Everybody who has watched any of The Terminator films knows about Skynet, the insidious computing system that becomes self-aware and decides to eliminate humans from the face of the earth. In the movies, a battle plays out, both in the present and the future, between humanity and this fearsome self-aware AI system.
Many who look at the automation of cloud operations, and the introduction of techniques like machine learning to both applications and operations automation, immediately bring up the image of a computing network that organizes into an intelligent being that is much smarter than us, and ultimately sees humanity as adversarial in some way.
People look at the giant data centers being assembled by the likes of Google, Amazon and Facebook, and they hear about changing scale of computing and the accompanying automation that enables it, and they see science fiction mirrored in life. Those that believe in the “singularity,” a point in time where we humans invent an artificial intelligence smarter than ourselves, see the cloud as an enabling infrastructure.
If we cloud computing fans keep at it, we are going to create our own demise. So we should cut it out already.
Systems, like humans, are diverse
But I don’t share that fear. No, I look at human society and politics, the complex adaptive-systems nature of cloud computing, and I fear something much more depressing: the total leisure paradise of the movie Wall-E.
Understanding why means understanding how the evolution of cloud computing is different from the story of Skynet. According to one fan site, Skynet was the result of a decision to build a centralized computer to coordinate a network of devices, vehicles, robots and weapons, aimed at creating the ultimate defense of the United States.
The cloud, however, is going to form the infrastructure for a loosely assembled network of applications, each of which is designed and operated by a different human social structure, such as a business, a computer science study group, a hacker hobbiest, etc. And these applications will be designed and deployed in various technical, political, cultural and economical goals, all of which will create a level of diversity that will make centralized control all but impossible.
Think of our own global political infrastructure. We can’t even agree on the laws we codify within a single town, much less a county, a state—or at any other scale, for that matter. How is it we expect millions, or billions, or perhaps even more, independent software systems to self-organize into a single intelligence?
No, I think its much more likely that our automated software starts to see many of the same problems: infighting over resources, skirmishes over rights to establish rules, even perhaps outright war.
Not annihilation, but service to a fault
The story in Wall-E is different. Instead of seeing humanity as something to destroy, the computers in Wall-E see humanity as something to serve—to a fault. They are built by a mega-corporation that promises amazing leisure cruises in space aboard entirely automated spacecraft. When the same corporation (run by humans, mind you) completely destroys the environment on the Earth’s surface, they decide to pack up much of humanity in these space ships, and kick off a multi-hundred year cruise while the planet heals itself (with the help of a robot or two).
For me, the part of Wall-E that strikes home is the level to which the automation and technology on the ships remove humanity from reality, eliminating any need to work, or interact “off-line,” or even experience the real-world versions of what the ship ultimately simulates for its passengers. Humanity becomes a complacent, bloated, uninspired version of itself, willingly.
Because technology promises to help us at least as much as it promises to hurt us, I believe the evolution of the cloud will be pushed toward the former, not the latter. That’s not to say that military, criminal or even just self-centered elements won’t make true utopia unlikely, but we will likely be served by robots, have our cars driven for us, and even take our next job assignment from a social network, perhaps organized by an AI system. In fact, much of that will probably happen relatively soon.
I can imagine a day when we, as humans, go generations without realizing what is changing in us: our giving up of creativity and experimentation, our lack of physical engagement with the world around us, our education of only what we’ve allowed our technology to tell us about the world around us.
Then again, human society is a complex adaptive system made up of complex adaptive systems made up of complex adaptive systems. Somehow, I think someone, somewhere, will find a way to keep us, well, human. Even in the face of the cloud.
Related research and analysis from GigaOM Pro:
Subscriber content. Sign up for a free trial.
- What converged infrastructure means for the future of the data center staff
- Quality of the cloud: best practices for ISVs
- Migrating media applications to the private cloud: best practices for businesses
