Are we facing killer robots or personal shopping assistants?
AI or Artificial Intelligence has been floating around in Video Games, TV and Movies for a long time, but only recently it has been appearing more often in the news. It has become a regular occurrence in the news, however it is rarely properly defined.
Most people have a vision of AI as being a sentient, conscious robotic being, which is a hard image to shake as movies have firmly planted this idea in our minds with films like Terminator, Blade Runner, Westworld, and more recently Ex-Machina. However, this is very far from the truth and even the definition of AI. AI is currently defined and thought of as being a ‘Branch of Computer Science dealing with the simulation of intelligent behaviour in computers. The capability of a machine to imitate intelligent human behaviour’. The most important word in that definition is ‘imitate’. Its not ‘become’ or ‘improve’ or even ‘emulate’, is ‘imitate’. This means that the goal is for AI to become ‘human-like’, not become human.
This misconception of definitions is further driving misconceptions about AI itself and what it can and will be able to do for us. AI has already had massive impacts across many sectors and in the first part of my blog below I have outlined the most exciting and relevant so far.
AI in transport has been all over the news for the past couple years. Tesla and Google paved the way for autonomous driving cars, with Google having clocked over 5 million miles in its self driving car and Tesla reporting over 300 million miles in their autopilot mode.
Due to the early work that Tesla and Google have completed, more and more companies are taking on the challenge of AI transport. Uber and Lyft have been developing self-driving cars to pick up customers, and all these systems have been well received, however the ethical dilemmas posed by AI cars are endless and will be nearly impossible to solve and please everyone. Furthermore, everyone gets up in arms anytime an accident occurs with a form of AI, however the biggest element that is overlooked is the fact that AI is constantly learning. From every minor incident, to full blown accidents, the AI crunches the data to ensure it will never happen again, and while there will forever be an argument against AI transportation, there is no denying the statistics that it is the safest mode of transport.
You can just imagine an entire cities transportation system being a living, thinking organism that works in perfect harmony with one another. No one would ever be late, there would be no traffic, the sheer precision of the system would allow for maximisation of road use and the best part about it, is that all the AI cars now are electric meaning environmental impacts would be slashed.
Entertainment is maybe the last place you would expect to encounter AI but not only is there a big focus in this sector, but it is already being used in applications we use every day. Netflix runs an intelligent AI system to track what you watch and recommend new shows you would like. The same happens on YouTube, however many creators hate the system and claim it hides their content from many of their subscribers.
However, where it is exciting is in the movie industry. A few years ago, 20th Century Fox used IBM AI systems to create a trailer for a horror film called ‘Morgan’ about a synthetic AI human. The IBM AI system watched the film hundreds of times and then selected 6 scenes with 10 minutes of footage that it thought would make the best trailer and then a film editor spliced the footage together to make a 1 minute trailer which you can watch here https://www.youtube.com/watch?v=gJEzuYynaiw
Alongside assistance in editing, they have attempted to make AI write film scripts on very few prompts. The biggest success in this area has been the creation of a short 9-minute sci-fi film called Sunspring. The film itself is impressive given the context of its creation, however if you had not known it was made by a computer system you would think it was the worst film ever made. The script is utter gibberish and the scripted actions of the actors are completely random, including one of the characters producing a fake eye from his mouth in the middle of a conversation. The biggest issue with the ‘AI’ film is that the AI is not operating the cameras or editing the film which means that the reason is comes across as a relatively good film is because of the human effort put into the film. Which is also why it has been received relatively well due to the effort from the human side of the project. However, it is evident that with some tweaking, the film could be good, and does show that with practice AI could be used in the future for some filmmaking but I don’t think its going to be producing the next Star Wars anytime soon…
These are just a handful of the changes that we are experiencing and next week I will explore the other impacts we are experiencing, and the possible future developments of AI
To be continued next week…