Top 6 Examples of How AI & Machine Learning Are Shaping, Shifting and Revolutionizing the Future

     Technology is being evolved continuously at a rapid pace. Smart machines and advanced mobile applications are becoming a daily phenomenon, helping us make faster and more precise decisions.

About 75% of businesses have already started investing in Big Data thus making us rethink about the demand of AI and Machine Learning which is likely to increase exponentially over the coming decade.

         If we peep through the last year (2017), some of the organisations have initiated in spending around 15 percent of their IT Budget in machine learning capabilities. It is needless to state that the number will rise in coming years.

Why We Are Certainly Seeing a Lot of Hype of AI & Machine Learning?

          Digital transformation is taking shape since a very long time and it is bound to grow over the coming years. The Involvement of AI and machine learning now-a-days is on hype due to its wide range of advantages over conventional techniques. The implementation of machine learning have proved to make computing process more reliable, efficient and cost-effective that will finally make decision-making more data-driven.

Here are the top 6 examples that are using machine learning effectively and full fledgedly.

# Voice Recognition Systems like Siri, Alexa, Google Assistant, Cortana and Bixby

         These systems also called as personal assistants use machine learning to emulate deep human interactions. As the use advances, these assistants will learn and make out the modulation and semantics of language.

         Each assistants identifies their own trigger phrases. For example Apple’s Siri identifies the phrase ‘Hey Siri’, while Samsung’s Bixby catches the Phrase ‘Hey Bixby’. All these assistants have their own speech segments in their database when used shows responses that closely resembles real-life interactions.

        Statistically, Google shared of about 25 percent in the personal assistant market in 2017. This is now predicted to rise till 43% in 3 years time.

            

# Facebook

        Prompting to tag your friends in a post has now become an old story. Social network’s algorithms cleverly recognises familiar faces from your contacts lists with means of a jaw dropping AI technology called as DeepFace.

      "DeepFace" a technology that Facebook has implemented to spot the differences in human’s face anatomy with a 97.25 percent degree of accuracy.

      “You normally don’t see that sort of improvement,” says Yaniv Taigman, is one of the genius behind DeepFace. Further he says “We closely approach human performance”.

There’s still more to come from Facebook as they continuously focus on delivering high quality user experience for their users.

# Google

Google introduced a smart reply function back in 2015 which help users handle their inbox more effectively and is time saving. Here the technology used is “Deep Neural Network” which breakdowns and encodes the received email.

Source: https://ai.googleblog.com/2015/11/computer-respond-to-this-email.html

          These network creates and identifies “vectors” for the communication. To well understand the intent, the system comes up with direction vector, for example, “are you ready for the interview,” “are you prepared for the interview,” “Yes I am ready” comes under the same vector and intent.

In this way, these networks works in cycle to decode the meaning behind the incoming message and like-wise suggests three possible responses.

# Uber

         Uber is entirely based on Michelangelo which works on AI and machine learning platform. When you book a cab, Uber’s main task is to calculate estimated time of arrival as quickly as possible. Machine learning helps in this task by precisely analysing data from various previous trips thus giving you best results according to the user’s situation. The Uber model utilizes these algorithms to determine pick-up locations, arrival times and drop times.

         Uber doesn’t seems to stay back with Michelangelo. Surprisingly, Uber has filed a patent for its AI system that can detect drunk passengers.

To know more about the patent click below link:

https://interestingengineering.com/uber-files-patent-for-ai-system-that-can-tell-when-passengers-are-drunk

# Netflix

        Catering over 100 million subscribers, Netflix uses machine learning as the major part for this process. More than 80 percent of TV shows and movies are found through its in-house developed recommendation engine.

       While the details of machine learning algorithms used in Netflix are still not disclosed, Tod Yellin, VP of Netflix states, there are two things that regulates the neural network, user behaviour and programme content. When worked together these neural networks tells the recommendation engine which programmes to display.

# AI powered Cameras

• Computational Photography in Phones

        “It’s about taking a studio effects that you achieve with Lightroom and Photoshop and making them accessible to people at the click of a button,” says Simon Fitzpatrick, Senior Director, Product Management at FotoNation, which provides this technology to most camera brands.

         Single tap beautification mode is now becoming an outdated feature. This is getting replaced by the most anticipated AI technology which plays a key role in creating images that look believable.

• What about AI in DSLRs?

        Users are already familiar with automatic red-eye removal as this feature is provided in almost all DSLRs. Face detection, smile detection where a selfie is automatically clicked when the subject grins. All of these are AI. But the implementation of more advanced AI in DSLRs is still a question.



Written By: Ishan Dodke