The big question of the year is on ... the machines that think

No, not my question.  Though, I have obsessed enough over it, even right here in this blog.

It is Edge's annual event--the big question for the year.  For 2015, it is: What do you think about machines that think?

As always, it is an impressive lineup of thinkers who respond to that question.  The ideas there are way more than my limited understanding capacity can handle.  So, I picked my way through that intellectual smorgasbord.

I decided that I would read contributions from non-white males.  Why?  This is a field that has been the domain of white males.  Slowly females and other males are cracking through it all and, thus, I figured it might be interesting to read their perspectives.

Even that was one too many people.  I figured I would develop another filter based on the names that I read: there were two Marias and one Mary.  So, hey, responses by three Marys.  There was one Indian name; so, of course, yes to that.  And then one of my all-time favorite polymaths ever: Freeman Dyson.

I had a game plan.  I went in.

First up, the Indian, Satyajit Das, a "former banker."
I read his piece, re-read it, and wondered why he was in the lineup.  A rambling, broad, blah response.

Will Mary save me?
Nope. More blah!

I bet that this Maria will have something profound; after all, I have often read her blog-posts and they have always been insightful.

Thinking is not mere computation—it is also cognition and contemplation, which inevitably lead to imagination. Imagination is how we elevate the real toward the ideal, and this requires a moral framework of what is ideal. Morality is predicated on consciousness and on having a self-conscious inner life rich enough to contemplate the question of what is ideal.
The famous aphorism often attributed to Einstein—"imagination is more important than knowledge"—is thus only interesting because it exposes the real question worth contemplating: not that of artificial intelligence but that of artificial imagination.
Of course, imagination is always "artificial" in the sense of being concerned with the un-real or trans-real—of transcending reality to envision alternatives to it—and this requires a capacity for holding uncertainty. But the algorithms that drive machine computation thrive on goal-oriented executions, in which there is no room for uncertainty—"if this, then that" is the antithesis of the imagination, which lives in the unanswered and often, vitally, unanswerable realm of "what if?" As Hannah Arendt once wrote, to lose our capacity for asking such unanswerable questions would be to "lose not only the ability to produce those thought-things that we call works of art but also the capacity to ask all the answerable questions upon which every civilization is founded." 

Score!

Off to the other Maria then.  What does the Caltech professor has to say?

I for one, am more concerned about humans who drop thinking or are brainwashed, than smart thinking machines taking over.

I, too, worry about humans who are ready and willing to be brainwashed and do not care to think for themselves.  Go on, professor.

 I foresee the emergence of hybrid human-machine chimeras: human-born beings augmented with new machine abilities that enhance all or most of their human capacities, pleasures and psychological needs. To the point that thinking might be rendered irrelevant and strictly speaking unnecessary. That might provide the ordinary thinking humans a better set of servants they have been looking for in machines. 

Oh, no!  This is the kind of a scenario that worries me.

Finally, Freeman Dyson.  I can always count on him to be clear and direct.  No blah. No unnecessary qualifiers.  No hedging.  A wonderful thinker.  Will he let me down?  I hope not.

Turns out that Dyson had the shortest response of all, which I provide you in its entirety:

I do not believe that machines that think exist, or that they are likely to exist in the foreseeable future. If I am wrong, as I often am, any thoughts I might have about the question are irrelevant.
If I am right, then the whole question is irrelevant.

Awesome!  I knew that they guy would be direct.  Maybe two generations from now, we will find out Dyson was way wrong.  I don't care.  I love the way he clearly articulates his thoughts based on a remarkable knowledge-base.

Here's to hoping that "nature cannot be fooled."

By now, you know where this is from, right?

Think about this: Science is only a small part of human capability

One of my favorite living intellectuals, Freeman Dyson, yet again delivers an essay in which he so easily weaves his polymath nature and his interactions with many remarkable people in his long life.  All in what one would expect to be a dull and boring topic of physics at the fringe, when pursued by amateurs.  In a book review essay in the NYRB, Dyson places science in its place:

[Science] is only a small part of human capability. We gain knowledge of our place in the universe not only from science but also from history, art, and literature. Science is a creative interaction of observation with imagination. ... Imagination by itself can still enlarge our vision when observation fails. 

"creative interaction of observation with imagination" ... Isn't that a lovely description?  If only most science teachers would keep that in mind when they teach their students, from elementary school to college.

Dyson reminds us that there are limits to imagination, and goes after string theorists who conjure up visions that seem to be no different from those offered by highly creative amateurs without solid scientific backing:

The fringe of physics is not a sharp boundary with truth on one side and fantasy on the other. All of science is uncertain and subject to revision. The glory of science is to imagine more than we can prove. The fringe is the unexplored territory where truth and fantasy are not yet disentangled. Hermann Weyl, who was one of the main architects of the relativity and quantum revolutions, said to me once, “I always try to combine the true with the beautiful, but when I have to choose one or the other, I usually choose the beautiful.” Following Weyl’s good example, our string cosmologists are making the same choice.

What is wrong with string cosmology?

String cosmology is a part of theoretical physics that has become detached from experiments. String cosmologists are free to imagine universes and multiverses, guided by intuition and aesthetic judgment alone. Their creations must be logically consistent and mathematically elegant, but they are otherwise unconstrained.

The disconnect from "observation" is sometimes the case even when I listen to faculty colleagues or politicians, whose imaginations simply run amok :)

Why a sababtical? To be a stranger!

I thought this advice for a sabbatical was neat.  But, Freeman Dyson does it better, of course (ht):

As my friend the physicist Leo Szilard said (number nine in his list of ten commandments): “Do your work for six years; but in the seventh, go into solitude or among strangers, so that the memory of your friends does not hinder you from being what you have become.”

Ok!

Why is my teaching an "alternative" style? It is all about thinking!

For a couple of years now, I have made sure to emphasize to students my view that this is a world that is vastly different from even a decade ago.  From even five years ago.  I mean this in the context of teaching and learning.  What s that difference?

Students can access all the information they could possibly ever want without coming to the classroom.  Enrolling in a university or coming to the classroom after registering for classes is not about gaining access to information and facts, but is to make sense of them.  To understand their meaning.  To be able to then ask questions, many of which, hopefully, will be troubling questions.

If they don't trust me on this ... wait, that too is an advice I give them: don't trust me.  Well, don't merely trust me, but verify too.

Anyway, I can now supplement that with observations from George Dyson--yes, the son of Freeman Dyson.

We now live in a world where information is potentially unlimited. Information is cheap, but meaning is expensive. Where is the meaning? Only human beings can tell you where it is.

I tell students that this is where the power of Google (or any such tool) comes in--once we are able to craft questions that help us understand an aspect of this highly complex world, then getting information related to that is immensely less complex than ever before in history.  Again, Dyson:

Finding answers is easy. The hard part is creating the map that matches specific answers to the right question. That’s what Google did: They used the power of computing – which is cheap and really does not have any limits – to crawl the entire internet and collected and index all the answers. And then,by letting human beings spend their precious time asking the right questions, they created a map between the two. That is a clever way of approaching a problem that would otherwise be incomprehensibly difficult.

The net result is that my teaching and testing comes across as unconventional and as an alternative style.  In my classes, I engage students with questions.  When it comes to tests, it is not their ability to recall information that I am interested in; I am, instead, far more interested in providing them opportunities where students can demonstrate their thinking skills.  Which is why I remark to students that I feel like puking when I see tests in which questions are true/false or multiple-choice.  These do nothing to develop in students the ability to make sense of the fuzziness that surrounds us.

Even worse is to simply tell students that the course will require essays on topics of their choice.  Why?  Because most students have no clue how to ask questions.  Writing papers require that ability to formulate a question and then going after the supporting evidence when answering the question.  Higher education is then about understanding concepts enough to be able to ask questions.  In the format that most of higher education is, students rarely are taught how to ask questions.  And if they never figure this out after four years of university education, then all the access that Google provides will be of zero value.

One final comment from Dyson:

The danger is not that machines are advancing. The danger is that we are losing our intelligence if we rely on computers instead of our own minds. On a fundamental level, we have to ask ourselves: Do we need human intelligence? And what happens if we fail to exercise it?

But then, it is not the students I am worried about: I am disappointed with faculty--not merely at my university--who think that their role is to be that cliched "sage on the stage."  I feel like puking--yes, my favorite phrase!--when I walk by classrooms and see nothing but text-filled PowerPoint slides.  And even that "information" is often incorrect.  Once, I paused outside a classroom where a full professor of criminal justice was lecturing about the US Supreme Court.  It was depressing to see the slide listing the justices because the information was old--no Sotomayor or Alito, but the names of retired justices!  The guy hadn't even bothered to update this basic information.  When a google search would provide the accurate information. 

A bonus for you, dear reader, if you read until here: a while ago, the Atlantic had a fantastic essay about Freeman Dyson's climate change oddities, and in that context also discussed the father/son relationship.

Richard Feynman, I hardly knew ye :(

I could not have started my day any better, it feels like.

Freeman Dyson, whose essays in the NYRB are always ones I look forward to, has a wonderful piece on Richard Feynman, while reviewing two biographies of that phenomenal thinker as a person (ht).  Dyson, a gifted and  accomplished polymath himself, writes that perhaps "Richard Feynman is rising to the status of superstar" among the worldwide masses of non-physicists, similar to the standing that Albert Einstein and Stephen Hawking have.

I loved Dyson's summary of Feynman's picture of the classical and quantum layers of nature.  It is amazing how these gifted folks are able to narrate complex scientific ideas to those of us whose intellectual abilities are way, way down the ladder.  Dyson writes:

Feynman’s picture of the world starts from the idea that the world has two layers, a classical layer and a quantum layer. Classical means that things are ordinary. Quantum means that things are weird. We live in the classical layer. All the things that we can see and touch and measure, such as bricks and people and energies, are classical. We see them with classical devices such as eyes and cameras, and we measure them with classical instruments such as thermometers and clocks. The pictures that Feynman invented to describe the world are classical pictures of objects moving in the classical layer. Each picture represents a possible history of the classical layer. But the real world of atoms and particles is not classical. Atoms and particles appear in Feynman’s pictures as classical objects, but they actually obey quite different laws. They obey the quantum laws that Feynman showed us how to describe by using his pictures. The world of atoms belongs to the quantum layer, which we cannot touch directly.
The primary difference between the classical layer and the quantum layer is that the classical layer deals with facts and the quantum layer deals with probabilities. In situations where classical laws are valid, we can predict the future by observing the past. In situations where quantum laws are valid, we can observe the past but we cannot predict the future. In the quantum layer, events are unpredictable. The Feynman pictures only allow us to calculate the probabilities that various alternative futures may happen.
The quantum layer is related to the classical layer in two ways. First, the state of the quantum layer is what is called “a sum-over-histories,” that is, a combination of every possible history of the classical layer leading up to that state. Each possible classical history is given a quantum amplitude. The quantum amplitude, otherwise known as a wave function, is a number defining the contribution of that classical history to that quantum state. Second, the quantum amplitude is obtained from the picture of that classical history by following a simple set of rules. The rules are pictorial, translating the picture directly into a number. The difficult part of the calculation is to add up the sum-over-histories correctly. The great achievement of Feynman was to show that this sum-over-histories view of the quantum world reproduces all the known results of quantum theory, and allows an exact description of quantum processes in situations where earlier versions of quantum theory had broken down.

It was neat to read this especially because a couple of weeks ago I read the interview with Leonard Susskind, whose string theory ideas attempted to provide a unified theory of physics and nature.  Too bad that Scientific American has this behind a subscriber wall!

Dyson's essay comes at a time when we are marking another important milestone in science and technology that also, in a way, introduced Feynman in a big way to the American public and the world--his work with the committee that inquired into the causes for the space shuttle Challenger to explode moments after liftoff.  Here we are having witnessed the final launch of the spaceshuttle, and after Atllantis returns, the space shuttle program will end.  Feynman's simple demonstration of the failure of the "O-ring" when exposed to icy conditions is legendary.  Dyson writes that Feynman was critical of the political culture at NASA that vastly overestimated the safety and success of the spaceshuttle.

The second opportunity to educate the public concerned the culture of NASA. Feynman wrote an account of the cultural situation as he saw it, with the fatal division of the NASA administration into two noncommunicating cultures, engineers and managers. The political dogma of the managers, declaring risks to be a thousand times smaller than the technical facts would indicate, was the cultural cause of the disaster. The political dogma arose from a long history of public statements by political leaders that the Shuttle was safe and reliable. Feynman ended his account with the famous declaration: “For a successful technology, reality must take precedence over public relations, for nature cannot be fooled.”

A profound statement that "nature cannot be fooled."  It is not merely about physics, of course.

I was a first year graduate student at USC when Feynman died.  Shift the time-line a tad, I would have had enjoyed a talk of his at Caltech. But then time flows only in one direction and "nature cannot be fooled."