Ten ways the machines are taking over the cloud

You don't have to be human to get the most out of the cloud - machines are getting in on the act too

Like the dinosaurs, we humans could be doomed by the cloud as it slowly changes the environment against us. Only this time the carbon based life form seems to be the architect of its own extinction at the hands of the cloud.

Consider the evidence. Humans have long been obsessed with creating intelligent machines that can talk to each other and make decisions - cutting us out of the loop. For decades these servile machines were happy to carry out relatively simple. But now they can learn and make decisions they are running everything.

1) Telemetry was the traditional domain of machine to machine (M2M) communications. A reservoir in East Anglia, for example, too big and remote to be cost effectively guarded, 24/7, by humans, would be empowered by Jabbakam surveillance cameras. These have enough intelligence to detect motion and send an alert back to HQ, via a Hughes satellite network. Machines in the cloud are catching humans, who are innocently going about their business of stealing cables in order to feed their family.

Now, with mobile broadband becoming pervasive, and with tiny gadgets packing mainframe-like processing power, linking to virtual machines, humans are barely needed in running utilities.

2) Thames Water, for example has a system installed by i20 Water that automatically regulates the water supply without any human intervention.

The company’s intelligent hardware nodes are parked at various points on the network of pipes, in chambers and hydrants, each of which is incredibly sensitive to water pressure, flow and temperature. These controlling units gather all this information and feed it back via SIM cards, modems and across GPRS, to i20 Water’s even more powerful cloud based Microsoft virtualised servers in Docklands, which seems to be the HQ of our new cloud overlords.

The cloud based servers run algorithms that calculate their decisions over control of the network. If they detect a pattern that suggests leakage from one section of a pipe, they can automatically lower the pressure and minimize the loss of water. The machines can be 70 percent more effective at cutting water loss than humans.

These controlling units also have the capacity to learn and recognise patterns of human activity (such as rises in consumption when sporting events or music festivals are on). The machines then, are in control of our water supply and constantly learning how to stay in charge.

3) Not content with seizing control of our water supply, the cloud machines are now taking over our media too. If you walk into any public space, cameras are reporting back to cloud computers, which make all kinds of snap decisions about you and take action.

Researchers at Kingston University, for example, have developed an algorithm that makes instant judgments about your suitability for freedom. Images are gathered from one set of machines (video cameras) and this digital data is then fed to computers that rationalise the information.

Your gait, your pattern of movements and your lack of confidence are all quantified. These images are then cross-checked with another intelligent machine that is capable of learning. It can then check whether your general information is consistent with, say, that of a potential terrorist. All that’s missing from the M2M chain of command is a robot that could be instructed to ‘take you out’.

4) If you’re not being singled out as a terrorist or shoplifter, the machines are colluding in the cloud about how to manipulate you. There are intelligent systems such as Neocast, a cloud-based SaaS, which study your physical dimensions, compare them against existing data and automatically stereotype you. They then talk to another machine, a media server, to instruct them on which type of advert to play as you walk past the next plasma screen. So a machine judges me, and decides I’m the type that likes pies, and whacks up an advert for pasties just as I’m nearing Greggs. The machines have taken over the marketing industry now.

Sign up for our free newsletter