The Decline of Hard Artificial Intelligence (revival of old Usenet discussion about the brain and A.I.)

A diagram of some Usenet servers and clients.  The blue, green, and red dots on the servers represent which groups they carry.  Arrows between servers indicate that the servers are sharing the articles from the groups.  Arrows between computers and servers indicate that the user is subscribed to a certain group, and uploads and downloads articles to and from that server.Usenet servers & clients (Wikipedia)

Someone else’s light-hearted Greek post in another blog (here), making fun of idealism and certain erroneous ideas about the brain, reminded me of an old Usenet discussion in the newsgroup alt.religion.buddhism.tiberan, back in 2000, where someone had posted a tribute to what is (commonly calledl) «Hard Artificial intelligence«; about which my views are (generally speaking) heretical; neither idealistic, nor materialistic (in a mechanistic way).

In response to that posting, I wrote:

A machine plays chess to perfection, but is not conscious it plays chess.

A machine kills you, but it not conscious that it kills you.
A machine performs actions, but is not conscious of his actions.
A machine will never be part of a SAMSARA cycle, ever.

However, the other participant in the debate (as usual) disagreed:

«Me Tarzan» <> wrote:

Machines may already be conscious.

Absolutely not so. Of this, we can be absolutely certain, at the present time.

I’ve been in A.I. work for more than fifteen years now. I’ve deleloped Expert systems that simulate debate, automatic translation programs, simulations of nerve cell action for OCR software (e.g. here) and practically the whole lot you can expect from software-based A.I.

I’ve also been an artist however, and expressed in many ways the view that simulations of intelligence can never be conscious, but can indeed exceed natural intelligence from the point of view of problem-solving ability and any other (objectifiable) aspect of intelligence.

Intelligence is a trivial, secondary property of sufficiently complicated information-processing structures. Consciousness is a non-trivial fundamental property of sentient beings and doesn’t need intelligence to exist (but it… helps).

There is some interesting work being done by a couple guys, I think in New Mexico. They are building machine insects, using chaos theory as their guiding principle. The insects respond to external stimuli according to their instinct to survive. Many think that this is all carbon-based insects do, all the way up to small mammals. In this way, these engineered insects are conscious. It was very interesting watching one of the engineers play with them. They acted as if they were really alive! He treated them as pets! (It was on the Discovery Channel).

Television, like all media, including the internet, is deceptive. It’s conditioning people into such stupidity because it is simply impossible to demonstrate consciousness on a television screen, or in Usenet writing space. It doesn’t manifest well in this way. The problem is objectification.

There is simply no limit to how many aspects of life one can simulate, and thus turn into an object which appears indistinguishable in behaviour from the «real thing» (which is not a «thing», but an awareness of things).

The fact that we can objectify consciousness infinitely, e.g. give drugs to political prisoners to make them compliant with interrogators, does not mean that consciousness can (ever be) equated with its objectifications. E.g. If I can… kill you, this is not a way to give you life, or create new life, and so on. Whatever manifests and is a part of observable reality, is by definition objectifiable. Thus we can choose to construct as many deceptive manifestations as we like, through simulation.

[ irrelevant speculation about animals deleted ]

The work being done in A.I. is trying to make machines human-like (with the exception I explained above). Naturally, machines will develop their own consciousness and the difficulty in saying that they will be human-like sentient beings and be able to transcend samsara exists because we do not know what they will become.

This is the prevalent mechanistic doctrine of our time. Like… selling you a sufficiently life-like… sex-doll, whom you can f@ck, never realising the difference between her and a… real lover. 🙂 Well, there IS a difference, no matter how perfect is the simulated robot-lover, although the latter… can sell more(!), if it can persuade you never to realize the …hmm… crucial difference. Basically, hence or otherwise, mechanism is the ultimate intellectual wank. 🙂

[ more speculation deleted ]

However, my feeling is that since we gear machines to interact with humans, their consciousness will develop into something that resembles human consciousness to the extent that the distinction between humans and machines becomes blurred.

Blurred awareness in human beings is the only way to achieve this. Like… never realising the difference between people on stage performing a play, and people in real life. This blurrring of human consciousness is called (in old European philosophy) «alienation«.

You could say that humans and machines will merge. Pretty cool huh?

Callously cold, mostly.

Humans can only merge with machines in the particular sense of acquiring mechanical body or brain parts.

Consciousness can be augmented, but simulations of consciousness are not conscious implementations of it.

– Read some Roger Penrose, for Buddha’s sake! 🙂

NOTES (2008):

POLIMEREK - a robot

Kismet, a robot with rudimentary social skills.

On reflection, that old heated Usenet discussion was fascinating. So many interesting ideas were discarded by the heat of the debate, where some… Mechanistically-minded (pseudo-)Buddhists and anti-mechanists (very few of these, in fact) kept lashing out against each other’s brains controversial ideas about the brain, mixed with misconceptions and prejudices.

In fact, quite a few important Computer Scientists and specialists in Artificial Intelligence, are inclined towards an anti-mechanistic stance against doctrinaire «Hard Artificial Intelligence». E.g. Nova Spivack, a distinguished scientist and founder of Radar Networks, the company of «Twine» (a tool for the Semantic Web). I am a regular fan of Nova Spivack’s blog, where I also wrote comments some time ago (e.g. in his post «Is Consciousness as Fundamental as Space, Time and Energy?», here).


Related articles

  • Later on, I will probably augment this post (or write a new post as a follow-up) with other extracts from the same discussion…

For the moment, however, enjoy the following (YouTube video of a) BBC interview with Roger Penrose,

YouTube – Roger Penrose – Cyclic Universe Model:

If you don’t believe in any particular metaphysical doctrine and you are an atheist or an agnostic (or even a… pantheist, like me!) with an open mind, It’s worth watching the following 10 videos. They present controversial ideas and open questions about (Quantum-)Consciousness and the mind’s mysteries. They feature important biographical and philosophical facts about some great scientists, including the mathematician Georg Cantor and the controversial physicist Roger Penrose:

Dangerous Knowledge part 1/10:

Dangerous Knowledge part 2/10:

Dangerous Knowledge part 3/10:

Dangerous Knowledge part 4/10:

Dangerous Knowledge part 5/10:

Dangerous Knowledge part 6/10:

Dangerous Knowledge part 7/10:

Dangerous Knowledge part 8/10:

Dangerous Knowledge part 9/10:

Dangerous Knowledge part 10/10:

One comment


Εισάγετε τα παρακάτω στοιχεία ή επιλέξτε ένα εικονίδιο για να συνδεθείτε:


Σχολιάζετε χρησιμοποιώντας τον λογαριασμό Αποσύνδεση /  Αλλαγή )

Φωτογραφία Google

Σχολιάζετε χρησιμοποιώντας τον λογαριασμό Google. Αποσύνδεση /  Αλλαγή )

Φωτογραφία Twitter

Σχολιάζετε χρησιμοποιώντας τον λογαριασμό Twitter. Αποσύνδεση /  Αλλαγή )

Φωτογραφία Facebook

Σχολιάζετε χρησιμοποιώντας τον λογαριασμό Facebook. Αποσύνδεση /  Αλλαγή )

Σύνδεση με %s