Artificial Intelligence 2017

What We Can Expect From Artificial Intelligence in 2017

Brad Zepecki, Managing Partner | October 28, 2017

As the human race continues to improve its technological capabilities, artificial intelligence has come to the forefront of continued improvement. In the past two years alone, dramatic improvements have been made in the capabilities of artificial intelligence and its applications within the real world. On all fronts of education and learning, AI is continuing to surpass itself and grow more and more impressive. Four significantly different advancements can be expected this year in the field of artificial intelligence and the application architecture surrounding them.

Language Learning

This is an area of artificial intelligence that has many far-reaching applications. As machines continue to improve their natural language processing abilities, they can better interact with and help provide direction within a variety of different endeavors. Things like voice recognition can be expected to be drastically improved upon, and it will be possible to apply this ability to many different devices with which people interact on a daily basis.

Another possibility is that artificial intelligences could start communicating with each other in languages that human beings don’t understand in order for them to work faster and more efficiently. Human languages are very crude and inefficient with respect to computational communication, and there have already been instances of AI’s talking to each other in languages people don’t understand. In the future, it can be expected that this to continue and be improved upon as AI’s keep improving.

Generative Adversarial Networks

GANs, or generative adversarial networks, are systems that consist of one network generating new data after learning from a training set, and another network that seeks to tell the difference between fake and real data.

The interaction of these two types of networks can produce some very realistic synthetic data. This type of interaction between systems could be used in the future to help generate better video game scenery and stylistic upgrades to computer-generated designs. It can also be used to help de-blur pixelated footage, which would have huge future implications for security and law enforcement.

Positive Reinforcement

This is an area that could have serious potential in the fields of industrial robotics and automated driving. Reinforcement learning actually takes its inspiration from the way animals learn by associating certain behaviors with either a negative or a positive outcome. An AI can figure out how to navigate a maze by using a process of trial and error, and then associate the eventual positive outcome with the actions it performed. As we move towards a future with more self-driving cars, this could be one of the most impactful innovations to be seen from AI this year.

One example of this technology already blowing up is the Alpha Go program that was created to play the ancient Chinese game of Go. Alpha Go used a system of positive reinforcement learning to quickly master the game, and just recently it defeated one of the most accomplished Go players of all time in a historic match. The idea of positive reinforcement learning was developed well before current types of technology existed, but now, by combining it with deep neural networks, a level of exponential growth can be achieved that could lead to the AI mastery of things in the blink of an eye, no matter how complex or difficult they may seem to human beings.

When it comes to issues facing our species, whether it be how to get rid of trash, how to lower pollution, or how to leave the planet Earth and colonize others, positive reinforcement learning could be the key to extending the lifespan of our species far beyond what in the past was ever thought to be possible.

Startups Focused on AI

Obviously, as the AI industry continues to blossom, many more companies will become devoted to the use of AI technology. As countries around the world continue to improve upon their own systems rather than copying the West, improvements could be seen across the board and all over the world with respect to all areas of AI application.

Countries like China could well begin to catch up or even surpass the Western world when it comes to digital advertising, digital transformation, and voice recognition. A huge boom in AI-focused startups can be expected to be observed during 2017, not just in the US but all around the world.

Artificial intelligence and its application architecture are continuing to grow and expand at an extremely rapid rate. In the coming year alone, massive improvements and shifts in focus toward more AI-based technologies can be expected to take place. As we continue to divert more energy and resources toward artificial intelligence, we could find that the world’s most difficult problems are not as hard to solve as may have once been thought. Also, those who fear that artificial intelligence might outsmart and enslave humanity could find that it will actually allow people to reach their full potential as a carbon-based life form within the universe.

back to blog