Artificial Intelligence (AI) is thoroughly changing limitations of what intelligent mind is capable to memorise. Because of AI’s sudden emergence and a huge pace it evolves in, there is a misconception about it being only a distant, extra-innovative geek thing that has nothing to do with the non-scientific people’s interests. Let us deny this belief and see how Artificial Intelligence can be as useful to our businesses as able to become our co-partner in playing games.
ARTIFICIAL INTELLIGENCE SERVING OUR BUSINESSES
Many companies take up AI in order to reduce operational costs, increase efficiency, grow revenue and improve customer experience. In these cases, AI is inseparable from the Internet of Things, which provides AI with data. Using IoT-enabled information, AI analyses, memorises, learns and thus helps us.
How exactly IoT-provided information contributes to businesses:
- Automated working methods;
- Optimised routine processes and tasks;
- Increase in productivity and operational efficiencies;
- Avoided mistakes and human error;
- Predicted customer‘s preferences using insights;
- Increased revenue by identifying and maximising sales opportunities.
Employee-tracking systems can be a great instance of the IoT-AI duo helping businesses to grow.
Take the employee surveillance system developed in 2013 as an example. Tracking sensors were created by curious students who wanted to know what makes certain workspaces like Google great at building teams.
Once tracking sensors were built in the employees‘ IDs, they collected data about workers‘ movements, speaking tone, volume, speed and other parameters but did not record any conversations and could not associate data with particular individuals, so the whole process is anonymized. By collecting this data, employers got useful insights without any legal issues or concerns from the employees. Collected information showed WHEN and HOW productivity of employees reaches highest figures and thus suggested how employees and the company as a whole, can improve and work more efficiently.
The Bank of America tried the employee surveillance system and the company eventually experienced a 10 % improvement in productivity by making some workplace culture changes.
There are many similar and different-functioned systems on the market, however, not all the managers know about the mentioned variety of AI or IoT possibilities in a business.
Maybe the problem is that some executives have not even seen AI’s advantages over human minds in basic activities and thus they unintentionally ignore obvious AI’s superiority over human mind capabilities in serious businesses, too.
AN EXAMPLE OF HOW ARTIFICIAL MIND SURPASSES HUMANS IN GAMING:
Games have served as benchmarks for artificial intelligence (AI) for decades now. It all evolved from a simple human-computer chess match to what we have right now – an AI that can self-learn and master nearly every game that is available today.
For example, in October 2015, AlphaGo became the first computer Go program to beat (5-0) a human professional Go player without obstacles on a full-sized board. Fan Hui, a professional Chinese-born Go player who lost against AlphaGo, told international weekly science journal, Nature, that humans make very big mistakes, because they are humans: “Sometimes we are tired, sometimes we so want to win the game, we have this pressure. The programme is not like this. It’s very strong and stable, it seems like a wall.”
After that, in 2016, AlphaGo won (4-1) a $1 million match against 18-time world champion Lee Sedol. He is the fifth youngest (12 years 4 months) to become a professional Go player in South Korean history. The match was broadcasted live (with more than 65,000 Youtube viewers on its peak!) and compared with the historic chess match between Deep Blue and Garry Kasparov in 1997, which was a subject to documentary film, The Man vs. The Machine.
If it is not impressive enough because AlphaGo have trained on thousands of human players, maybe the new version of AlphaGo, AlphaGo Zero, which can beat the previous version, will impress you. How the Zero version of Alpha Go exceeded all the old versions in 40 days? Well, that is where machine learning comes.
AlphaGo Zero playing against the best player – itself:
- The system starts off with a neural network which knows nothing about “Go”;
- By combining neural network with a powerful search mechanism, the system starts playing games against itself;
- he neural network is tuned and thus constantly updated in order to predict moves and the winner;
- Updated neural network is recombined with the search algorithm and a new, smarter version of AlphaGo Zero is created;
- The process begins again and each time the performance insignificantly improves.
According to the creators of AlphaGo Zero, DeepMind Technologies, this technique is no longer constrained by the limits of human knowledge as it is able to learn from the strongest player in the world: AlphaGo itself.
We believe that seemingly endless AI’s learning capabilities can be simplified and even compared with the learning process that we manage ourselves – training of pets. In most cases, AI, just like a dog, is provided with many similar objects and situations in which teachers show what is good and wrong (by getting “rewarded” and “punished”). Thus, we should consider AI a friend who can be taught and, in return, help us in different situations, too.
In fact, we encounter AI’s help with spam filtering tools, instant face recognition on Facebook or recommended Spotify playlists every day, so why are we still scared of its’ abilities? Maybe we do not have to thoroughly understand how AI works, but instead, learn interesting, useful parts of AI and use it for our business or daily needs.