It’s important for stakeholders in strategic Artificial Intelligence (AI) projects to understand the difference between narrow AI and Artificial General Intelligence (AGI). We currently live in the age of narrow AI.
These are highly specialized algorithms focused on specific tasks. Think playing Go or Chess; or the example in computer vision where algorithms are trained to detect things like cancer. The list of very specific tasks is far-reaching into many fields like finance, insurance, and medicine. Even the development and deployment of self-driving cars falls into the domain of narrow AI.
Remember: Highly specialized for specific, single tasks.
Artificial General Intelligence (AGI) is an attempt to create a tool set which is able to approach any problem much like a human would. It’s about being able to make connections between different domains of knowledge within the knowledge already stored and transferring that learning to make complex decisions; all while absorbing and synthesizing new knowledge. We still lack the computational power to even approach AGI. At best guess, we’re probably 20 to 30 years away.
This is a far cry from the Hollywood killer robot scenario seen in movies like The Terminator. However, experts agree a point will come where a machine will learn faster and absorb more information than a human can. When the machine is smarter than us, we’ll have reached what is known as the Singularity. Then, all bets are off.