Google has been widely reported as making another quantum leap forwards, in terms of developing its Artificial Intelligence programs.
Google’s developers have reportedly developed a “learning algorithm”, which allows its system to learn and adapt while playing computer games!
More specifically while playing Space Invaders- the algorithm learned how to adapt and score higher points during game-play, showing off its Artificial Intelligence capabilities.
Google have named this algorithm and Artificial Intelligence program “The DQN Network”.
Quantum leaps for Artificial Intelligence, and for Google.
What makes this a fascinating development, is the fact that the algorithm managed to improve its ability and performance during game-play, on a number of different computer games.
The system was not pre-programmed, with scripted code, telling the system how to play the game, instead the algorithm was actively learning and developing ways too improve its own level of performance!
The Deep Q Network, was simply provided with pixel input screen information and details about the current game score, during game-play.
The algorithm then had to learn what it could directly influence within the game itself and how to improve its score.
The algorithm showed signs of Machine Learning, by learning tactics and strategies during game-play, which helped it enhance its own score.
“The computer program, which is inspired by the human brain, learned how to play 49 classic Atari games. In more than half, it was as good or better than a professional human player”.-By Rebecca Morelle
Learning how to manage game-play, without any instructions!
We are all familiar with computer games challenging a game player, through running scripted codes, what makes this remarkable is that the system and its capabilities of showing us an ability to adapt to unfamiliar challenges.
Therefore the system is showing the process of thinking and developing itself, so that it improves its performance.
How could this impact Search Engine Optimisation SEO in the future?
We think and predict that the major Search Engines algorithms in the future, will be more under machine control than human control.
What we mean by making this statement is that instead of pre-programmed algorithms stating what makes your website rank higher than another, instead Artificial Intelligence will be taken into account.
This AI technology will take into account, countless variables, which it is actively determining where a website should rank.
What’s fascinating is that Artificial Intelligence AI, is being combined with Machine Learning, these machines will be able to ascertain over a vast scale of factors, which will determine if the search results which have been provided and presented, actually do satisfy a users queries.
We could see these Artificial Intelligence programs, ensuring that websites presented are of higher enough quality, in terms of “user experience”.
This we predict, could be actively monitored, watched and learnt from, in terms of the algorithm learning what sites better answer a searchers queries.
We could see these Artificial Intelligence systems actively learning on a mass scale, if the websites presented do actually satisfy common search terms.
This could happen with pinpoint accuracy and monitoring of minute details, which will allow the systems to understand if certain web-pages, do in fact actually answer a searches queries or not.
We predict then from this information, we may see websites fluctuating in terms of their ranking performance, on a much more regular basis in the future.
We also predict that the systems will be very clever in understanding a websites purpose in the real world and whether content of these sites are actually satisfying what a searcher has set out to find information upon.
What’s for sure is Artificial Intelligence and Machine Learning will have an impact on the major search engines in the future.