Did you know? Artificial Intelligence is one of the top 10 strategic technology trends predicted by Gartner for 2018. McKinsey report on AI mentioned that more than $30 billion a year is recently spent on AI R&D which indicates proliferation of AI-based tools in the forthcoming years.
What is AI? – An Overview
Artificial Intelligence (AI) is an arm of Computer Science that focuses on enabling computers to imitate human behavior and make smarter decisions to solve problems such as visual perception, gesture recognition, decision making, and language translation.
Types of AI:
AI is broadly divided into three categories by their capabilities:
- Weak (Narrow) AI – Typically focuses on a narrow/specific single task with non-sentient machine intelligence. It is the only form of AI that human race could achieve so far.
- Strong AI – Also known as General AI, is functionally equal to human’s intellectuality, can comprehend environment around it just like humans. It’s a sentient machine (with consciousness and mind). It demonstrates cognitive computing capabilities and comprehensive knowledge. It applies intelligence to any problem and solves it as smart as a human can.
Though it’s been said that Strong AI is just around the corner, it’s not yet readily available. Some industry experts say that it might take few decades for us to witness Strong AI in operation.
- Super AI – Machines with intellectual capabilities and intelligence far beyond and surpassing the abilities of a brightest and most-gifted human mind.
Super AI is vaguer than Strong/General AI at this point, and the distance between them is very short.
Now, coming back to Weak (Narrow) AI which is available and in use, below are few areas where it is used and few relevant techniques:
- Computer Audition – Speech and Speaker Recognition
- Computer Vision – Image Processing, Facial Recognition, Object Recognition
- Game AI – Video game console, Computer game bot
- Knowledge Management – Concept Mining like Data Mining, Text mining, Email Spam Filtering
- Machine Learning – Deep Learning, Neural Modeling fields
- Natural Language Processing – Chatbots, Language identification and translation, Question Answering
- Pattern Recognition – Handwriting, Speech and Face recognition
Some AI techniques that help us in our daily chores
- Traffic Algorithms - Web navigation services use specific traffic algorithms to automatically detect destination and suggest us the best and alternative routes with minimum travel time. E.g., Google Maps.
- Text Classification - Every message that we receive in our email is classified based on the content and is categorized as either spam or non-spam and then pushed to appropriate folders like Inbox, Spam, and Promotions, etc. E.g., Gmail.
- Recommendation Engine – eCommerce websites heavily impart these engines and use Item-based and User-based collaborative matrices to provide relevant suggestions. E.g., All eCommerce Websites like Flipkart.
- Prediction Algorithms – Logistic Regression predicts the occurrence of events like rainfall, temperature and dew point, etc. E.g., Facebook. Linear Regression helps in measuring the degree of the occurrence of an event. E.g., Google offers both Linear and Logistic Regression.
- Facial Recognition (FR) – Most of the current generation mobile phones have inbuilt FR algorithm to identify facial patterns as well as the age for better capturing of events. E.g., all current gen phones like iPhone.
- Pattern and Gesture Recognition – Video game consoles use these techniques to attract better participation in games. E.g., Xbox, Kinect.
- Natural Language Processing (NLP) – Chat Bots heavily use NLP to give appropriate answers to the end users. NLP is also used in language identification and translation by search engines. E.g., Chat Bots, Search Engines
- Speech Recognition – Current generation personal assistant devices heavily use this and NLP techniques and act according to the instructions given by the users. E.g., Siri, Amazon Echo, and Amazon Alexa
- Principal Component Analysis – This technique is heavily used in Image Processing in search engines to produce relevant image results at a faster pace. E.g., Google Images
The AI techniques cited above are only a few relevant ones; various other techniques can also be used in the same scenarios.
Few holistic examples wherein multiple Narrow AI techniques are used: Self-driving cars, Search engines, Robots. These heavily implement almost all the available AI techniques to make jobs for humans easier. Especially, self-driving cars use synchronization of many narrow AIs. In all the scenarios mentioned above, AI produces needed insights as the machine is trained with loads of data to predict near-to-accurate values in live situations.
The way forward…
Apart from providing meaningful insights, AI also helps in reducing tedious manual hours spent by automating repetitive tasks. AI has its advantages and disadvantages but if we see the brighter side of it, using AI righteously helps individuals and businesses constructively. Especially, enterprises would have a pool of customer data generated by their CRM; they can analyze the data for meaningful and useful insights and implement relevant and suitable techniques to improve user/customer experience. Enterprises can also help their customers to make smarter decisions which would eventually help them grow their business.
Artificial intelligence for sure is going to create the next wave of digital disruption, and enterprises should be well-equipped with an appropriate AI strategy and investment plan. The early-adopters of AI have already been seeing real-time benefits, and the rest of the enterprises are also moving fast towards digital transformation and adopting AI strategies.