The Evolving Face of Artificial Intelligence
Machine intelligence backed by Deep Learning will get better at outsmarting humans
AI growth has dominated most of last year's technology news, from algorithms learning how to play Go without human input, through Elon Musk and Mark Zuckerberg disputing the dangers and benefits of AI to humanity, down to Russia and China declaring it to be a top priority. Looking ahead, 2018 will be a year of developing and deepening all available AI technologies. Deep learning, one of the most important subfields of research on artificial intelligence, will be especially promising.
Kshitiz Randhir Shori, Country Manager- RTB House India talks about AI’s upcoming developments in the marketing sector.
Generally speaking, the goal of artificial intelligence is to make computers as smart, or even smarter than human beings, by giving them human-like thinking and reasoning abilities. Among the many ways to achieve this, machine learning AI is the baseline technique, used widely across all industries. Intriguingly, more advanced approaches are beginning to proliferate.
This is especially true of deep learning—an innovative branch of machine learning that closely imitates the work of the human brain in processing data and creating patterns of decision making. This year it has become a must-have technology in many areas (like healthcare or car automation). From a marketer’s perspective, deep learning has made a huge impact on the entire advertising industry.
Inspired by the biological neurons in our brains, deep learning made it possible to get more reliable, richer, machine-interpretable user descriptions of customer’s buying potential without any human expertise. For example, recently at RTB House we have analyzed massive data sets to show that an AI-based approach can lead to a 35% better conversion rate than the marketer's natural instincts. But deep learning algorithms can achieve even more. This technology is able to predict a user’s unique habits and desires for the advertising industry. It is simplifying our everyday user experience by bringing deeply targeted ads that contain not only products we are more likely to buy, but also those which we haven’t seen or products we haven’t even thought about.
Ultra-precision targeting offered by deep learning has been also used in the advertising industry.Self-learning algorithms can help achieve super accurate product recommendations, while also better predicting the probability of a user clicking on an ad (conversion potential) or the value of purchase (conversion value) to make advertising activities up to 50% more efficient.
The power of deep learning in advertising is the way the algorithms use a massive amount of data and act like humans without specific instructions or rules. It can work inter alia in recommendations used not only by e-commerce companies to persuade customers to buy additional products, but also by other companies to suggest music, events or even dating profiles.
The typical approach to targeted ads (personalized retargeting) is as follows. A user sees a banner creative based on a matter of predefined assumptions: if you checked out black shoes with a gold clasp, the recommendations might want to show black boots with buckles from historical information gathered. With deep learning however, there are no pre assumed rules. Computers learn by practicing what will be the best combination, should it be the next pair of black shoes or a better option – brown sandals, matching bags, etc. The main drive is that no human being has programmed a computer to perform any of the specific actions described above – every display is driven by data and algorithmic learning. In other words, human action is only required at the step of programming the algorithm on how to program itself.
Many brands see the benefit in implementing AI and deep learning solutions into their products or tools. In 2018, we fully expect to see leading companies focus on developing their deep learning AI potential.
From supervised learning to new areas
In 2017 we saw a departure from so-called ‘supervised learning’, a standard approach used by machine learning. Its premise is based on a human giving instruction for computer to learn, taking into account patterns of pre-existing examples, datasets, and answers.
In 2018, AI research will delve into more sophisticated areas, like ‘transfer learning.’ This is a form of deep learning, where teaching a machine is based on various simulations. The machine learns to make decisions using the knowledge gained from many simulations, instead of data from reality. This makes the process a whole lot easier, quicker and cheaper. Using this method, a machine learns to make decisions with logical conclusions, analogy, or deduction by itself.
For example, in older machine learning models, a self-driving car would carry a human and drive for millions of kilometers while recording data. This data would be fed to a machine, which learns how to drive based on driver decisions. But thanks to transfer learning, there’s no need for a physical driver. Instead data can be drawn from hundreds of thousands of simulations, like a driving game. By simulating millions of driving hours, a machine learns by itself how to drive and it can transfer that knowledge into the real world.
The second approach is referred to as reinforced learning. Its purpose is to have machine make the best decisions, based on the feedback it receives from the environment and its actions. For example, it applies to advertisers when bidding to buy ads. Auction systems are very complicated. Even specialists often have problems with determining the optimal rate that will allow them to achieve their desired results at the lowest cost. A machine will also meet similar problems at the beginning of its journey. However, unlike a human being, it can work and bid 24 hours a day in a simulation environment. It can also learn much, much faster than a human. Based on the results of their auction simulations, it can learn how to bet the most effectively, and thus, how to win the auction.
New jobs and new tasks
Deep learning algorithms learn the same way people do. But unlike humans, the machine learns incomparably faster and is able to analyze unimaginable amounts of data. This is where the super human performance factor comes into play. In a very simple way, AI will aim to outperform people in every possible area. Nowadays, self-learning algorithms are able to better recognize the image than the human.
Does it mean that machines while outperforming humans will also take their jobs? Not exactly. According to the World Economic Forum, 65% of children entering primary school today will end up in jobs that currently don’t exist. The current rate of AI development enables more companies to look for more IT specialists, data analysts, programmers. Next year we will probably have a boom of new job offers for data scientists - a position that has not been very popular so far.
2017 innovations will be enhanced in 2018
The ultimate goal for deep learning is to make our lives easier and our work more effective. Therefore, the use of AI is no longer a standard, but a necessity for companies that want to be competitive on a global market. It is not about the ability to personalize or improve the capabilities of a final product, but also about a range of other indirect product activities - such as the collection and analysis of data. Companies now have such a large amount of data to analyze that they do not keep up with processing. This directly affects the decisions made by their employees, and consequently financial results. Companies that are specialized in collecting and analyzing data for various business will play an increasing role. Companies with larger budgets will use AI in turn to suggest what to offer customers, recommend terms to give suppliers, and instruct employees on what to say and do - in real time. It must also be assumed that many new startups will arise soon, offering solutions based on self-learning algorithms, as this technology is going to spread.
There is no doubt that Artificial intelligence has become part of our everyday lives as well as public debate. The coming years will focus on developing AI-based technologies that will replace humans in many difficult tasks, ultimately making our lives a lot easier. But there is still a lot of work ahead of us!
Disclaimer: The views expressed in the article above are those of the authors' and do not necessarily represent or reflect the views of this publishing house
Around The World