Technology has affected our daily life to the point that we can’t live without them. They are essential in communication, industry, and even in our households. Technology has evolved to the point that we have made programs and applications that learn and grow as time goes on. However before we can talk about robotics today, we need to look at where it began and how it’s changed.

With the introduction of technology came the introduction of fear, scepticism and amazement. With the introduction of the first programmable “Computer” named the Colossus in 1942 it was clear that technology was going to play a role in at least the manufacturing industry and with the first working AI programs being created soon after in 1951 on the Ferranti Mark 1 in the University of Manchester. The first AI was a chess game, but the computer was not powerful enough to play a full game, it could only find the best possible moves, looking only 2 moves ahead. A full chess game could be played by a later program written in 1958 by Alex Berstein.

Artificial intelligence along side with technology experienced rapid growth with each passing year discovering new ways to interact with the system and new formulas to allow the code to run. The new discoveries seam to be never ending and with exponential growth the question occurs, at what point will technology surpass us, this point of technological singularity. Nobody knows what will happen after the singularity, or when it will happen. It is dangerous and surrounded with mystery.

One of the ways we have gotten so close to the singularity is by means of machine learning. Machine learning helps find the best fit way of a problem, many cases the solution is better and years ahead what a human would have found. A more advanced form of machine learning called neural networks. Neural networks act more similarly to a brain, it gathers data and makes decisions based on them. The network then the results are the feed through the system to change the neurons and alter future results for the better. Neural networks could be used in many applications to improve upon what humans have accomplished.

Neural networks could help with the medical industry. They could observe and perform medical procedures on patients more accurate, quicker, and more cost-effectively than a human counterpart. In some cases, they are already in use! For example, in the radiology field, neural networks are observing tumors very accurately. Because of this medical treatment for medical issues that can be diagnosed and treated by technology has had cheaper prices than if it had been done by a doctor.

Neural networks could be tomorrow's ray tracing technology, not only by making them look better but have the physics and entities in them act more realistic rather than floaty and inept. In order to achieve this, it would be required that the technology be accounted for in the hardware to achieve better performance on the user's behalf, as well as the difficulties in training such a technology. We already see this technology in some industries, such as the gaming industry with “Super Smash Brothers Ultimate” on the switch. This game has amiibos that you can buy and fight against in the game. At the conclusions of every match the amiibo will adjust to better suit the play style of the player with the goal of winning.

Facial recognition, fingerprint readers, and other biometric scanners are pathing the way for an easier tomorrow. With this ease of access to our devices come a wave of people that wish to infiltrate other security systems. If a neural is in a majority of the code it could create flaws, exploits in the code in order to gain access to the system. One could create a neuron with a weight so great that a skews any true neurons that are actually getting the information.

More encryption helps alleviate this problem somewhat but for neural networks to be introduced into A larger system, but this has not stoped its growth. Neural networks have also allowed computers to have “fuzzy logic”. Fuzzy logic is defined as logic that does not have a true yes or no answer, The logic could be 60% true and 40% false. An example is the chance of rain given the variables.

Companies such as Tesla, Google, and Toyota have self-driving cars with an array of sensors in and on the car. The sensors will send data to the brain of the car which contains a neural network to decide what the car will do in these decisions. These cars have one problem, The Trolley problem, the problem of choosing who dies. As these cars are becoming more prevalent the trolley problem will occur more often.

Author's Bio: 

I am a computer science professor. Being a tech enthusiast I keep close tabs on trends and will be glad to share and discuss the latest wrapups in the field with the community.