As we live in the Information Age, we are all witnesses of the great technological development that our world is going through. Apart from all the scientific and technological breakthroughs, the field of Artificial Intelligence (AI) is growing rapidly and is starting to have a great impact in our everyday life as part of the new industrial revolution. Andrew Ng, who is one of the most influential people in AI, even compared it with electricity from a century ago. He mentions the facts how it transformed industries at that time like communications, agriculture, manufacturing, and transportation by replacing steam powered machines with those using electricity. Similarly, AI is transforming almost all industries, including healthcare, education, finance, maintenance etc. The good thing is that anyone can be part of this new technological revolution.
AI is defined as the theory and development of computer systems that are able to perform tasks normally requiring human intelligence, such as visual perception, speech recognition, decision-making, and translation between languages. Nowadays, its applications are numerous and some examples are: virtual personal assistants like Siri, Google Now and Cortana, recommender systems for books and movies, self-driving cars, e-mail classification and spam filters, advertisement targeting, purchase prediction, computer players in video games, machine translation, autonomous robots, fraud detection, security surveillance etc. After the IT transformational wave where almost everything got digitalized, there is a lot of data to be generated and the data science opportunities are expanding everyday. Having a lot of data with high quality means AI can become better. The field of AI that provides systems the ability to automatically learn from data and improve from experience without being explicitly programmed is called Machine Learning. Practically such computer programs having enough data are getting better than humans at many tasks. There is also the type of AI like DeepBlue, which is considered an AI just because it can play the chess game very good, but in reality such programs only use sophisticated procedures that perform some very complicated calculations. These types of algorithms sometimes are called classic AI, because they are limited to achieving great results in complex tasks requiring intelligence. Even though these types of algorithms are not obsolete and are applicable in a variety of industries, computer systems that are able to learn themselves, especially those based on artificial neural networks, are more prone to succeeding when it comes to building AI.
There are AI systems that can carry out multiple tasks and perform better than humans. In 2015 Google DeepMind developed a computer program named General Atari Player that could learn to play more than 50 Atari-2600 games from just the pixels in the screen without any info given about the games. The same program can beat humans at most of these games, even though they range from 2D to 3D. In some games, it even figured out the best playing strategies all by itself. DeepBlue is different because it was only able to generate and evaluate millions of moves per second in the game of chess. Thus, it was able to find the best moves against chess grandmasters and managed to even defeat the champion Garry Kasparov during the 1990s.
Considering also other breakthroughs like the software system that learned to pilot a model helicopter from just a few hours of observation and the IBM Watson capabilities, there are some concerns raised about the threat of AI. It was the release of AlphaGo that totally caused influential people to talk about the threat to humanity from AI in the future. In the past scientists believed that it will take a very long time for computers to beat humans in the Chinese game Go. In the game of GO, the number of moves is higher than the number of observable atoms in the universe and that is why it was considered very difficult for computers to become better than humans in it, because using classic methods like DeepBlue, it is impossible for computers to evaluate that huge number of possibilities. Google DeepMind proved them wrong when AlphaGo crushed the best GO player in the world, and the way it learned is by just playing thousands of games with itself, which is very fascinating. Theoretically it might be the case that Artificial Intelligence could be dangerous on its own, but as Andrew Ng said, “fearing a rise of killer robots is like worrying about overpopulation in Mars”. We know that we’ve almost reached the limits of silicon chips and without other means of computing, AI is very limited when it is compared to human intelligence. Contemplating about quantum computing and things like Neuromorphic chips then there might be a chance. A real problem nowadays could be the shortage of jobs because of massive task automations happening around all industries. The good news is that AI is also opening more doors than it is closing so we can easily say that AI is about empowering people and definitely not about the rise of machines taking over the world.
AI is definitely one of the most potentially transformative and exciting opportunities that mankind encountered in recent history. It is going to be a key investment target very soon and many organizations are already trying to capitalize on its potential. More than 30 AI startups got acquired in 2016 by tech giants. As businesses are starting to embrace its value, a myriad of other opportunities are just yet to come. This kind of exponential growth is going to create even more demand for AI talent and business opportunities, and right now we can easily say that there are around 20 jobs available for every AI engineer.
I believe that many ICT enthusiasts’ interest about AI should have been sparked already. The good news is that anyone can learn about AI from many free beneficial online courses and thousands of amazing books ranging from beginners to experienced computer scientists. Just very recently a famous AI scientist named Geoffrey Hinton, who keeps contributing to Artificial Neural Networks (ANN) field from the 80s, announced a new type of model named capsule neural networks that has a chance of replacing all other ANNs. A few days after this announcement, many free tutorials about capsule nets got released online and also a few implementations in different programming languages. Anyone can learn and experiment with the latest tech for free and it is quite easy to learn the basics from the online resource and be able to apply AI in real-life projects at some level. Even top influential people in AI like Geoffrey Hinton and Andrew Ng also teach online courses for free.
As we can easily follow up with the latest technological trends, we can try to contribute and change Kosovo and the region for the better. IPKO Foundation provides great support to all the Kosovars wishing to pursue such journey. In my case, it gave me a financial grant to finish the Udacity Machine Learning Engineer Nanodegree program, as I demonstrated my interest and future vision. The skills and expertise gained from such program started to help me gain a competitive edge in my newly co-founded startup named ALGORITHMICS that is about using AI to solve industry problems starting from Kosovo.
To conclude, I believe AI is a really potent means of improving the quality of our lives, as has been shown from the paragraphs above. We are very fortunate to have the opportunity to see this improvement, but also have the chance to be part of endeavors that contribute towards it. There are many free online resources that we can all use and learn from. You can also access even highly valuable resources that are not free from grant supports from the generosity of IPKO Foundation. We are only supposed to believe, have the willingness and the discipline to make contributions in the effort to improve human life on Earth. As the founder of Ford Motor Company, Henry Ford puts it, “Whether you think you can or can’t, you’re right”.
– Muhamed Retkoceri