I believe Artificial intelligence is becoming increasingly more accessible. When AI was in its infancy in the 1950’s it was described as “astonishing” and only used for basic and mathematical tasks, Learning checkers strategies and algebraic proof for example. At this time AI was very rare and only became more commercially accessible In the late 1990 with AI being used for logistics, data mining, medical diagnosis. This jump in commercial use was due to more widely available and affordable computers capable of running these AI. This increase in availability lead to increased funding for further research. At present AI is used in lots of fields; healthcare, art, video games, military, with AI for Good being a movement to tackle key social economic challenges. The University of Southern California is using AI with the goal of addressing socially relevant problems such as homelessness. While at Stanford, researchers are using AI to analyse satellite images to identify which areas have the highest poverty levels.
In Art and Video games AI can be used in image editing. AI can be used to effectively upscale images. This allows higher resolutions in games without having to have the higher resolution resources taking up space on the hard drive and in art/photography it can help create higher resolution photos from low resolution for example:
AI can also be used to mix photos and videos together. This is best used on faces and face animation. While largely used for good in film making it has also been used maliciously and so shows how AI becoming more accessible to the public is not always a good idea. At the moment AI upscaling is quite advanced however still struggles with artefacts in the picture and faces. In future AI will be able to deal with these better and understand how to work around them.
AI is also becoming more prominent in our day to day lives currently being used for predictive searches to GPS routing. This will only increase given time and the accuracy and variety of the jobs it will perform will increase. Currently self driving cars use AI to help navigate these are often very safe however there have been incidents where the AI has miss interpreted data and caused a crash. As AI becomes more widely used in day to day tasks it will become more effective as it will become a wide network where everything is reporting back and improving the AI. This is called Machine Learning and will let the AI adapt and evolve over time making it increasingly more effective. Machine learning is commonly used in speech recognition and chat bots. This can be both good and bad as it often learns from human interaction. Depending on what demographics it is presented with will shape the outcome of the AI. This was clearly shown to be an issue with Microsoft’s Twitter chatbot “Tay” which was quickly shut down after it began post inflammatory and offensive tweets.
AI is also used in military applications. At present it’s main military use it threat detection and radar. However targeting and decision making AI is being developed and some autonomous vehicles have been created this creates the most risk with potential errors in targeting as it will not have the morals of a human and so may miss identify or target bystanders. As a result I believe the future of AI in the military could be a potential dangerous gateway if appropriate testing and precautions aren’t taken.
