The next time you ask Siri to set a reminder or Cortana to assist you with checking your mails, pause and think about this. Both Siri and Cortana are mere software systems created by human-beings like you and me. Yet, they can not only understand and assist you in your tasks, but also promptly change the flow of their conversations based on your commands! Over the years, we have successfully managed to create applications that can not only satisfy the user’s immediate demands, but also display innate emotional intelligence and the ability to independently adapt themselves to new situations and information.
We’ve already discussed the key differences between Artificial Intelligence, Machine Learning and Deep Learning. But can Artificial Intelligence by itself consist of further bifurcations? We’re back with a blog; and today, let’s have a look at the further classifications of AI:
1. REACTIVE MACHINES: LIVING. IN. THE. PRESENT. MOMENT.
Living in the present is definitely a positive attribute. But does that mean one forgets every event or action that occurred in the past? Well, reactive technology functions somewhat like that. Yes, these machines can adapt to the situations at hand and react accordingly. However, they do not have the ability to learn from previous experiences, let alone apply such knowledge to the present.
An interesting example of a reactive AI machine is Google’s AlphaGo. This game is the online version of the popular board-game Go, and it applies a neural network system to evaluate each player’s move. This game has managed to defeat Go experts by reacting to their moves with its ‘intellect’. But it cannot predict further moves of the player by his/her previous actions, or based on general inferences drawn from the various games played earlier. Another very popular instance of reactive machines is Deep Blue, the supercomputer by IBM that is an expert at chess.
2. LIMITED MEMORY: ‘CAUSE SHORT-TERM MEMORY ALSO GOES A LONG WAY SOMETIMES!
And then there’s another category of AI machines – limited memory. As the name suggests, these machines do have memory, but it’s rather limited. They can base their actions on immediate previous learning, but cannot store learned information to utilize in future. For instance, self-driving vehicles have the ability to read sign-posts, traffic signals etc. and navigate accordingly. They can also convey their decisions to change lanes and follow traffic rules. However, these machines do not have a ‘long-term’ memory and function merely on the basis of their perceptions of very recent instances.
3. IT’S ALL IN THE MIND – THE THEORY OF MIND IN AI
Undoubtedly, the human mind still reigns supreme over all advancements in science and technology. And rightly so, because the human brain is the ultimate source of all inventions and discoveries we so heavily rely on. But…what if we manage to produce technology that can recognize human emotions and predict behavior? While we still have a long way to go in terms of manufacturing such ‘sensitive’ machines, we are definitely on the way. The sentiment analyzer, for instance, is an important feature in chatbots and virtual assistants. With this feature, it is possible for these software systems to recognize the user’s emotional state by his/her tone of speech, and react accordingly. This stage in AI transcends the popular definitions of ‘machine’, thus bringing it (somewhat eerily) closer to human perceptions and behavior.
4. SELF-AWARENESS: WHAT IF TECHNOLOGY INDULGES IN SELF-REFLECTION?
Think about it: What if technology begins to think for itself? Mary Shelley’s Frankenstein is an early description of the chaos that may take place if technology begins to dominate over its own creators. At this stage, technology dons the cloak of consciousness, realizing not only that another being is acting a certain way, but also analyzing the reason behind this behavior. It’s not only about what one does, but also the whys and the hows of the same. And yes, you guessed right. Mankind hasn’t been able to device such types of technology yet. But…never say never, right?
So that was all about the 4 major types of AI technology. While the last 2 types of Artificial Intelligence have not yet been actively initiated on a large scale, technology is rapidly advancing to meet the ever-rising (and never-ending!) human needs. AI technology is definitely taking the world to places – but as always, we know one thing for sure: a proper balance between the two types of intelligence is the sure fire way to go!