An Intro to Artificial Intelligence

Artificial intelligence, or AI, refers to the ability for a computer to think like a human: to adapt and improvise in a new environment and to generalize its accumulated knowledge and apply it to unfamiliar scenarios.

This of course is nearly impossible for machines because humans are immensely complex creatures. We experience new things every second of our lives, with no filter as to what can be perceived. Humans are aware of time and space as a way of defining our existence. AIs define time as a sum of milliseconds. AIs only experience what we want them to experience. Computers experience intelligence through a series of ones and zeros. Humans aren’t binary thinkers. We can examine a wider scope of possibilities. The challenge is manipulating the ones and zeros into acting non-binary through the use of algorithms and neural networks.

Types of AI

AI can usually be split into two broad scopes: narrow AI and general AI.

Narrow AI

Narrow AIs are focused on specific tasks, such as voice recognition or shooting a basketball. Stepping out of that comfort zone would be impossible. For example, Siri on the iPhone would not be able to shoot a basketball and the shooter robot would not be able to hold a conversation. Narrow AIs can further be broken up into two types: Reactive and Limited Memory.

Reactive machines are the most basic type of AI having neither the ability to form memories nor to use past experiences to inform current decisions. For example, IBM’s chess-playing supercomputer Deep Blue is a reactive AI. It can identify the pieces on a chess board and has knowledge of how they move. It can consider the consequences of a move and the potential moves to be played after. But no previous moves from the current game are used when calculating what move to use next. No analysis of the player’s affinity or style is computed. In other words, Deep Blue focuses only on the future with no regard to the past.

Limited Memory AIs are machines that do look into the past. These machines can store memory and use it to make a decision. For example, Google’s self-driving cars detect other cars on the road and store values such as its speed and direction. These data points cannot be calculated in a single instant. Speed itself is a calculation of distance over time. Inherently, time must pass. These changing values are added to the preexisting representations of the world such as lane markings, traffic signs, and curves and included in the decision making process when changing lanes or making a turn.

General AI

General AI is the type of adaptable intellect found in humans, a flexible form of intelligence capable of learning how to carry out vastly different tasks. This is the type of AI found in science fiction movies, such as Skynet from “The Terminator”, which do not exist today, but may be a reality in the future. However, experts are divided over how soon. A survey conducted among four groups of experts in 2012/2013 by AI researchers Vincent C Muller and philosopher Nick Bostrom reported a 50% chance that Artificial General Intelligence would be developed between the years 2040 and 2050, that figure rising to a 90% chance by 2075. However, more recent assessments made by AI experts are much more cautious. They say that society is nowhere close to developing AGI. Rest easy; we won’t be seeing AI supervillains anytime soon. But these theoretical future AGIs also are divided into two types: “Theory of the Mind” and Self-Aware.

The so-called “Theory of the Mind” AIs get their name from the psychology term in which an individual can ascribe mental states to itself and others; mental states such as liking and disliking, knowledge, doubt, belief, happiness, etc. This capacity to at least acknowledge that other beings have their own thoughts, emotions, and experiences is what makes us human. This is the line that must be crossed for AIs to be considered General and Human-like.

The final step of AI evolution is the self-aware machine: the machine which has thoughts, emotions and experiences of its own and can communicate them to others. This leads to the most difficult barrier of all: the understanding of the human mind. Researchers are still understanding new things about how humans think and feel and how the brain makes it all possible. We may never know enough about ourselves to conquer this final hurdle.