Artificial Intelligence (AI) is almost everywhere these days. It surrounds us and it is involved in many of the actions we take. However, are we fully aware of this? Sometimes it is difficult for us to see other forms of Artificial Intelligence rather than the ones we see in Westworld and other science-fiction screen creations.
How is AI affecting our day-to-day?
Nowadays we can talk about two main categories of Artificial Intelligence: Narrow (or weak) Artificial Intelligence and General (or strong) Artificial Intelligence. What is the main difference between these two?
Weak AI encompasses all types of artificial intelligence mainly created to carry out specific tasks. From the type of artificial intelligence that we use in our daily lives without even noticing -think online searching tools or image recognition software- to more advanced technology like self-driving cars. This specific type of AI is the one that makes our lives easier because it is fundamentally a replication of basic human intelligence. And this comes, of course, with its restrictions as well.
When it comes to strong AI, the concept of Artificial Intelligence gets wider and more conceptual. We are talking about cognitive abilities, not only primary intelligence. Devices with Artificial General Intelligence (AGI) act as a cognitive agent, having the ability to think and act rationally, using its intelligence to solve problems.
AGI, getting to the next level
To get these concepts straight, let’s take an example of Narrow AI mentioned above and see what its evolution has been with the irruption of Strong AI.
We talked about self-driving cars as an application of the most basic form of Artificial Intelligence -that is the one that basically reproduces human intelligence. Nonetheless, when applying a cognitive level of artificial intelligence to self-driving cars, we get results like the one achieved by Google in 2004: an autonomous vehicle that passed a U.S. driving test.
With progress like this, it is understandable that, a lot of times, we wonder if we are close to live a real-life Westworld scenario. The answer is no. As far as we succeed in setting goals for AI that align well with our own goals as human beings, we can accomplish really successful advances. Therefore, the main and urgent task nowadays regarding AI should be focusing on competence.
Prioritizing AI safety research
For all these reasons, it is important that we now work towards getting the required research on AI safety, taking the prudent amount of time before we fully develop more of its capacities. We are facing a challenge with a totally open book since there is no past record of technological advances like this that we can look up to.
The IoT ecosystem now needs to answer a fundamental question in the AI field: what are the human qualities that will always remain human and will make us keep the power over AI?