2016 is the Year of AI

16th February 2016
2016 is the Year of  AI

Sahil Sood, author of the book, “Back to  the Future”,  suggests why  AI  could find its tipping point in 2016

The year 2015 saw a large number of notable developments in the technology sector. The coming years would see a revolution in the era with most of us relying on science to perform our day to day work from buying grocery to controlling our homes with intelligent devices.   Something that was a figment of imagination few years ago, is taking a definite place in our everyday lives. 

One of the major technological trends to look out for in 2016 would definitely be Artificial Intelligence. Despite being over sixty years old it continues to be intriguing even today. It is poised to become ubiquitous in our society in the coming century, with applications in image understanding, mapping, medicine, drones, and self-driving cars. This trend is expected to continue in 2016 with companies like Google, Amazon, and Facebook making huge investments in it.

Artificial Intelligence (AI) is an idea that has withstood the test of time. It continues to taunt our imaginations from time immemorial the iconic duo of C3PO and R2D2 from the Star wars franchise.

AI at its very core is all about mimicking the human learning process. But learning requires data.Gigabytes and gigabytes of data and data at that… But how do we get such large amounts of data? Well data can be acquired from the internet, case studies, government records, search engines, past experiences etc… AI adds an intelligence layer to such large amounts of data to process complex analytical queries much faster and much more accurately than any humans ever can.

AI systems make use of a host of sophisticated training algorithms like machine-learning, probabilistic models, Case-based reasoning (CBR) and neural networks. Often more that, one learning paradigms are used in tandem to provide better actionable results.

We see the key word in any AI systems in data analysis. And the exponential growth in the technological capabilities and processing power (courtesy Moore’s law) of modern computers has assured us that reality might soon match the rhetoric. Data is increasingly tagged and categorized using sophisticated tools developed for the sole purpose of converting raw data into usable knowledge.

The mere collection and categorization of data used to incur an enormous cost earlier. Almost all information regarding our day to day needs can be found on the internet. We inadvertently contribute to these datasets by participating in discussions on social forums and thus leaving behind our digital footprint. Leveraging these datasets machine learning algorithms can be trained at a much lower cost. Given the plummeting costs of storage and increasing options for parallel data processing the intrinsic appeal of intelligent application continues to grow unabated. Therefore it may be fair to say Artificial intelligence provides one of the most enticing opportunities of the current decade.

The enigmatic Turing Test
The Turing test is a test of a machine's ability to exhibit intelligent behavior equivalent to, or indistinguishable from, that of a human. The test was introduced by Alan Turing in his 1950 paper "Computing Machinery and Intelligence," while working at The University of Manchester (Turing, 1950; it opens with his famous words: "I propose to consider the question, 'can machines think?’ The Turing Test is quite popular with Hollywood. The humanoid Ava undergoes a modified version of the Turing Test in the 2015 film "Ex Machina."

In 2014 a computer program developed by a team of Russian computer scientists fooled one in three people on a team of judges, into thinking that it was Eugene Goostman, a 13-year-old Ukrainian boy. Though some might cite it as being a stunning example of the Turing test being cracked, stalwarts in the field of AI are quick to point out that the original Turing test hints at a much higher level of conversation between the human and the bot, one which Goostman can never hope to achieve. Therefore the Turing test continues to be as evasive as ever computers have continued to become more intelligent with time.

Recent Development in Artificial Intelligence
Search giant Google is also ready to cash in on the Artificial Intelligence phenomenon. In November 2015, it announced TensorFlow, its open ­source platform for machine learning. Multiple Google products harness TensorFlow.

Tensorflow is one of the best applications for deep learning in general. Deep learning is a machine learning model which helps classify and cluster data with an innate accuracy. TensorFlow goes beyond traditional realms of deep learning. It makes use of data flow graphs. Multidimensional data arrays called tensors are passed along in the nodes of graphs to compute complex mathematical operations. The official website of TensorFlow provides tutorials to both beginner and expert level of users on using the TensorFlow software along with code snippets in Python.

Thus, open-­sourcing TensorFlow allows researchers, grad students the opportunity to work with professionally-built software.

As the official Google Blog puts it: "We've seen firsthand what TensorFlow can do, and we think it could make an even bigger impact outside Google. So today we’re also open-sourcing TensorFlow. We hope this will let the machine learning community—everyone from academic researchers, to engineers, to hobbyists—exchange ideas much more quickly, through working code rather than just research papers. And that, in turn, will accelerate research on machine learning, in the end making technology work better for everyone. Bonus: TensorFlow is for more than just machine learning. It may be useful wherever researchers are trying to make sense of very complex data—everything from protein folding to crunching astronomy data.”

Facebook is all not very far behind. The Facebook AI Research team more popularly called (FAIR) is also working hard to enhance our lives through the use of Artificial intelligence.

Though we might not be making super-intelligent droids like Ava any time soon, we are definitely taking steps in the right direction as proved by the Scientists at Nanyang Technological University in Singapore. In December 2015 the researchers at NTU unveiled "Nadine," a socially intelligent, android complete with her own personality, mood and emotions. “Nadine” was presented at a new media showcase at NTU

 “Nadine” has been created to be a doppelganger of Thalmann. She has the capacity to show visible traits of human emotions like happiness and sadness along with the ability to recall a person she has had prior contact with. Algorithms similar to those used in Apple’s Siri or Microsoft’s Cortana provide the required social intelligence to Nadine.

AI takeover refers to a hypothetical scenario in which artificial intelligence (AI) becomes the force on Earth, with computers or robots usurping control of the planet.

Stalwarts like Stephen Hawking and Elon Musk have also highlighted the threat highly intelligent machines pose to our very existence. Despite the apparent pitfalls,  it is a common consensus  that  Artificial intelligence has a long way to go before we get to machines that can pass off as human, but that's not necessarily the kind of AI that we should be worried about. While we spend our time debating about "Strong AI," i.e. artificial general intelligence indistinguishable from human a more realistic threat  to our current society would be  "Weak AI" gone rogue. The principle behind Weak AI is simply the fact that machines can be made to act as if they are intelligent. Specialized systems, like those used for maintaining high-frequency stock trading (HFT) could be crippled by persistent attacks.

AI for robotics will allow us to address the many challenges the aging human population now face.

 Most of us already have our own personal assistants on our smartphones be it Apple’s Siri or Microsoft’s Cortana. The trend is expected to continue in the future with better versions of these software’s being created. In near future we might have AI’s in the form of butlers and housemaids.

 The health industry is probably going to apply machine learning tools to provide personalized medications. Everyone’s genome would be sequenced and medical records would be scrutinized in details to find possible patterns. Such large amounts of data would help health facilities to carefully examine individual problems and provide specialized treatments dedicated to the requirements of the person concerned.

 Another popular theory is the possibility of the combination of artificial and human intelligence to produce super-humans. In the near future systems powered by AI might augment reality by giving us sensory abilities far beyond our natural powers of vision, hearing, and manipulation. 

It could help us in providing relief work in case of natural disasters which make a terrain inaccessible for human navigation (example, earthquakes, avalanches etc…). It could also help in case of man-made industrial disaster like gas leaks, nuclear meltdowns.

 Apart from benefits, AI’s might take over many of our jobs by the next century. With specialized AI’s being developed chances are AI’s will replace many of us if don’t stay ahead of our times.

Final Note
Despite what cynics say, we are far from machines taking over anytime soon. The benefits of AI systems far surpass the possible threats it possess.

The $650M DeepMind Technologies acquisition by Google in 2014 re-established the notion that there is money to be made in artificial intelligence investments. 2014 saw a 302% increase in the funding of artificial intelligence startups. With a plethora of venture capitalists placing their bets on AI based startups; Artificial Intelligence is definitely the place to be in 2016.
Feb 16 2016

Sahil Sood, is a technology and business performance professional who lives in Mumbai. An information technology graduate with Masters in His debut novel – “Back to the future” can be found  here