The 3rd article by Rishabh Arora in the Grandmother Speak series is here. Tune in as our in-house experts demystify Artificial Intelligence and Machine Learning concepts to simple, practical and snackable bites, just for you!
Is AI, same as automation?🤔
If yes, why do we have two different terms? If no, then where do we draw the line or where is the overlap? I get asked these questions very often. Let us try to answer some of them here.
As discussed in the first post in this series, AI is intelligence built-in non-living beings (machines in this case). Intelligence, as we know, is the ability to remember past events, comprehend the present, and make predictions for the future. When a machine is able to analyze the data generated from past events as well as a real-time stream of data from current events and create intelligence out of it, we say that it has AI!
After the machine has gained intelligence, we can choose for it to do three things:
- train the machine to take an appropriate action
- take an action ourselves
- do nothing
-1 above is Automation
Time for a Grandmother Example — A typical pizza ordering chatbot has intelligence built in to give customers options on the type of crust, toppings, and considerations like dietary requirements, calorific value, etc. This intelligence has been built in the machine, the chatbot, in this case, using data from past orders, availability of stock in stores, and other business considerations like promotions, product line specific sales goals, etc. This is as we know called AI. The AI grows and evolves over time as machine processes (non-assisted learning) or is made to process (assisted learning) more customer orders and patterns arising out of them. This is ML.💡
Coming back to the example, now once a chatbot has helped the customer find the right pizza, they have three options:
- ask the chatbot to place the order itself
- call the store and place an order
- do nothing or make pizza at home
-1 above is Automation.
Simple isn’t it?
One could argue that even assisting the customer in selecting the pizza is Automation. Yes, that is true but as Automated tasks become mainstream, we start calling then Computational and not Automation.
Time for another Grandmother Example. Spam mail detection was considered Automation till a decade ago, however, is now considered Computational. Another name for Computational can be take automation for granted.
Remember — if my grandmother can learn it, so can you!
In case you missed out reading the previous articles in this series, you can read it here on Medium link:
Katonic is an AI start-up that assimilates data from cross-domain sources & combines machine intelligence with human…
If you enjoyed this piece, we’d ❤️ it if you hit the clap button 👏 so others might stumble upon it.