0
$\begingroup$

Placing Hype and Fame Aside

Marketing departments for technology corporations throw the word smart around with cavalier inaccuracy. We know, if we have written any mobile phone applications, that there is nothing notably smarter about smart phones than the New England telegraph system of the 1800s, unless the application loaded into the mobile device provides some smartness. Hype and overconfidence aside, the voice communication system was several orders of magnitude smarter when people were operators and you could talk with them to resolve your connection.

Cognitive digital systems may come soon, as the media hype proclaims, but this prediction was made several times before. Such predictions are notoriously optimistic. Is the current set of predictions made by those who present themselves as mavens of technology similarly optimistic?

The Folly of the Aeronautics Analogy

People who argue that skeptics felt that way about many aeronautics achievements before the first sustained, controlled flight at Kitty Hawk, don't understand the foundations of aeronautics.

The ancient desire to fly like a bird is evident in ancient times. Images of an early model airplane found in a tomb at Saqquara, Egypt in 1898 can be displayed using a simple image search on the web.

"cairo museum" "model airplane"

Some think that the artifact supports the passage of extraterrestrial intelligence through our system, which is possible, but an absurd explanation for the existence of a model airplane in an ancient tomb. The object is likely simply the result of people watching birds and wishing they could fly.

The assembly of the object, especially the static uni-wing replacing the dynamic wing of a bird, clearly shows the emergence of the idea of gliding without propulsion not unlike the first flight at Kitty Hawk.

The dating estimates for the Egyptian model airplane are in the neighborhood of 200 B.C., which may be the oldest record of the vision of artificial flight, although it is easy to argue that there may have been much earlier conceptions of human flight than that. Conservatively, lets go with 150 B.C. as the worst case scenario.

The flight at Kitty Hawk was in 1903. So we have 1903 - (-150).

It took at least 2,053 years from the ontological birth of the idea of human flight to reach the realization of practical aeronautics.

The Illusion of Technology Recency

The claim, "But they didn't have technology back then," is untrue. They didn't have technology developed to the point where they could apply mechanics to the design of the wing, the supports, light weight frames, petroleum refinement, propellers, and internal combustion engines yet.

Consider that the and the birth of technology was clearly visible before that, in 3500 B.C. in Mesopotamia.

  • Friction bearings that held potter's wheels in such a way that they didn't tip over and could be turned by hand
  • Irrigation systems that regulated the amount of moisture to optimize plant growth and limit root rot.
  • Plowing devices that balanced the muscular force of two oxen in such a way as to direct plow motion in a straight line (a yoke)

These are all, without any doubt, evidence of mechanical design. The ancient designs were highly practical, still in current use today, and arose in the absence of several other important factors that postmodern people take for granted.

  • Faraday's work on electricity and magnetism along with others
  • Lavoisier's work on oxygen that led to molecular comprehension
  • Newton's work on unifying the field of physics and introducing calculus

Certainly the pioneering work over the last 5,500 years provide a foundation for further discovery, but none of that work, including the von Neumann architecture and the integration of transistor circuits into surface mount devices on mother boards provides any opportunity beyond what can be thought about thought without computers.

Notice that Aristotle, Leibniz, Galileo, Copernicus, Newton, Bohr, Gödel, Wiener, E. Mach, and Einstein had no test bed or computers on which they relied. Science and technology was born in their absence. One could reasonably argue that the progress in science has slowed because of obsession with browsing, blogging, hacking, and IT in general.

When Was the Ontological Birth of AI?

Although the sexagesimal (base 60) calculating apparatus of the Sumerian abacus appeared around 2,500 B.C., it was not an attempt to produce automation. The abacus was a tool to extend the mind, not create a technical offspring from it. Calculating tools result from the biological fact that numerical aptitude is only wired for zero (the absence of any) and small counting numbers from 1 to around 7. Working with other numbers requires training of the brain's neural networks. What we call math class is required to do even the simplest arithmetic.

The idea that AI was born with the ancient Hebrew desire to forge gods, an idea proposed elsewhere, is ill-founded. The concept of creating a mechanical mind was not the objective in those ancient stories of golden gods. The objective in those accounts was gaining advantage by creating an alternative control path through polytheism. (Not exactly a scientific approach, and not very open-source minded either.)

The emergence of the prominence of the idea that human intelligence could be simulated electronically probably began with the creative competition between Claude Shannon, Norbert Wiener, and their contemporaries around the time of WWII. The construction of digital computers were certainly funded by the United States and the United Kingdom in response to European and Pacific crises.

But neither the early musings of creating gods nor the more recent emergence of digital electronics is a realistic ontological birth date.

The first recorded engineering attempt for intelligence is probably the recipes for the creation of life progressing from reptiles to humans in Jābir ibn Hayyān's Kitab al-Ahjar (Book of Stones), around the year 800. He further explained, "Many things are agents, for which the realm of nature becomes a matter and in which they act in order to bring to existence that which is supposed to exist, like the human being and other things. ... He -- that is, the human being -- is an agent acting in matters other than him when creating artificial forms, and object[s] of the realm of nature, ... and his activity comes by reason[oning about] his being and the act of another who undertook his actualization, that is his creation."

If you understand the cultural framework in which ancients speak, this statement from Jābir ibn Hayyān, combined with his recipes and his discourse regarding the distinction of man from other organisms on the basis of intellectual capacities, is a clear indication that a man is meant to ultimately create an artificially intelligent man.

Islamic sexism aside, this is probably the first historical conceptualization of artificial intelligence as a realistic (and inevitable) scientific endeavor.

Now for Rational Prediction

Assuming only a few things, we can estimate the emergence of the electronic brain. These are assumptions of a much more rational nature than that technological advancement is exponential or that computers can be used to understand human intelligence, neither of which have any legitimate supporting evidence.

  • That creating cognition and the many other prominent features of human intelligence (of which predicate logic and pattern recognition are only two) is at least as technically difficult as creating an airplane that can land without crashing
  • That the deeply held belief of those living in an world where the sustenance of the global economy depends so heavily on technology, that it is a trivial fact that technology grows exponentially, is false (on the basis that humans tend to gloss over the effort required to achieve what has already been achieved, as discussed above)
  • That people wished to fly like birds just as strongly as that they wanted a competitive new species to take their jobs

Adding 2,093 (the temporal span between ontological aeronautics and practical aeronautics) to 800 A.D. (the ontological birth of AI) we get the year 2893.

Back To the Question

As disappointing as this prediction result may be to those who won't live to the 29th Century, is this not a more realistic and logical estimate for the earliest likely realization of an artificial brain?

Are not the current predictions grossly overoptimistic?

$\endgroup$
  • 5
    $\begingroup$ I think you need to edit this question down. It is well written but reads more like a blog entry presenting your opinion then polling for feedback at the end than a Stack Exchange question. In addition, I think that to gain focus it could do with a "notable claim" that can be addressed in an answer (c.f. Skeptics Stack Exchange) rather than a generic concern about timing. Proponents of The Singularity can be very bullish about timing, but it is not clear whether that will actually happen. Whilst invertebrate brains are being transcoded to silicon today. $\endgroup$ – Neil Slater Aug 5 '18 at 6:41
  • 1
    $\begingroup$ No literature or research proposals exist for whole mind uploads at this time. It is a futuristic project proposal. The highest power scanning devices (large field functional MRI systems) do not have the capability of reading neural circuits in vivo and the action of the organelles is still under study. It may be decades or centuries before a whole mind upload is accomplished for even the smallest invertebrates if a requirement is that the state of the organic memory inside the neurons is included. $\endgroup$ – FauChristian Aug 5 '18 at 20:55
  • $\begingroup$ Ok, thanks for taking a look. It may be worth bearing in mind for future questions, as the subject matter and sentiment are a good match to the site IMO, but the length and breadth of introductory material are a problem. A good answer needs to address points raised in the question, and there is simply too much here in one place. $\endgroup$ – Neil Slater Aug 6 '18 at 8:30
  • 2
    $\begingroup$ @NeilSlater I'm torn here. This is definitely a thesis leading up to a question. (I'm inclined to believe it causes no harm as these types of questions are relatively rare.) I've provisionally created a "predicting-ai-milestones" tag to disambiguate from the function-oriented "prediction" tag, and wondering if we need to also add "social" or "soft-question"... $\endgroup$ – DukeZhou Aug 7 '18 at 19:09
  • $\begingroup$ Links: The OpenWorm project plus an example from last year where the "brain" was used to control a lego robot. The honeybee brain variant was the vision system only, and I think this one only transferred the architecture, and was then separately trained for connection strength: iflscience.com/plants-and-animals/… $\endgroup$ – Neil Slater Aug 9 '18 at 10:21
2
$\begingroup$

The primary proponent of the singularity is Ray Kurzweil, and his writings permeate the first pages of search results in Google for singularity related search terms. His SEO is probably helped along by the fact that he is the director of research for Google.

One person's analysis of former R.K. predictions, The Singularity is Near: How Kurzweil's Predictions Are Faring, Posted by Paul, 2017, Antropy eCommerce, based on Kurzweils survey of his own former predictions, does not indicate a high reliability in past predictions. The current prediction that the singularity will occur by 2045 is unsure, but if the actuarial tables are representative, I will be around to witness whether or not that prediction better than some of the previous ones.

This more refined statistical analysis provides a better picture. At the end of the summary, these two items stand out.

  • There is little difference between current predictions, and those known to have been wrong previously.
  • It is not unlikely that recent predictions are suffering from the same biases and errors as their predecessors.

Removing the double negative, that last item is, "It is likely that recent predictions are suffering from the same biases and errors as their predecessors." Based on the analysis above the summary, the supporting data, and the legitimacy of the assumptions made in this question, I concur.

$\endgroup$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.