Artificial Intelligence – How far are we?

February 5, 2018 12:00 pm

As you should well know, we are in the age of computers. The world of technology is ever-changing, with innovative ideas and technological advancements coming out almost daily. One such technology is AI or artificial intelligence. This is our attempt at playing God: creating machines that display intelligence comparable to that of humans, with the hope of giving these machines a sort of consciousness. This, understandably, is no easy feat, as the brain is by far the most complex organ, able to process unparalleled amounts of information and communicate that information effectively throughout the body. It’s truly telling how efficient the human body is.

While AI has been prominent in various works of science fiction (e.g. HAL 9000 in 2001: a Space Odyssey), in reality we are still trying to figure out all the complexities of AI. There are obstacles around almost every corner that scientists, technicians, and engineers will have to tackle in order to proceed to the next steps.

There are different levels when it comes to AI technology. Some we have attained, others we’re still working out the kinks. Let’s take a look:

  • Generality – Generality is the concept that a machine can solve any problem, regardless of the discipline. We’ve already achieved this level, more or less.
  • Self-teaching – There are various goals every AI system works to achieve. Being able to teach itself allows the AI system to create its own sort of neural network (similar to the human brain) so as to rely only on itself and not outside sources. This is akin to how humans learn throughout their lifetimes. We have achieved this level in AI technology.
  • Transfer learning – This is similar to the above-mentioned level. With transfer learning, the machine will extrapolate knowledge gained and processes used to solve one problem and implement them to solve a similar problem down the line. Scientists are still working to achieve this across all AI systems.
  • Common sense – This is one of the harder problems to solve. When humans communicate to others, we make inferences based off that information to come to conclusions about the topic. This, however, is hard to translate into code, so we’re still working at it.
  • Self-awareness – This level focuses on the idea of consciousness. Consciousness is the one thing that differentiates the human mind from an AI system, every human has a conscience. It will be difficult to tell if a machine in the future will truly be self-aware, as we would have to solve the Hard Problem of Consciousness first. In the meantime, though, we can program a machine to exude a personality, a sort of superficial consciousness, but we are still a long way off from machines becoming fully self-aware.

While AI seems like the work of science fiction, scientists are teaming together around the world to make it a reality. The ultimate goal of AI is to behave like humans, but at a faster, all around better capacity. The capabilities of AI will be seemingly endless. For example, IBM’s Watson, a machine learning system, beat its human competitors in a game of Jeopardy! back in 2011 and has since been focusing on the field of medicine, specifically diagnosing patients correctly and sooner than human doctors. It’s exciting to anticipate what the future holds in terms of technology.

Categorised in: