Humanity has long been interested in the potential of artificial intelligence, ever since the first computers were designed back in the 1930s.
While humans have natural intelligence, most machines are limited by internal rules. Any machine that mimics the function of human intelligence can be defined as AI, and this technology has improved exponentially in the past few decades. Machines are now faster than us, and there’s little room for human error. However, they don’t improve with experience, and can’t replicate our decision-making processes.
For all the talk of a Skynet-esque future, there’s little to worry about when we consider modern AI. The Terminators aren’t coming for us just yet – but that doesn’t mean we shouldn’t be wary about the next stages of AI development.
Elon Musk has spoken about how AI is beginning to outpace our ability to understand it, which is dangerous when considering implications for humanity as a whole. Musk has also tweeted about the dangers of advanced AI manipulating social media (or the way we think); another valid worry for the near future.
Machine learning is currently in vogue when building an AI system. (While all machine learning is AI, not all AI is machine learning.) A data analytics technique that allows computers to mimic our learning capabilities, the machine uses data to come to a decision, rather than following a predetermined equation. It’s not particularly advanced, but that doesn’t mean that we won’t see progress eventually.
Examples of AI include systems used by ridesharing apps like Lyft and Uber, while aeroplanes are flown using an AI autopilot. Anyone who plays videogames will have probably encountered enemy AI at some point, although limitations are clear for all to see. Image- and speech-recognition are common examples of machine learning, as well as efficient extraction and structuring of data. The transformation of translation services is another example of improved AI that is usable in the present.
👉 Looking to the next few decades, many jobs will be lost to automation, and AI is likely to be the reason machines are suddenly able to do a human’s work. There’s also the prospect of self-driving cars, which would also be powered by an AI system. It sounds like science-fiction, but it doesn’t stop there.
Google employees had to protest to force the company to drop a contract with the US Department of Defense, where machine learning was to be used for military purposes like drone imaging. Anything from facial-recognition tech to autonomous weapons are plausible in coming years; a scary prospect.
AI has both positive and negative connotations, but the ethics depend on factors like who built it, and for what purpose? Considering the nature of AI, it’s hard for a programmer to tell what their creation will do, which is all the more worrying. For now, computers aren’t really that smart, but they are becoming increasingly capable.