A glorious accident, de Kaizer

Este libro, ya antiguo, y que además leí hace ya tiempo (pero que no había podido reseñar aún) transcribe un conjunto de entrevistas emitidas en TV en 1993 (y disponibles en Youtube).

Sin embargo, y a pesar de la edad, que se nota en algunas partes, es un gusto de libro, básicamente porque da acceso al pensamiento de varias mentes brillantes que se prestan a dar su visión sobre algunas preguntas “incontestables”, y que, desde luego, darían para varios miles de libros. Los expertos son Oliver Sacks, Daniel Dennett, Stephen Jay Gould, Freeman Dyson, Rupert Sheldrake, Stephen Toulmin. Unos más que otros, por supuesto, y también con distintos niveles de credibilidad. En todo caso, aquí tenemos la oportunidad de disfrutar de una versión condensada de su pensamiento y sus inquietudes, y, por supuesto, de que aparezcan muchas más preguntas que respuestas, algo que, en mi opinión, también tiene mucho valor.

Todo viene de una idea algo peregrina de Kaizer, como él mismo explica:

One languid summer night, I wrote down some questions on the back of an old hotel bill, for no reason except to kill a little time. Why do some bamboos flower only once every 120 years, and how do they count the passing of the years? Why do we call grass green? How can someone love Rilke and Beethoven with all his heart, and at the same time butcher people as methodically as if he were stamping forms? How does a homing pigeon, released hundred of miles from its loft, find its way back to those few square inches? Was the appearance of our consciousness in an insignificant group of monkeys swinging through the trees intentional, pure chance, or an accident? Will there really be a time when we can download our consciousness in a computer and achieve immortality? Why is it that, to quote my beloved Immanuel Kant, at one and the same time our existence can seem as majestic as it seems ridiculous?

Oliver Sacks es el primero en intervenir, y su discusión sobre la analogía entre el cerebro y el ordenador es muy interesante, aunque creo que se ha quedado anticuada, porque él se quedó en un concepto de computación en el que las redes neuronales todavía existían al nivel actual.

p22. I am now convinced there is no computer in my head, that natural intelligence is not like artificial intelligence, and tha thinking is not like computation. We do not have hardware and software in the usual sense, nor do we have a memory store, or information processing in the mechanical sense. Our brain is altogether more of a biological nature, in some ways much more powerful than computation, in other ways less powerful.

A robot seems to have a repertoire that it cannot leave, whereas creatures seem to be above all nonprogrammatic, to change the experience and deal with events.

Pero esto no significa que todo lo que dice haya que descartarlo. Tiene reflexiones muy interesantes sobre la conciencia:

p23. I am worried about the reduction of thought to algorithms. An algorithm is a mechanical procedure to solve a problem. I think one should ask: Is human thought, or for that matter animal thought, algorithmic?

I do think that our consciousness and for that matter the responses of many animals, at least higher animals, depend on constructing a scene or a scenario. I think this scene may consist of, if not an infinite, at least a very large number of details, arranged in some sort of organic and dramatic unity, and also a sort of resilience, so that things can be slightly wrong or slightly distorted.

I cannot in my mind isolate movement from consciousness.

O sobre la dificultad de replicar el cerebro particular de una persona (incluso con redes neuronales)

Even in genetically identical people, various epigenetic influences such as cell migration and cell death ensure that by the time of birth every nervous system is unique. It would not be possible, given such diversity of nervous systems, to run algorithms and programs. The hardware is too variable to allow computation. In a computer any variation is an error, but in life variation is of the essence.

Basically the nervous system is individual from the start, continues to be individual, and becomes more and more richly individuated throughout the life of the person or the animal. So the tormenting dichotomy between the computer and the person doesn’t arise, because your brain is molded by your experience even as it molds your experience.

A continuación interviene Daniel Dennett, cuyas ideas no me llaman demasiado la atención, la verdad. No tomé ni una sola nota de su parte, quizá porque me parece casi todo wishful thinking.

En cambio, Stephen Jay Gould es mucho más provocador, con un par de ideas interesantes sobre la libre voluntad. La primera me recuerda a algo que comentaba Santi hace poco en el blog, y es que la clave es cómo definimos la libre voluntad en términos prácticos.

p103. “I may not be the master of my own destiny, but my sense of integration probably reflects a biological truth”. We are so complex as to be unpredictable; that what we call free will covers a great range, and that’s why we do have free will, in the traditional sense of freedom.

La segunda idea me parece interesante porque me recuerda bastante a Sen:

Someone who’s terribly poor and who lives in a ghetto without any means of existence has little free will

Freeman Dyson tiene unos planteamientos que me parecen muy sensatos, aunque a veces quizá demasiado optimista sobre el avance de la ciencia:

There are no unanswerable questions. But it all depends on how long a time scale you’re thinking of.

Sobre inteligencia artificial hace una reflexión muy interesante:

(p129) It’s likely there will be a merger. So that we will have neural circuitry doing the jobs that neural circuitry does well, and electronic circuitry doing the jobs that electronic circuitry does well. Very likely the two will be linked together, so you will not be able to tell in the end whether a robot is alive or not…Whether you consider it to be alive is a matter of definition….so that the question whether we will have conscious machines is in a way meaningless, because it’s the question of what you call a machine.

Rupert Sheldrake, en cambio, no sé muy bien qué pinta en el libro (creo que Kaizer tampoco lo tiene muy claro). No entiendo cómo se puede poner una idea tan estrambótica como la de la “morphic resonance” junto a las demás: No, la gente no tiene más fácil resolver algo porque ya se haya resuelto antes…Simplemente nos indica que hay problemas que sabemos resolver mejor que otros, o acciones a las que estamos más determinados por nuestras estructuras mentales. Y eso que el hombre no parece falto de sentido común en otras cosas:

p152. The creation of new forms and patterns, new species, new ideas is one of the ultimate problems. I don’t think that science can address it, because science confines itself to the study of regularities and repetitions.

El que más me gustó de todos fue Stephen Toulmin:

p191. It’s a curious kind of narcisism to think that every individual creates his own mind and personality from scratch in his own lifetime. There’s none of this sense that we are the inheritors and transmitters of traditions, that our world is shaped for us by the things we grow up into, and which we internalize and make our own.

p192. Wittgenstein showed that things people took as self-evident were based on all kinds of unjustified assumptions

Kaizer: We have plenty of beautiful answers, but formulating the most beautiful question is what it’s all about.

p203. The world is predictable within limits. But the limits begin to appear larger and larger.

Y el libro concluye con una mesa redonda que tuvo que ser apasionante en directo, y en la que Gould le da un par de tortas a Teilhard de Chardin:

p231. Take TdC, who had his momentary run of popularity. It was basic old-fashioned mysticism put into such an incomprehensible motion that the history of life was the inexorable progressive increase of spirit over matter, which will eventually culminate in the union of that growing of spirit with God. It’s something he called the omega point. That’s a kind of evolutionary theory that’s much more acceptable to people who want to see the history of life and the universe as progressive. Darwinism is indeed challenging.

Y Sacks, sin saberlo, también se la da a Harari (o a Dennett):

p268. I do think that we are, if you want, the sum of the processes and activities in us, but I also think these are not only uniquely marked by our experiences and values but uniquely coded. And I therefore think it would be difficult or impossible to transfer the information to another brain or download it to a machine. It is not information in that sense. It’s a unique organization of incredible complexity, which sort of dies with you.

En resumen, un libro que da gusto leer, incluso aunque no estés de acuerdo con mucho de lo que se dice.