I spent most of my working life designing and implementing IT systems. I believed and still do that information technology can be used for the common good if used wisely and appropriately. We are living in the digital machine age where much of our communication, commerce, health and industrial life has become dominated by computers. The existence of the digital world is one of so called "virtual reality" where the real world is represented by digits. This "virtual reality" does not have a real existence but it is a mathematical representation of hard reality. As an example it is possible to record and playback music according to a digital algorithm, this music to your ears can be made to sound almost exactly the same as the original music when it was recorded live. The same can be said for visual media, but in all cases the digital representation requires human interpretation for the musical or visual media to appear real.
When you take a digital photograph of a fellow human being or a natural scene the image is recorded as a mathematical figure; it is easy to manipulate the image to change its appearance or colour and to copy the image. One image can easily be superimposed on another. It is therefore possible to create something entirely different to the original image, but this new image is interpreted by the human observer as being entirely genuine and real. In this way it is possible to fake a new "virtual reality" which is indistinguishable from a truthful "reality".
At the moment it is still possible to tell a fake image or audio presentation but it won't be long before this is not possible. Banks are using voice recognition to permit customers to make financial transactions across their accounts, the voice recognition software sometimes does not recognise a voice especially when the customer has got a cold or is talking emotionally, but very soon machine learning software will be able to identify a voice whatever the circumstances. What happens when machine learning software can exactly mimic a voice? It may not be possible to to secure any voice driven transaction. The same principles apply to visual recognition; so it may not be possible to assure yourself that your voice and image are not being "faked". At present the number of scam and fake transactions is limited; however, the capacity of computers is increasing exponentially, so we could all, become victims of scams be they personal, financial or political. Digital parents: beware a scam paedophile could be coming for your child.
A normal life where we can trust each other and impersonal organisations may become impossible; we might have to resort to going back to face to face transactions and ignore the computers.
Machine learning or AI systems are under the control of humans at present, and are mostly held in check by benign organisations and people; but malign influences could use machine learning to wreck our lives via scams or even kill us by official and unofficial military means. Don't think that this cannot happen: it is already possible.
To guard against the malicious use of machine technology by humans we might choose to develop IT systems which are completely autonomous, and which are not under human control. Machines could be set in motion which only act rationally and logically and which are devoid of emotion and motivation, they would have no desire to harm humans directly. Machine learning machines, however, do not need food or organic nervous systems, they can operate in all weathers and climates. They can grow their networks for as long as they have an energy supply from the sun or nuclear power. Robots could serve them by digging for minerals. Robotic factories could make new machinery to replace and enhance their physical networks. Robot drones could launch their own communications satellites. There would be no need for humans to attend to their needs. They could completely ignore us.
However, we have grown to rely on machines for our needs. What would happen if the machine learning systems decide without menace or motivation to not repair our transport and energy systems? What would happen if they do not till our fields and produce our food? We could all starve; the human race would be set back by thousands of years, so we might have to become hunter gatherers again. The big danger of relying on 'Artificial Intelligence' might not be war but our own ignorance to sleep walk into a world where the machines we have come to rely upon just ignore us. This is food for thought; as no-one is trying to prevent this from happening. Our future could be bleak. "
