writ large, each person revealing his or her "philosophy," neither expecting nor offering criticism, incapable of even simple rebuttal or defense, respecting each other's right to his or her own opinion, indeed, respecting each other like crazy.

Exercises in civility and understanding "where everyone is coming from" complete the ritual; critical examination is foregone and people finally got down to political "brass tacks."

Are we surprised that our educational leaders, professed "change agents", uplifters and reclaimers, ultimately engage in but one perennial pursuit: bandwagoneering?

To examine these issues further, see CRITICAL ASSESSMENT AND VALUES IN EDUCATION


Cordially
--- EGR

Teachers: to make things clearer, talk less, much less and, instead, …!

In the course of his or her professional preparation, every teacher has had the experience of having to listen to a long lecture on the evils of lecturing. Lecturing is a high status activity: not only do professors do it almost exclusively; but, generally so do teachers down through middle school.

Why do teachers at all levels talk so much? Isn't a picture "worth a thousand words"? Not only do teachers tend to underutilize visuals in their teaching in the middle and high schools, but university faculty (and students, too) tend to react to the use of visual aids with hardly disguised condescension. Rapid-fire loquaciousness is mistaken for profundity of thought. Lips move and heads bob and understanding remains untested.

We are trapped in antique metaphor. For example, when we understand something, we say that we "see" it. As teachers we chatter away to make things "transparent" to our students. We take joy in a "bright" student, who is a "clear" thinker. But these visual metaphors are misleading, the rubble of ancient philosophies. Something else is needed.

To examine these issues further, see The Mind’s Eye and Pedagogical Practice

Cordially
-- EGR