The Dark Side of AI Part 3
For the final part in this series I wanted to get away from the doom and gloom of A.I. being the end of the world. Ultimately it is this sort of fear that could lead to secrecy and a dangerous lack of transparency emerging amongst potentially condemned scientists. So to counter the problems and dangers associated with AI, I wanted to explore the most abstract and intelligent applications of Artificial Intelligence that is already having an impact on our lives.
Ross, the first A.I. lawyer, has been designed and built atop IBM’s cognitive computer Watson to handle the Baker & Hostetler bankruptcy practice (which is currently made up of a team of 50 lawyers). It has been built to read and understand language, suggest hypotheses when asked a question, and research case law and precedent to back up its conclusions. Ross also incorporates deep learning, allowing it gain speed and knowledge the more you interact with it.
Rather than relying on researchers and experts to find obscure precedents and case law, Ross can read through the entire history to help you get the most accurate information quicker and more efficiently. Ross can even monitor ongoing cases and new court decisions that may affect the verdict of your case! As if we didn’t already have enough lawyers…
A.I. Personal DJ
SoundHound is a music and artificial intelligence company that is attempting to merge the two into a brand new speaker – the Hurricane Speaker. The speaker combines a voice controlled personal DJ/assistant, music recognition software (that allows you to sing a tune to it for recognition), and a vast music collection from which to draw from.
The speaker will be capable of selecting music based on a your mood, creating personalised playlists with its Predictive Analysis Library (PAL) algorithm, as well as providing updates on weather, sports, setting alarms, and generally helping to organise your life.
ResApp Health is an Australian “digital healthcare solutions” company, who have been working on an app that can diagnose respiratory conditions using the microphone on a smartphone (acting like a stethoscope). The app applies deep learning algorithms to analyse cough sounds in an attempt to identify conditions such as pneumonia, asthma, bronchiolitis and chronic obstructive pulmonary disease (COPD).
But this is not an isolated use of machine learning in medicine. Enlitic is using Google’s deep learning open source tech to build an A.I. capable of diagnosing and suggesting treatment in order to help doctors solve medical issue much like a complex data problem. As a test they ran their algorithm with lung CT scans in an attempt to diagnose potentially cancerous growths, comparing it to results given by a panel of world’s top radiologists. Enlitic beat this panel comprehensively, successfully diagnosing every case of cancer when the panel missed 7%, whilst misdiagnosing 19% less cases than the human experts. Survival rates for cancer grow exponentially the earlier it is detected! For bonus points enclitic also helps doctors by showing them similar cases and helping to analyse trends that would be impossible for one doctor to see or consider.
A.I. Journalists Aid
With the associated press already using automation to cover minor league baseball games, it was only a matter of time before A.I. grew into a larger part of journalism. The next step in that growth comes via JUICE, a project funded by Google’s Digital News Initiative, and has been described as a tool to help journalists “discover and explore new creative angles on stories they write”. JUICE is being designed as an add-on to Google Docs, it uses AI systems to analyse what you have written and find creative and productive angles from which to approach the article or story. It is connected to around 470 news sites and automatically runs what they call “creative searches” to pull up relevant articles, cartoons, and multimedia that could be useful to the story. The project is aimed at improving the quality of journalism and helping writers find new ways of approaching their work. The system has had successful trials on journalism students and is expected to be more widely available at some point next year!
Although artificial intelligence can seem like an incredibly scary prospect, it is definitely a tool that has been and can continue to be used to improve many people’s lives and generally be a fantastic aid to the progression of society. However, there is a great deal of caution required in the pursuit of this technology. We cannot allow complacency of the same magnitude that we have allowed in nuclear power, climate change, and cyber security.
By Josh Hamilton