It has been a while since I talked about the technical Singularity, so I thought I might return to this topic. There is a question I wish to explore, namely: Why should anyone believe in the Singularity?

Consider the eye. It is a well-established fact that, since the eye does not possess limitless capabilities, it cannot reveal to us the whole of the visible universe. Indeed, the portion of the visible universe that we can see is but an infinitesimally small window and we are simply blind to most of the universe. There are things too far away to be seen by the naked eye, or too small. There are objects and phenomenon that can only be detected by instruments that pick up wavelengths the eye is unable to detect. 

I see little reason to suppose that similar limitations do not apply to the brain. The ‘universe’ of imagination, knowledge, inspiration, intuition, etc that we perceive is not the universe entire, but merely that tiny portion that the naked brain is capable of perceiving.

Returning to the eye, there is an obvious way around its limitations, one we have been applying for centuries. That is, the use of technology to augment the eye’s capabilities. With microscopes we can examine that which is normally too small to see. With telescopes we can see further than was previously possible. With instruments designed to detect radio waves or infrared light or gamma rays we unveil or at least infer the existence of phenomenon like cosmic background radiation, dark energy, dark matter, or black holes. A reasonably well-educated child could probably rattle off dozens of cosmic phenomenon and objects that Gallileo never heard or dreamed of. Thanks to technology, our visible universe is immeasurably  bigger, more complex, and full of questions than the universe we understood before technology augmented the naked eye.

Obviously, then, the aforementioned technologies do not merely augment the eye. They, supported by other advances that refine the scientific method, expand our knowledge. It can also, I would argue, expand intelligence. It seems to me that refuting the singularity hypothesis entails establishing at least one of two things:

1. The human brain is, or close to, the absolute best ‘engine’ of general intelligence there can ever be.

2. While it is possible in principle for something to be capable of general intelligence surpassing that of the human brain, the creation of such a thing lies beyond the powers of the human brain.

I do not think either possibility is true. Throughout history, it has been the case that what was once beyond the realm of possibility for one generation was technically possible for future generations. We are quite capable of constructing machines or concepts that are beyond the intelligence of any one person. Could an individual possibly design and build the Large Hadron Collider without help in its conception and construction? Of course, not, the very idea is ridiculous. Neither could any one person, working in isolation, establish the ideas in theoretical physics that required us to build the LHC in the first place. Newton’s point that one needs to stand on the shoulders of giants (in other words, you cannot work alone but need the support of other experts present and past) is just as true today as it ever was.

OK so obviously the human race itself is capable of more than an individual. But is the human race capable of more than the human race? That sounds impossible, but we should not be hasty in declaring it so. After all, isn’t it true that the human race today is capable of feats that were beyond our ancestors? We can attribute this capability to scientific progress, which Michael Shermer defined as:

“the cumulative growth of a system of knowledge over time, in which useful features are retained and nonuseful features are abandoned, based on rejection or confirmation of testable knowledge”.

It seems to me to take this growing system of knowledge and apply it to that which is the driving force behind it, so that we may better understand it. In other words, we can use science and other systems of knowledge like philosophy, in order to answer questions such as ‘how does the mind work’? We may then set about using our growing system of knowledge and expanding technological capabilities in order to create brains whose intellect surpass our own, or form collaborations between collective human intelligence and machine intelligence until there is effectively a super organism with an intelligence surpassing the human race. More and more I am reading about technologies surpassing human intelligence in some narrow field:

“Most [astrophysical] data and data constructs, and patterns in them, cannot be comprehended by humans directly….This requires the use or development of novel data mining or knowledge discovery in databases and data understanding technologies, hyper-dimensional visualization etc. The use of AI/machine-assisted discovery may become a standard scientific practice”- S.G Djorgovski and R. Williams, “Virtual Observatory: From Concept to Implimentation”.

“Sophisticated models derived computationally from big data- and consequently tuned by feeding results back in- might produce reliable results from processes too complex for the human brain. We would have knowledge but no understanding”- David Weinberger, Scientific American: The Machine That Would Predict The Future.

Both these quotes inform us that it is not beyond the realm of possibility to assemble systems of technology that can possess a comprehension above and beyond the human mind, indeed, all human minds. None of the scientists quoted throw up their hands and cry in despair, “well we have all this data and patterns within them, but we cannot understand it and nor can we ever have the technology to help us make sense of it”. No, they talk soberly about methods of data-intensive scientific discovery relying on processes too complex for the human brain. 

In other words, that window into the abstract universe of knowledge is expanding and is now beginning to allow the embryonic human-technology super organism to access knowledge that is beyond human comprehension, at least in narrow fields like astrophysics. Is this the Singularity? Perhaps not. But I would argue that it is an intermediary step towards it. I would not be at all surprised if future generations are not only capable of answering questions that are fundamentally too hard for us to solve, but also capable of asking questions we are not capable of asking or understanding (unless, of course, we can sufficiently augment our natural abilities using their technology).  Because I believe that, I have to believe in the possibility of a technological Singularity.
This entry was posted in technology and us. Bookmark the permalink.


  1. Great analysis, and I agree in part. But let me counter.

    The metaphor for the Singularity is of course the Black Hole, or vice versa. And if you fall in a black whole parts want OUT and parts want IN, because of tidal shear, as you whirlygig around the centrifuge. But even a direct accelerated trajectory straight in to the Singularity won’t do the trick… there will be a phenomenon of sphaghettification.
    If we have a take-off point, won’t a small segment of society be fully benefitting these advances, while most of society will be desoriented, or worse, left helpless?
    In fact I can argue we are entering that era allready. My neighbour is well-entrenched in VHS syndrome, and while she is a whizz with mobile phones, she has no clue what a “skype” is. I can explain it for an hour, and she won’t get it. My other neighbor burned a CD with songs for her. I proposed to do that as well and send her an email with music. One email! She clicked a single link and …and … it kept playing music..! She was in a state. How does that fit on my computer, there is no space!

    She is just clueless. I explained to her “It’s radio”. Just keep it simple, the poor lamb.

    I suggest to no amount of gentle education will ever help my neighbour to understand anything of the proper implications of celestial mechanics, nanofabrication, cloud computing, ubiquitous computing, the implications of life extension, etc. She actively purports to no way want to live forever, as in “and work this horrible job forever as a wrinkly old hag??”. Seriously she is not capable of permutating even two or three compound implication stacks.

    The Singularity is about Love, Trophy. To love or not love. If we have one without, the consequence will be that we will leave behind, respectively are likely to Be Left Behind, by those ahead of us. And we will never have a clue what the hell it’s all about.

    Or worse we may not even have the ability to care.

  2. Jdlaw says:

    All observation – even those through instrumentation – still require some medium of transmission, like light, rays, or particle. We arrive then at a most fudamental truth- there is evidence of an unobservable.

  3. You don’t need cumulative knowledge per se to get to a > human AGI, except as a truism applicable to technological progress generally. What you do need to overcome the conundrum of whether humans can create something smarter than humans is exploration of a couple of things: (a) we create machines more capable that us in other areas and respects all the time so it is not obvious we can’t do so when it comes to what we loosely call “intelligence”; (b) it is not necessary to fully create a smarter than human “mind” but only to create the seed or a mechanism which will recursively improve to such a level.

    I take the opposite point of view from what you posited. I believe that the limits of the human brain impose pretty stiff limits on what we can understand and accomplish unless we overcome those limitations, i.e., create a > human intelligence. Collective efforts among humans are still limited by the limits of individual humans and by the sheer overhead of coordination of efforts of a lot of humans. There is a reason the “design by committee” is not praise.

    Yes, we do a great job of winnowing large collections. Yes machine learning can do much more with the data than we could alone. Yes you can build machines that manipulate larger patterns than we can hold in our head individually or collectively. But the latter is one capability of AGI and a lead to AGI.

    I don’t believe we will muddle along to Singularity doing pretty much what we are already doing. I think it requires AGI to get there.

    It could also be pointed out that a great many smart people are now rather concerned that the rate of human progress is faltering. Personally I don’t see this monitoring several types of technological development. But I do think they have a point as far as applying what is learned and rendering discovery into quite new invention. I also think they have a point when I see mundane things like telework not being embraced just so we can perpetuate the monkey level oversight and in-person communication of grimaces and grunts in their full glory. I see it when a brilliant analysis is not immediately embraced even after it is understood but is endlessly haggled over and bitched and moaned about to turn out a ‘consensus’ that is far more political than intellectual. In short I don’t think technological progress will go on indefinitely without AGI.

  4. I am reluctant to equate the Singularity with AGI because I think it gives people the wrong impression. If, by <AGI you are referring to a technological capability then, yes, that is definitely something the Singularity requires. A technological capability can manifest itself in many different ways and its existence may not manifest in any particular tech but rather emerge out of the interactions of many systems of tech. The problem with using the term <AGI is that people assume you are talking about a particular machine. You know, like 'one day IPad^N' will have a super-smart brain!". Well, yes, one day there may be a super-smart device and this possibility is one of the pathways to the technological Singularity. But there are other pathways in which a super-human intelligence might arise as something distributed across a vast number of systems, for example, 'The Internet Scenario: Humanity, its networks, computers, and databases become sufficiently effective to be considered a superhuman being'. In such a scenario any one of these 'components' (humanity, communications networks, computers, databases) may be unintelligent but nevertheless superintelligence is an emergent property of the interactions going on among all of them working together. But if we make the mistake of believing the Singularity is dependent on AN AGI, we may be comparable to someone on her knees with a magnifying glass, looking for the termite that understands engineering in order to account for the intelligence that constructed the termites' nest.

  5. >I take the opposite point of view from what you posited. I believe that the limits of the human brain impose pretty stiff limits on what we can understand and accomplish unless we overcome those limitations, i.e., create a > human intelligence<

    As I just said, I am loathe to equate the Singularity with A < human intelligence. It seems to ignore what, to me, are the two most promising pathways toward it (the Internet scenario and the digital gaia scenario). Albert Einstein once commented 'my pencil and I are smarter than I am', referring to the fact that we have long been augmenting not just our physical capabilities with technology, but our mental abilities too. Of course I am not arguing that Einstein became a superhuman intelligence thanks to his pencil, but like I said above, in the future an individual might find herself part of a system in which its components (including herself) considered in isolation, are not intelligent but nevertheless superintelligence exists as an epiphenomenon—a characteristic of the system as a whole.

  6. HIya Khannea. Yeah, my primary often witnesses vis niece routinely engaged in activities that were once the soul province of adults. She is not yet ten but already she is doing things like filming and editing her own little movies. When my primary was a child the very idea that the technology for filming and editing movies could be given away as a toy was just crazy, and the expertise required to use the equipment would have made it impossible for any kid save a genius to make much use of it. But my primary’s niece with her ipod touch has easy access to all kinds of apps that allow her to do creative things that were once beyond almost any child. Well, if kids are so augmented these days, just imagine what adult geniuses can achieve with all this technology at their beck and call.

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google+ photo

You are commenting using your Google+ account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )


Connecting to %s