Application of AI across industries: A short overview of who’s doing what?

I have been meaning to track it for a while. And now here’s a brief overview. Which industries are experimenting with AI and where have they reached. What I have tried to do is find out the following:

• Which areas of the industry are getting impacted by AI the most.
• Which problems are getting solved in the process
• Who are the major players in this area and what does the playerlist stack upto
• Breakthroughs that they have already achieved and future progress.

But, before going into that, let’s start with the definition. Artificial Intelligence is the intelligence exhibited by machines. So when a machine starts to mimic cognitive functions that humans associate with other human minds, we start to refer it as Artificial Intelligence. The other aspect that I wanted to highlight is that what falls within the definition of AI is flexible to a large extent. For example, Optical character recognition is no longer being treated as AI, becoming a part of routine technology now.

The field of AI research itself is divided into several sub-fields that focuses on specific problems or specific approaches or on the use of specific tools to satisfy specific requirements. We will not be going in-depth into them in this article. This is more of skimming the surface. So here we roll.

AI IN HEALTHCARE

Healthcare is where it gets really big. From virtual nurses to drug discovery, from diagnostics to personalized care, AI is creating major breakthroughs in the field of healthcare. Here are the areas where they become most effective.

Diagnostics: As per CB Insights about 22 companies are developing new programs for imaging and diagnostics. As deep learning algorithms become more and more adept at recognizing patterns, this has become a lucrative field of study. Cancer diagnostics is a major field of study by some of the key startups here including the IBM backed Pathway Genomics. Pathway Genomics has come up with its blood test kit for cancer called CancerInterceptDetect for people who have never been detected with cancer to see if early diagnosis is possible. Butterfly Network, is reinventing the ultrasound machine by squeezing all its components onto a single silicon chip with the intention of making the technology available across masses and making affordable diagnostics easy.The stealth startup Imagen Technologies is working on creating a world without diagnostic errors.

Drug Discovery: Reducing time for drug discovery is another area which has received considerable attention from major AI and machine learning players of late. It takes upto six years to build a body of evidence to support a new drug candidate for commercial development. twoXAR which has created the DUMA drug discovery platform intends to reduce it to minutes. DUMA evaluateS large public and proprietary datasets to identify and rank high probability drug-disease matches orders of magnitude faster than wet-lab approaches. These matches can then be used to cross validate clinical research and identify novel drug candidature. NuMedii is creating pipeline drug candidates with a higher probability of therapeutic success. The Company uses its Life Sciences Big Data technology to translate novel disease-drug connections into new indications for existing drugs.

Mining Medical Records: This is the most obvious area where AI has initiated sea changes. From collecting to storing, from normalizing to tracing its lineage — data management is undergoing a major change under AI. Recently, Google DeepMind launched the Google DeepMind Health project which intends to give faster, better treatment to patients.

Designing Treatment Plans: Identifying potential treatment plans for cancer patients is what IBM WATSON ONCOLOGY is focusing on. Watson for Oncology analyzes the meaning and context of structured and unstructured data in clinical notes and reports and then by combining attributes from the patient’s file with clinical expertise, external research and data, the program identifies potential treatment plans for a patient. Memorial Sloan Kettering Cancer Center and Manipal Hospital are two healthcare institutions that have already established WATSON FOR ONCOLOGY to create potential treatment paths.

Virtual Nursing Assistant: Molly, developed by the startup sense.ly is the world’s first virtual nurse. With a smiling, amicable face coupled with a pleasant voice Molly uses machine learning to support patients with chronic conditions in-between doctor’s visits. It provides proven, customized monitoring and follow-up care, with a strong focus on chronic diseases.

AI IN FINANCIAL SERVICES

Anything that deals with massive amounts of data and repetitive processes can do with digitization and automation. Therfore, financial services has not only been a major but also one of the oldest playing grounds for artificial intelligence. From customer service to fund management, from robo advisors to security systems, AI seems to pervading the financial services industry.

Here is a quick glimpse at some of the major players and the areas that they concentrate on.

Personal Finance: Before we reach the bank, we can always refer to wallet.AI. Founded in 2012 in San Francisco, wallet.AI builds AI that pieces together millions of of pieces of data to help make informed decisions. As we move around and make thousands of decisions about what to eat, where to go, which cab to take we leave breadcrumbs of data behind us which, when put together, creates a pattern of our spending. wallet.AI works like a watchdog, figuring out the spending patterns that are leading to money wastages and informs the user about making better financial decisions.

Wealth Management and stock trading: The Wolf of Wall Street might just live in the clouds. The Wall Street trading pits are getting transformed with a number of transactions being performed by computers now instead of fund managers in flashing suits. Sentient Technologies for example has been doing algorthmic trading with it’s AI platform for more than a year now. Emma AI is another example. Emma, a startup intends to revolutionize ‘quant’ data science altogether. Emma AI differs from current finance computing because its system of neural nets takes into account a more complex set of factors affecting stocks, like management changes or monetary policy in Europe, that other programs miss. IBM’s Watson is also partnering with Citibank in offering wealth management solutions.

Customer Service: At the Commercial Bank of Dubai, a virtual customer assistant named Sara is available 24/7 to assist visitors in filling out forms and getting up to date answers on saving and investing. Moreover, you don’t need to write your questions. You can just talk to Sara, the virtual assistant. From emotion recognition to chatbots, the use of artificial intelligence in banking has become a major field of study in itself. The impact of implementing AI in banking has been a major reduction in resource costs as well as efficiency. The virtual assistant is available 24 X 7 and has answers to the most asked questions. Aldebaran Robotics has gone a step beyond and has implemented Nao ( a humanoid robotic companion) which implements a similar scenario by talking to the customer in person.

Emotionscan is another technology that is helping banks understand better. The software analyzes the facial expressions of the customers as they listen to a series of scenarios designed around cash flow.

Banks now use a combination of technologies to bring down costs and increase reach and artificial intelligence is only helping them pave the way. DBS Bank Singapore launched DBS Digibank in India where the actual bank was present on the customer’s smartphone as an app, the account opening could be done through a physical kiosk present in coffee shops and malls, and the entire customer service is managed by an artificial intelligence created by Kasisto.

AI IN EDUCATION

The education industry had already been at the forefront of technology through adoption of smartclasses, tablets, and new ways of creating content beyond textbooks. The implications that AI is displaying in the education industry also indicate profound changes that we are about to come across in the way we learn, process and execute information.

Intelligent Tutors:Intelligent or Cognitive Tutors are is a common example of how AI is pervading K-12 segments.I Cognitive tutors are able to track the mental steps of the learner during problem-solving tasks to diagnose misconceptions and estimate the learner’s understanding of the domain. Furthermore, Intelligent Tutor Systems can also prescribe learning activities at the level of difficulty and with the content most appropriate for the learner. Tabtor is an example which learns about the student just as the students learn about the subject. At the other end is a human tutor though, but the human tutor is fed a ton of data that the machine has learned about the student. Through Tabtor’s “Digital Paper” technology and Point of Learning Analytics POLATM, Tutors track every pen stroke and the timing for each equation. They can see how a Student answered a question, including their scribbles and what they erased.

Educational Companions: In 2015, a group of Romanian programmers created Woogie a voice enabled AI device. It is forthcoming for kids aged between 6 and 12 years, native English speakers. It will be able to detect, read, process and understand the human language along with the ability to converttext to speech and vice versa, play radio stations, podcasts and shows according to the user’s age. Developers hope that Woogie will help children to memorize different information from multiple areas. It does that based on interactivity. It acknowledges the kid’s presence in the room and reacts according to this.

Personal Training: Internet of Things has anyway pervaded our fitness universe with devices like FitBit and others. It will soon be the time for the personal trainer to come in the form of an AI who can give personalized advice on fitness routines while ensuring that we avoid over exerting or injuring ourselves.

Custom Textbooks and Lectures: Content Technologies, Inc. a startup created by Dr. Scott. R. Perfitt is experimenting with content related technologies with an AI backbone. Content Technologies is addressing custom content through a series of solutions. Palitt, an AI solution reimagines how content is assembled. Until now, custom content and custom textbooks were an expensive and time consuming process. Using Palitt, people can create their custom lecture series, syllabus or textbook in a jiffy. Another product, Cram101’s AI technology can turn any textbook into a smart study guide complete with chapter summaries, unlimited true-false and multiple choice practice tests and flashcards all drilled down to a specific textbook, ISBN number, author and chapter. And finally, NursingEd101 utilizes AI technology to help nursing students take the stress out of learning. I The NursingEd101 system breaks down the content into smaller increments in order to help student feel less overwhelmed and retain more information.

AI still has a substantial way to go when it comes to education compared to the way it has pervaded healthcare or finance or manufacturing, but with predictive analysis picking up in open online courses, it might just be a matter of time before AI becomes a natural part of our education telling us what to learn and how to learn it very soon.

AI IN MANUFACTURING

AI’s can be fully paid for employees and a Japanese Company Deep Knowledge has recently appointed an AI as its Director due to its ability to predict market knowledge that is not immediately apparent to humans.

Manufacturing seems to be the industry that is ripe for AI intervention. The processes are clear and robots can be programmed to perform them. Foxconn, the №1 EMS manufacturer have been planning to deploy 10,000 robots to offset increasing cost of labour in their Chinese development center.

From automotive cars to self running flights, AI has pervaded industries which deal with manufacturing in a number of ways. But let’s just look at how it has impacted the manufacturing process over here.

The Intelligent Factory: When it comes to the factory floor, it was initially just the hard working robots who were taking over. But not anymore. AI has come a long way from big scary looking machines working monotonously. Instead, the factories themselves are becoming smart thereby creating Industry 4.0. A smart factory is a networked factory, in which data from supply chains, design teams, production lines and quality control are linked to form a highly integrated, intelligent creation engine.

The ideal intelligent factory, or as GE calls it, Brilliant Factory, looks something like this. Software, such as GE’s Production Execution Supervisor, captures order data from customers. If, say, an aircraft maker is running low on fuel nozzles, the software can automatically order production on new units. Using 3D printing technologies, they can design, prototype and test a new part in hours instead of days or weeks. The part then goes into production on a line that’s mostly ‘staffed’ by intelligent robots. Each stage is monitored by sensors that feed data to AI and analytics software, such as GE’s OEE Performance Analyzer, that reside in the cloud. If a defect is spotted, or a new part needs to enter production, the software orders the part and the process begins anew. GE’s brilliant factory is already in use by Procter & Gamble which is experiencing sufficient savings in downtime. In fact, the market for smart manufacturing tools is poised to hit $250 Billion in 2018.

AI IN FILMS

Imagine yourself as a producer. An independent one who is about to produce his first film. Would you be interested to know the chances you have in ensuring it’s box office success? Of course, you will. It’s your money and you would like to see returns on that. Even though the film and television industry has been promoting AI in its content for years (be it through Spielberg or Star Trek), the actual application of AI on the industry itself has been very limited. But over the last few years, we have witnessed some major changes. AI has slowly been creeping in almost some of the major tasks — such as scripting, trailers and even movie making.

In 2016, IBM Watson came up with the first AI made trailer for the movie Morgan. Morgan is a psychological thriller. To create the trailer, researchers fed WATSON with 100 horror movie trailers. WATSON processed them to understand the nuances of a trailer and then delivered the movie trailer in 24 hours by selecting 10 scenes totaling six minutes of video.

While IBM WATSON gets into serious moviemaking, the Israeli startup Vaulthas created a program that claims to be able to tell whether a movie will be a hit or a flop. David Stiff, Vault’s CEO and co-founder, says Vault focuses on an intensive analysis of 300,000 to 400,000 story “features,” which can be things like themes or level of violence. All these story features are pulled from the script by his program with no human input. The Vault AI has been trained with script data going back to the 80’s. Currently Stiff says that the AI can predict a hit or a flop with almost 65 to 70% accuracy. That’s a lot considering that only 15–20% movies really make money.

Postscript:

AI is not just revolutionizing industries. It is also changing roles and functions at a random. This was a brief list, and the implications of AI in areas like air traffic control for aviation, connecting automobiles for transportation, smart infrastructure in cities are also far reaching.

But this could be just the beginning…. and we still have time to speculate if governments should move for a minimum basic income policy or can AI exist with humans in the coming future.

(Reposted from Linkedin Pulse: https://www.linkedin.com/pulse/applications-ai-different-industries-list-whos-doing-what-ghosh?trk=pulse_spock-articles)