Imagine if your phone could call up your near and dear once it senses that you are not being normal in terms of the activities you do or in terms of your dexterity or mobility ? Imagine if your computer could call the tech support chap since it senses that your hard drive is going to crash ? Welcome to the world of Artificial Intelligent (AI) Systems .
I read a gartner report that 85% of customer interactions are projected to be managed without a human by 2020, I also read that more than 1.6 B USD has been invested into AI uptil last year. AI has been around for a very long time ( right from my engineering school days) but of late has been garnering a lot of interest. I have to hazard a guess and state that it’s because of lower cost of computing and storage including cloud computing and big data concepts.
AI can be referred to as the brain , an intelligent agent in an autonomous entity. AI typically encompasses machine learning, natural language processing and understanding, deep learning , context awareness and image recognition and processing. As you can probably envisage , AI can potentially be an extremely powerful tool to help improve quality of life if data privacy and other considerations are managed well.
3 top perspectives about AI which you may want to keep in mind as the industry goes ga-ga over this technology ….
Machine Learning is not the only aspect of AI , machine learning is an algorithmic ability to learn from prior data to produce a desired behavior. Typically this would involve showing the algorithm a set of situational data and telling it what the right decision is. Once the model has been trained, you can feed in new sets of data into the algorithm and have it make decisions for you. Its like playing the guitar – you learn the main chords and once you do that, you can try and play songs …. Another example could be using various EKG’s in an algorithm and teach it to make the right doctor intervention, Over a period of time, the algorithm can learn to read any kind of an EKG and do a cardiac event interpretation.
Deep learning is an offshoot of machine learning …it also involves something which all who did engineering may have heard of – nerual networks. It is a form of biomimicry where algorithms are inspired by the way neurons work in the brain. Finding insights and meaning based on mutiple layers of artficial neurons. Building layers of insights one upon another until you get a breakthrough insight.
Natural language processing and understanding is probably the trickiest as this involves making sense of how humans communicate and buiding that perspective into the system. A word like “ that’s interesting “ could mean many things based on context of usage. For example – somebody could really mean that something was interesting , while somebody else could be using “that’s interesting” to mean that its puzzling or something strange or peculiar. The ability to capture context is one of the biggest challenges in NLP since teaching algorithms to understand context is a challenge.
I really think we will get to a situation very soon where low level support activities get automated through AI and over a period of time high level insightful activities maybe performed or assisted by AI systems.
Here’s wishing that my washing machine calls up the support service on its own once it senses that it has a problem sooner than later :) and saves me on my housework and chores.
Have a good rest of the week.
Comments