Being an underdog has not been a new feeling to Google. When it launched its search engine in the late 90’s, there were multiple other such services backed by big corporations. No one took Google seriously. But who ultimately won the game is no hidden fact. Today, Google is symbolically almost synonymous to the internet itself. But that’s not it, with the fortunes garnered over the years from ‘search’, Google has invested a lot in other technologies. Services and products like Chromecast, which launched today in India are results of company’s such aggressive ventures. The fact that Google Inc acquires two to three companies on a weekly average is enough to show its resolve to expand into several domains.
Although, the most impressive and astonishing field of Google’s interest after search is Artificial Intelligence. At present, they own fourteen AI and robotics companies. It appears as if Google must be using AI techniques to boost up its search potential. But no. In fact, search is being used to fuel its AI, backed on various cloud centers. The system is learning from every query we make, from every entry into the search box, from every relation between query and results. We are feeding the giant. When it is fed completely with the food of the 21st century, i.e. knowledge, it would send its first cute little “Hello World” to us. Or at least that is how people would like to think of it. A monster, but a cute one. Once it is awake, it will help us solve crimes, treat patients, you name it.
Surprisingly, from the perspective of Larry Page, this extension to artificial intelligence was very natural. The founders always knew they were building an AI. It appears as if they have been marching slowly from being an underdog to a major player; handling each nuance with patience.
The AI dream is as old as independent India, or older. Experts in late 50s bet on having machines which could think in less than a generation’s time. However, the task proved more tedious than expected. Progress was made, but very gradually; across various AI summers and winters, terms attributed to exciting and dull phases respectively in AI research. Owing to back-breaking research over the 20th century, the world has in its hands a few essential technologies which can make the break-even possible. The recent upsurge in Big Data handling is one of them. Although in simple language, it is the manipulation of huge amounts of data but everyone seems to be going lady gaga over big data. That because it is actually quiet important, at least that is what the industry thinks as of now. Also, introduction to very cheap computational power has been a boost to the industry.
Google have been the master of scaling. They are not the best with every product they build, but the scale on which they build it is so humongous, the demerits of perfection fade away. Things are finally looking peppy for AI research. With the monetary and academic backing from a firm as stable as Google, we are looking at an artificial future. Only recently did Google acquire DeepMind, a startup based in England which has announced the building of a Neural Turing Machine. A NTM would be capable of combining the potential of a human brain’s neural structure with the design of digital computer. It has already demonstrated its capability by learning simple algorithms on its own.
Another example of Google’s interest is the renovated reCaptcha service. It is a system basically designed to distinguish humans from robots, based on whether they can recognize photographs of cats, dogs and other objects from daily life. What’s special is, that this is another way of feeding the AI giant. The humans would type in “dog” if they see a dog, and the giant would know how a dog looks like. Fantastic. Many would bet on AI becoming the biggest product of Google, overshadowing search as early as in 2025. Hope this summer doesn’t end.
And before I forget, Courage the Cowardly dog had already made a sentient system back in ’95. No one beats him. Long live Sherdil. I dedicate this blog to you.
Although, the most impressive and astonishing field of Google’s interest after search is Artificial Intelligence. At present, they own fourteen AI and robotics companies. It appears as if Google must be using AI techniques to boost up its search potential. But no. In fact, search is being used to fuel its AI, backed on various cloud centers. The system is learning from every query we make, from every entry into the search box, from every relation between query and results. We are feeding the giant. When it is fed completely with the food of the 21st century, i.e. knowledge, it would send its first cute little “Hello World” to us. Or at least that is how people would like to think of it. A monster, but a cute one. Once it is awake, it will help us solve crimes, treat patients, you name it.
Surprisingly, from the perspective of Larry Page, this extension to artificial intelligence was very natural. The founders always knew they were building an AI. It appears as if they have been marching slowly from being an underdog to a major player; handling each nuance with patience.
The AI dream is as old as independent India, or older. Experts in late 50s bet on having machines which could think in less than a generation’s time. However, the task proved more tedious than expected. Progress was made, but very gradually; across various AI summers and winters, terms attributed to exciting and dull phases respectively in AI research. Owing to back-breaking research over the 20th century, the world has in its hands a few essential technologies which can make the break-even possible. The recent upsurge in Big Data handling is one of them. Although in simple language, it is the manipulation of huge amounts of data but everyone seems to be going lady gaga over big data. That because it is actually quiet important, at least that is what the industry thinks as of now. Also, introduction to very cheap computational power has been a boost to the industry.
Google have been the master of scaling. They are not the best with every product they build, but the scale on which they build it is so humongous, the demerits of perfection fade away. Things are finally looking peppy for AI research. With the monetary and academic backing from a firm as stable as Google, we are looking at an artificial future. Only recently did Google acquire DeepMind, a startup based in England which has announced the building of a Neural Turing Machine. A NTM would be capable of combining the potential of a human brain’s neural structure with the design of digital computer. It has already demonstrated its capability by learning simple algorithms on its own.
Another example of Google’s interest is the renovated reCaptcha service. It is a system basically designed to distinguish humans from robots, based on whether they can recognize photographs of cats, dogs and other objects from daily life. What’s special is, that this is another way of feeding the AI giant. The humans would type in “dog” if they see a dog, and the giant would know how a dog looks like. Fantastic. Many would bet on AI becoming the biggest product of Google, overshadowing search as early as in 2025. Hope this summer doesn’t end.
And before I forget, Courage the Cowardly dog had already made a sentient system back in ’95. No one beats him. Long live Sherdil. I dedicate this blog to you.