Breakpoint, book review: Is the internet really a brain?

Breakpoint, book review: Is the internet really a brain?

Summary: According to neuroscientist and entrepreneur Jeff Stibel, the overgrown internet will eventually collapse to an efficient configuration and develop intelligence.

TOPICS: After Hours, Reviews

Our brains are shrinking, and this is a good thing. Having read, long ago, Stephen Jay Gould's book The Mismeasure of Man, I'd argue it's probably an irrelevant thing: the quality of brains is not effectively measured by their size. Otherwise, elephants would be running Congress and the shutdown wouldn't have occurred.


But this is sort of Jeff Stibel's point in Breakpoint (subtitled 'Why the Web will Implode, Search will be Obsolete, and Everything Else you Need to Know about Technology is in Your Brain'): highly interconnected systems grow like crazy until they reach some inherently unsustainable size limit and then they break. After that, they improve in quality by pruning substandard connections. Those who grew up reading Sir Arthur Conan Doyle may recognise in this idea Sherlock Holmes's belief that the brain was like an empty attic, and must be carefully stocked only with the furniture you actually need (this was his reason for not bothering to remember that the earth goes round the sun).

Now, I'm willing to accept that Stibel, the chair and CEO of Dun & Bradstreet, with a string of past startups to his name as well as degrees in brain science (Brown) and business, marketing and economics (MIT Sloan School of Business), knows more about the brain than I do. But when he starts talking about the web as having reached the digital analogue of this biological breakpoint and saying that we will now improve its quality by pruning the deadwood…well, I wonder which web he's talking about and why. Is he suggesting that the web will run out of space? The far more likely limiting factors are human attention and the availability of energy sources, neither of which is close to running out yet. Besides, one person's deadwood is another person's cute cat video.

Is AI already with us?

Stibel suggests that in future we won't do anything so clunky as type search terms into a box when we want to know something: we will think, and the information will find us

Stibel is more interesting, and much wilder, when discussing innovations like BrainGate, a brain-machine interface that makes it possible to control physical objects via thought — not, as paranormal believers might think, through telekinesis, but via a sensor embedded in the brain that senses the pattern of neurons firing and translates it into electrical signals. Stibel suggests, therefore, that in future we won't do anything so clunky as type search terms into a box when we want to know something: we will think, and the information will find us. He also believes that artificial intelligence is already with us — it's just that we keep changing the rules to avoid having to admit it. I'm going to argue with him on that last assertion: the Turing test hasn't changed since Turing proposed it.

I'm also going to argue with Stibel's conclusion: that the internet is a brain — not like a brain, but an actual brain. One day, as its interconnections continue to increase — perhaps in 20 years — intelligence will emerge, says Stibel. Well, maybe. But my inner biological supremacist says there's more to creating intelligence than connecting up a load of wires and software — even if the number of connections follows the neat little curve that's displayed by every system he studies.


Breakpoint: Why the Web will Implode, Search will be Obsolete, and Everything Else you Need to Know about Technology is in Your Brain
By Jeff Stibel
Palgrave Macmillan
256 pages
ISBN: 978-1-137-27878-4
£17.29 / $28

Topics: After Hours, Reviews

Kick off your day with ZDNet's daily email newsletter. It's the freshest tech news and opinion, served hot. Get it.


Log in or register to join the discussion
  • Obviously doesn't understand what he's talking about.

    For a "brain" each node would have to cooperate with the other nodes for everything. Not just simple communication protocol.

    That isn't what the internet does. Each node on the internet CAN function without the internet. But a brain cannot function unless the functioning of the nodes in distributing computation for a goal. There is no goal in the internet other than allowing each node to communicate with other nodes. The computation accomplished by each node is not really shared.

    Clusters share. Clusters may use TCP/IP... but the goal of that computation is not usually intelligence, unless that was the goal of the programmers of the cluster (as in the development of Watson).

    Unless he thinks intelligence can spontaneously arise out of the routing control flow programs...
    • Exactly

      Its akin to saying all the nerves in your body below the neck are a brain and can spawn its own consciousness.
    • Obviously don't understand what he's talking about

      It seems you're basing your comment on the reviewer's rather simple minded synopsis. You should read the book before you jump to such conclusions.

      And btw, the Turing test has been judged inadequate for some time. After all, some unsophisticated users thought Eliza was a real psychiatrist.
    • I dont know about obviously

      My daughter was born with Dysgenesis of the Corpus Callosum - thats the bit of the brain that routes signals through from one hemisphere to the other - and she has dual aphasia as a result. She's now 20 and because its so rare I've had to study her, and it myself to gain insight into her world. When she was a baby there were only 12 cases known in the UK and no-one could tell us whether she would even survive childhood, and information was scarce. I have Synesthesia myself and do not perceive the world as most do, and have developed a rapport with computers since childhood as well as studying the brain's physiology.

      Today I am working with synthetic intelligence built into interactive toys, after experimenting with smart equipment that prompts her to complete sequences, and with touchscreen speaking devices. BLAIR (bear-like autonomous interactive robotics) started life as a room-guard and scheduler but grew to include face recognition and messaging, so I'm now working on AIME, a humanoid the size of a small child that can make and recognise my daughter's signs, work with PECS and Makaton tiles, play simple matching games, and act as an advocate and companion.

      AIME isnt that smart, but to make her as smart as I have has required some lateral thinking. She learns much as we do, trial and error, and requires a pretty large database to hold everything her cameras can differentiate between. The mechanism is based much like our brains are, in sections - she has equivalents to memory, a visual cortex, a language centre and most importantly a hypothalamus.

      An information system without some kind of driving mechanism - an attention - does nothing. It stores information, and can be directed to extract extra information by interpolation but it has to be directed intelligently before reason can occur. This is the difference between animalistic reactive intelligence and what we define as full intelligence. If the Internet developed a processing hub that was able to reason, and act as a hypothalamus, then we'd be in trouble I guess. But until then we are the internet's hypothalamus and I think its much more likely to become an addendum and interconnect to our own neural processing power rather than an intellect in its own right.

      Thats the trouble with Turing, it seeks to test a human intellect inside a machine, and the only way we can do that is by either building a machine that simulates or mimics human knowledge architecture and its contained knowledge. Turing tests anthropomorphic thinking rather than true intelligence: If an alien contacted us from beyond our known stars, and we applied the Turing Test to it's responses, it would very likely fail to convince us it was another human being regardless of how intelligent it is.

      Artificial intelligence has largely fallen out of favour and been replaced by synthetic intelligence systems like the ones in your digital camera that sense scenes, recognise faces and adjust for red-eye or glass reflection etc. They have become so second-nature we dont even recognise them as machine intelligence any more, but they are. And they are getting better at taking photos than your average human is, quietly and efficiently learning what we like to see.

      I try to take a more open view on what intelligence is, rather than define what is intelligent.
  • Kneel before Zod

    I for one welcome our new computer overlords.
    Robert Hahn
  • Arrrrrgh

    SkyNet!!!! Run away!