The Four Horsemen of the Datapocalypse

All of us are affected by four accelerating and converging megatrends: pervasive computing, big data analytics, artificial intelligence, and cyber hostility. 

I have observed and participated in many of the developments in information and communication technology (ICT)  over the past 40 years, as a philosopher and cognitive scientist, and (more recently) as a software architect.  One overriding feature stands out: the rate of technical advance always accelerates.  The power and complexity of ICT is not only growing, it is growing more rapidly every day.  The rate of change is exponential.

Meanwhile, our human brains and bodies deal with change in the same old biological fashion, as dictated by the DNA blueprint we’ve had for hundreds of thousands of years.  We tend to respond to what we can see right in front of us.

We can’t always see the biggest dangers before its too late.  Advanced technologies not only change much more quickly than our basic biological capacities, but they are much more complex than technologies we have developed in past centuries.  Their interactions are intricate and the implications of those interactions reach us in every corner of our lives.

The greatest risk we face is no single threat, it is the convergence of several IT megatrends.  They are highly interrelated and becoming more so all the time.  It is the confluence of these four categories of technological development that constitutes the major threat. I call these converging megatrends the “four horsemen of the Datapocalypse.”

  1. Pervasive computing:  Regular old computers, social media, phones & tablets & other mobile devices, the cloud, internet of things, wearables, robots, drones, 3D printing, radical miniaturization, etc. We are increasingly immersed in a digital ocean.  It’s an ocean that is artificial, but completely real – and it is constantly changing and growing, enveloping every kind of technological advancement, including those that are inherently biological and genetic. That ocean is made of data.
  2. Big data analytics:  The enormous trails of data that each of us creates (thanks to pervasive computing) are continually sliced and diced, mined and analyzed, and shared throughout cyberspace at the speed of light.  And they never go away.  Our activities, our interests, our communications — virtually our entire lives — are available for intimate inspection by any intelligence sufficiently connected to the data centers and their channels of communication.
  3. Artificial Intelligence (AI): 

    AI is closely related to big data analytics, because both categories involve sophisticated software algorithms.  But AI takes them to a different level.  Instead of using the algorithms only for clearly specified and defined purposes, AI has the additional goal of incorporating learning and changing capabilities into the very structure of those algorithms.

    So rather than being confined only to what can be programmed into the systems initially, AI algorithms learn from their digital experiences.  They become more and more capable through continued refinement of their processing activities. AI, when combined with big data (including the virtually infinite resources of the internet) and with pervasive computing, can achieve virtually unfathomable intelligence capabilities.  We can barely begin to conceive of an intelligence as far beyond anything human as ours is beyond that of an ant.

    Moreover, the nature of artificial intelligence is profoundly alien, relative to our own. Human intelligence evolved along with human biology, but artificial intelligence is built around fundamentally different principles and goals.

    And the power of AI is still in its infancy.  Every year, more and more extremely smart people are mastering the existing techniques available to AI programming and, in universities, in corporations, and in military research facilities all around the world, AI scientists are developing newer and more powerful techniques.

  4. Cyber Hostility: 

    Information and communication technologies will never be universally peaceful, beneficial, and non-criminal.  They are among the most multi-purpose of all human toolsets and, like all tools, their use reflects the purposes and motivations of the people behind them.  As long as there are powerful interests competing for scarce resources or diligently working to conform the world to their own visions and ideologies, ICTs will be used for purposes of hostility (whether interpersonal, military, or criminal).

    Values and interests are inherently at odds, and people will use the most effective tools they have to advance their interests.  The more readily available and cheaply deployable the tools become, the more rapidly they will proliferate, and the more likely it is that they will be employed with hostile intent.

The combination of these four megatrends is the greatest threat.  They are converging rapidly; each one contributes to the scale of the danger posed by the others. We should not think about any of them or assess their potential dangers in isolation, because they are becoming literally inseparable.

For example, malware is continually becoming more sophisticated and difficult to detect (see my post “Malevolentware“).  As AI capabilities get folded into the basic structure of hostile software, it will become more and more able to detect and even predict defensive tactics.  It will attempt to hijack the efforts to protect systems and it may even piggyback on the processes used to monitor network security.

In addition, the exposure that individuals and companies face is escalating dramatically as everyone becomes ever-more dependent on the internet and digital devices.  When the loss of individual privacy is added to the mix, as is virtually assured by the availability of big data analytics, the result is that there is no place to hide. 

The “four horsemen” could be characterized according to their respective threats, which become more and more dangerous to humanity as their effects combine and each one reinforces the menace from the others.

  1. Pervasive Computing: No escape
  2. Big Data Analytics:  Nowhere to hide
  3. Strong AI: No contest
  4. Cyber Hostility: No morality, no mercy, no kindness, no quarter

Whatever risks that advancing computer technologies may pose, we can be fairly sure that they will not be isolated to any one of these converging trajectories.  The discussions of technological risks that typically focus on specific types of technology, rather than the cross-pollination of ideas that commonly result in new inventions and which are also most likely to produce unanticipated consequences.

(revised 5/10/2015)

, , , ,

Leave a Reply

Your email address will not be published.

This site uses Akismet to reduce spam. Learn how your comment data is processed.