Navigating through the world of technology starts with understanding some key concepts. With artificial intelligence penetrating more and more areas of both the localisation industry and everyday life, being clear on what artificial intelligence really means and how it affects our daily lives can make a big difference.
An article in one of the early issues of our machine translation newsletter aimed to help get to grips with some of the fundamental principles that working with machine translation is based on. Machine translation is a narrow branch of machine learning, which in turn belongs to a broader field of artificial intelligence, or AI for short. Since AI is evolving rapidly, this time, we are going to explore some in-depth concepts related to it, bringing this somewhat misunderstood technology closer to users.
What is artificial intelligence
It comes as no surprise that machine translation is the automated translation generated by a computer, without any human involvement. But how does machine translation relate to artificial intelligence? Undeniably, for quite some time now, artificial intelligence has been a buzzword heard in a variety of contexts. However, it often isn’t crystal clear what this, quite frankly, very broad term actually encompasses.
Disappointingly, there is no single exhaustive definition that everyone would be happy with. AI is often defined as a multi-faceted and multi-disciplinary branch of computer science focused on creating smart machines that are capable of performing tasks that would typically need human intelligence.
Advances in AI are currently resulting in a major shift in pretty much every sector of the tech industry.
As broad as it might sound, this definition focuses on how decision making in certain scenarios can be handed over to machines. AI is not an entirely new concept as it was born in mid-20th century from the visions of two scientists, Alan Turing and Marvin Minsky. Ever since then, AI has been gaining more and more traction.
Artificial Intelligence in everyday life
Examples of applying AI in everyday life include, but are not limited to, virtual personal assistants such as Siri and Alexa, Google Search typing predictions that we all know very well, spam filters, ridesharing apps such as Uber, plagiarism checkers, supporting decisions made by banks to award credit to their customers, facial filters on Snapchat and many many more.
It is not uncommon for banks to use AI-based systems for risk-assessment purposes.
There are different types of AI but the typologies are not all uniform. Some distinguish only two types – weak and strong AI, others three – narrow, general and super AI, and some other typologies go into more detail and divide AI into several different categories.
Apple’s Siri is an example of weak AI. In brief, strong AI, sometimes also called artificial general intelligence or AGI, is a system with generalised cognitive abilities of a human. When facing an unfamiliar task, a strong AI system is capable of finding a solution without human intervention – something that Siri or Alexa would normally address with “Sorry, I don’t know the answer to this question”.
Given the fact there is much disagreement and confusion regarding the exact definition of AI, it might be helpful to look at what AI isn’t.
What artificial intelligence isn’t
As opposed to us humans, AI systems are not quite alive or conscious, meaning they don’t have genuine creativity, emotions or desires on their own; they work only with what we program into them or what they detect from the environment or the data that they are fed.
Therefore, apocalyptic visions of AI breaking free from human control and becoming a threat to the human race are not quite realistic or at least, not when we consider the current shape of AI technologies.
Additionally, AI systems are not always androids and not all androids feature AI – contrary to what some of the science fiction movies would have us believe. Thinking of AI as robots dramatically limits our ability to fully understand what artificial intelligence truly is and what it could become in the near and distant future.
Should we be wary of AI?
Although nobody knows the exact direction in which the advancements in AI are heading, we should not forget that the amalgamation of technologies that it encompasses wouldn’t have been born if it wasn’t for humans in the first place. The brilliance of some of the AI solutions are a direct result of the brilliance of minds of the people who created them.
Quite a few qualities of the human mind, like creativity and compassion, are still out of reach for machines.
Ada Lovelace, an English woman who is considered the first ever computer programmer, had no illusions about artificial intelligence and famously wrote “There are in all extensions of human power (…)” about counting machines she worked on. Artificial intelligence is considerably more human than it might seem at first glance.
Therefore, the path AI will follow in the future will be a direct reflection of human’s input and intentions – reflecting not only their excellence but also their biases and weaknesses.
