by Dr. Ricardo Baeza-Yates, NTENT
Change is always difficult. Especially when presented to a species advanced enough to recognize it as it’s happening. To the human mind, the concept of Artificial Intelligence swings a pendulum from fascination to fear. Will the inevitable, apocalyptic prophecy of machines taking over the world lead to the end of civilization as we know it, as depicted in countless science fiction scenarios to date? Not exactly.
Artificial Intelligence is here, happening now, intertwined with our daily lives, unbeknownst to most of us who recognize it as a convenience. AI is actively changing civilization as we know it, but don’t panic; (most of) it is for the better, and still requires the human element of actual intelligence throughout the process.
The fascination with Artificial Intelligence comes from its ability to redefine the human experience and make life easier. The fear stems from the possibility it makes life so easy we become extinct, in one form or another. Both arguments hold water in theory, but currently in practice, AI is improving the quality of human life in more ways than it causes harm.
Using Deep Data Science, Machine Learning and Natural Language Processing (NLP), Artificial Intelligence continues to enhance every industrial sector. Healthcare, security, travel, real estate, digital search, education and research have all seen increased efficiency, with more growth and progress expected in future trends. However, we can’t ignore the virtual elephant in the room and the prospect of lost jobs as automation from chatbots and computers continue to out-produce humans. Nothing frightens civilization more than a looming threat to survival. The real future of AI depends on how well humans will be able to adapt to an evolving world. Adaptation starts with understanding what Artificial Intelligence is, identifying its future trends and preparing for the cons that come with it.
Deep Data Science, Machine Learning and Natural Language Processing (NLP)
With each keystroke entered in the web, an element of data is born. Vast amounts of data processed on the web between all machines and devices require the development of systems that can automatically process large data sets and perform automated transactions. Artificial Intelligence relies on applications to help computer systems learn and predict user behavior without being specifically programmed to do so. The most visible face of AI technology is Machine Learning, particularly deep learning (based in neural networks) that learn to recognize objects in images, drive a car, or defeat the best human player in challenging games like chess.
Using algorithms that work with structured and unstructured data, Machine Learning captures and understands data transferred across trillions of points per second, from Internet traffic, web search queries, content, ad-clicks, and more. Machine learning allows for the understanding, interpretation and prediction of behavior as a way to take specific action. Data Science is the art of interpreting this data to glean valuable information and insight from a user base. Deep Data Science represents the point where deep Machine Learning and Data Science intersect.
Through the combination of Machine Learning and Deep Data Science, AI is able to understand human behavioral patterns on every level, then provide solutions according to those patterns. As a result, businesses across all industries, at every size, are able to see what information individuals want, compare it to mass amounts of data and deliver appropriate, relevant results.
The current and future impact of AI on five key industries
Healthcare – The World Health Organization (WHO) estimates the global needs-based shortage of health-care workers could exceed 14 million by 2030. Artificial Intelligence applications can drastically improve many areas of healthcare, freeing up doctors and medical personnel for more critical cases. AI Health Assistants use NLP and Machine Learning algorithms to help identify symptoms, suggest basic treatments for non-threatening illnesses and provide a personalized tracker that sense pattern deviations and warn to take further action.
At Stanford University, a team of researchers created an automated classification of skin lesions using deep convolutional neural networks (CNNs) to overcome the challenge of pixel and grain issues that come when viewing these images with the naked eye. Using datasets containing thousands of images and diseases, using pixels and disease labels as inputs, the team compared the automated results to those from 21 board-certified dermatologists who diagnosed similar images. Results showed the CNN in accordance with the doctors, indicating an AI capable of identifying skin cancer could potentially reach mobile devices, extending far beyond the waiting room door, helping millions of patients in the process.
The opportunities for AI in healthcare are endless, from predicting heart disease to analyzing data sets of molecular structures that aid in pharmaceutical drug development. But such sensitive information walks a fine line between ethics and privacy, not to mention the potential loss of administrative jobs, consumed by automation.
Security – Artificial Intelligence is being used as a form of protection, from evils in every direction. From hackers to home burglars, security requires extensive visibility into a given situation at any moment in time. This results in exorbitant amounts of data incapable for humans to process efficiently. Fraud detection technology, facial and vocal recognition arm the cyber space and Internet of Things, while physical security measures used by law enforcement such as drones, ground-based robots and body cams contain smart software with deep learning-based video analytics solutions.
Legal – In an industry reliant on paperwork, Artificial Intelligence is a gift. Natural Language Processing can sift through thousands of documents at a fraction of the pace of its human counterpart, with less error. AI tools can help lawyers scan historical content for precedent or evidential testimony, and analyze internal data to help define billing patterns, case wins and streamline operations.
Automotive – 2016 accounted for the highest number of motor vehicle deaths in the United States since 2007, according to the National Safety Council. At 40,200 deaths, the total cost including injuries and property damage reached over $432 billion. One of the most prominent uses of AI technology is for the development of self-driving vehicles, with its most valuable benefit being the estimated increased level of safety. AI is used to learn human behavior and react to driving conditions the way a human would, if a human had superhero-like powers to sense terrain changes, weather factors and remain free from distraction.
Manufacturers can use AI data to monitor vehicle performance, identify faulty parts and gain actionable information for improvement. Connected consumers benefit from a range of perks such as local information on nearby stores, eateries and fuel stations, or additional safety features such as software vehicle updates, biometric capabilities or applications that can detect inattentive or tired driving.
Search -The world of search is a wide one that connects to each industry. Generally, the first step to consumer action in any of the industries covered above starts with a quest for information. Search providers are using Artificial Intelligence to monitor the infinite bytes of data collected all over the world to help refine the results users see. Perhaps you need to search for a doctor nearby or a lawyer who specializes in estates. Maybe you’re interested in a smart home security system or you want to know how soon you can read a book while your car drives you to work. The combination of AI technology such as Machine Learning, Deep Data Science and NLP works together to interpret a query and provide accurate answers obtained from recognized patterns pulled from the data abyss. Users breed data. AI learns from that data, which leads to more insightful, powerful search applications.
Preparation and agility are the antidote to obsolescence
Artificial Intelligence bridges the gap that once existed between time and efficiency. Tasks that seemed impossible to financially account for in man-hours can now be done in seconds for a fraction of the cost. No human could ever collect, categorize and analyze data at the pace or with the precision of AI. This opens a world of opportunity for anyone willing to embrace and invest in an AI future, leaving all others caught in the dust of a changing world.
In order to set themselves apart from other retailers, brands are using AI to generate highly personalized experiences, dependent on demographic data to help with decisions ranging from inventory assortment to campaign decisions and loyalty programs. This may demonstrate an optimistic attitude towards Artificial Intelligence, but the reality is that in order for companies to compete, they must manage to adjust for the AI disruption.
- Ricardo Baeza-Yates is the CTO of NTENT, a company that specializes in semantic search and natural language processing
- This article was originally published on Tech Talks. Read the original article here.