Five industries that are nailing AI

The idea of computers programming themselves was nothing but science fiction less than a decade ago. We believed that computers were remarkable tools but that they would always require programming by humans. Now, computers learn from the data they analyse, fine-tuning themselves over time. Artificial intelligence (AI) is a reality.

The share of jobs requiring AI skills has increased 4.5 times since 2013. More professionals means quicker progress and ‘smarter’ algorithms. For instance, since 2010, errors for image recognition tasks have reduced from 28.5% to below 2.5%. AI is already just as good as us at recognising phone speech audio and by 2035, it is predicted to be better at driving – it’s thought that about ½ million lives could be saved by AI-driven cars.

Here are five industries that are already seeing major changes thanks to AI technology.

AI in Healthcare

The most common use of AI in medicine today is improved diagnostic procedure. Smart algorithms can be trained to operationalise large amounts of historic and current data, notice similarities and draw conclusions about the best treatment options.

Take a specific type of cancer as an example. With enough patient data (gender, genetics, cancer staging etc) and information about treatment protocols, AI can determine the most appropriate combination of chemotherapy drugs for a patient. By analysing other patient data (lifestyle, geographic environment, genetic makeup etc), it is hoped that AI will be able to perform predictive analysis and determine a person’s level of risk for developing cancer. This will allow healthcare providers to take preemptive measures with their patients.

A new Google-created system of AI software is now in use to predict and diagnose a patient’s potential to have a major cardiovascular event (heart attack or stroke). And it is based solely upon photographing the eye retina. According to a recent study, this approach works.

The study took retina photos and health data from 284,335 patients. Based on the photographs, risk factors for cardiovascular disease were identified, including age, smoking behaviour and blood pressure. Follow up research looked at 150 patients who had had major cardiovascular events within five years of their eye scan. The algorithm was then given two retina images and asked to predict which patient has had a health issue. The algorithm made the correct predictions 70% of the time.

Using AI to predict cardio events from retina scans
From images like this, the researchers were able to identify what exactly the AI was paying attention to while assessing risk. Source: Business Insider

In general, this study demonstrates the benefits of deep learning in transforming how medical researchers and practitioners can study the body and provide a more complete picture of a person’s current and future health.

Another result? As all of this data is made available to people, they will take a more active role in their own healthcare.

AI in Finance

The internet has changed how consumers look for loans, banking and investment opportunities. Financial institutions are well aware of shifting consumer behaviours and the rising demand for personalisation and seamless, on-the-go experience. In response, businesses have figured out how to use AI to better meet individual consumer demands.

What AI currently does for financial institutions is to collect data on past and current consumers across the industry. That data, when operationalised with the help of data science, can do several things:

  • It can demonstrate the types of loans/investment products that specific segments of consumers find most valuable.
  • It can predict risk factors of those who apply for loans.
  • It can provide data that drives decisions on which loan or investment products to develop for segments of consumers.
  • It can provide critical information on pricing so that banks can remain competitive in the marketplace.
  • Based on market segmentation, offers can be personalised and pushed out to target audiences.
  • It can predict potential churn so that measures can be taken to retain current customers through incentivisation. Let’s take an example of a mortgage loan. Data collected on current customers can be used to predict which types of mortgage loans are most attractive to them. A bank can then study what competitors are offering and develop new loan products that will be just as attractive to specific demographics among their customers. Thus, customers are retained.

The use of such insights can be further automated with the help of AI algorithms that would automatically interact with users to suggest the best products for them; pitching personalised financial services or providing unique financial advice.

Robo-advisors are already able to create personalised investors’ portfolios based on automated risk-profiling tools and to adjust the allocation of assets according to current market trends and shifts. Christian Spitz, Senior Manager at Accenture, states that we are gradually moving to the future of holistic and customised financial advice that is fully automated. “Enhancements in cognitive computing, big data and behavioural analytics will augment AI further and will allow machines to sense, comprehend, act and learn. This will quickly pave the way for technology to gather and understand complex client needs; propose and implement solutions based on historic client behavior, and explain complex concepts in the context of each particular client,” notes Spitz.

AI in influencer marketing

It is predicted that by 2019, influencer marketing will be a $2 billion industry. On Instagram alone, marketers are spending over $1 billion. Here is what AI can do for marketers who intend to level up their strategies in 2018.

AI can help identify the right influencers to partner with based on specific queries you provide – for example, their primary follower’s location or specific age brackets. The tools will notify you whether any parameters change over time e.g. if the follower count grows or drops. Some tools can also help you analyse the past performance of selected influencers by audience segmentation and partnerships with similar brands.

Smart social media platforms can now help predict the best times for launching a specific influencer campaign and auto-suggest the campaign duration, based on historical data. Marketers can look at influencers’ past campaign success data, and what specific products or services generated the most interest with the target audience of a specific influencer. The relaunched Linqia platform, for instance, uses machine learning to analyse an influencer’s past content, match it with the brand’s affinity and take note of patterns that can determine the success of partnering with them based on the brand’s goal. The test runs proved that campaigns optimised by AI over the course of time drove 51% more engagement.

Influencer marketing is already providing a powerful tool for marketers, as consumers become increasingly wary of brands blatantly promoting themselves. Using AI will ensure that brands partner up with the right influencers for their goals and plan their campaigns for optimal reach. And using AI will allow marketers to find the right influencer the first time, rather than trying a ‘hit and miss’ strategy – the same way AI is now helping marketers to develop better PPC campaigns.

AI in travel

Travel brands are mining customer data and using AI for predictive analysis regarding consumer behaviour and potential travel intent. AI is also widely in use for keeping travellers informed about journey updates and generally improving their experiences. Here are the key uses of AI in the travel industry.

Advanced Product Personalisation is based on previous individual behaviour and on the behaviour of similar consumer demographics. This includes hotels, cruises and cruise lines, package deals, hotels, car rentals, flights and fares. AI can also provide suggestions for better prices, based on customer travel details. Airlines are now actively exploring AI-driven dynamic pricing solutions that would allow carriers to create custom prices for each ticket. These would depend on multiple factors, including the customer’s previous history with the brand, loyalty status, add-on services purchased and so on.

Product personalisation also means that push marketing can be more effective. If a consumer usually books a specific type of vacation (e.g. a skiing holiday), special deals for lookalike trips will be delivered to them.

Chatbots and virtual assistants are now allowing travellers to be in charge of their own travel plans – searching for the best fights/flight prices, suggestions for places to visit, even providing information about traffic. Hipmunk is perhaps the best example of this – a virtual travel assistant that has incorporated messaging platforms from Facebook, Skype and Slack, and offers comprehensive advice to travellers. In responding to queries, it will access massive amounts of data regarding pricing, room inventories and itineraries. It can even be set up to provide alerts on special deals and pricing, and it remembers earlier conversations in order to provide the most personalised assistance.

hellohipmunk screenshot

Robotics has also gotten into the act. Hilton’s Connie is an intelligent robot that provides personal concierge services to guests in their hotels. While still in its early stages, operating as an answerer to specific questions, Hilton envisions a future in which Connie will recognise return customers and even offer translation services worldwide.

Hilton Worldwide's robot concierge
Hilton Worldwide is experimenting with a robot concierge named Connie (Source: USA Today)

AI in education

While change comes slowly to this sector, AI does have the potential to radically change education.

Now, using AI systems, with additional support, students can be anywhere in the world and still ‘attend’ supervised learning programs. The current and potential use of AI in education includes the following.

Progressive teachers now use AI systems for grading purposes. They have been in use for objective assignments and tests for some time. But teachers have not been able to input their rules for grading written work until fairly recently. With new technologies in place, machines can now learn how to implement custom factors in the grading of student written work and deliver a better assessment of individual performance.

AI has already shown great progress in personalising educational experiences for students. Based upon accumulated data and individual student profiles of learning, AI can make recommendations for the best remediation activities when a student has not mastered a skill or concept. AI can also set up individual learning programs, freeing up teachers to spend more time individually with students. For example, the online learning system Coursera, as well as Carnegie Learning, are already leveraging past student data to create custom programs/assignments for learners. Also, data is gathered about individual student performance and mastery in each course, as well as such things as dropout and retention rates. This informs decisions that instructors make to modify and improve their courses.


Conversational UIs (chatbots) are already assisting students in their queries about educational program options at various schools, as well as their potential for being admitted, based upon their profiles and gathered data of past admissions. Priansh Shah, an undergrad computer science major at Imperial College, London, actually developed an algorithm to see if he could predict which colleges would accept and reject him. He achieved an accuracy rate of between 75-80% and made his tool available on GitHub.

Are these the only industries being disrupted by AI? Not by a long shot. Virtually every sector of the economy and government will ultimately be impacted by the ability of computers to continue to learn.

Chris Pitt profile picture
Chris Pitt

Chris is Managing Director at Vertical Leap and has over 25 years' experience in sales and marketing. He is a keynote speaker and frequent blogger, with a particular interest in intelligent automation and data analytics. In his spare time, he enjoys playing the guitar and is a stage manager at the Victorious Festival.

More articles by Chris
Related articles
AI in marketing: we need more analysis, not analytics

AI in marketing: we need more analysis, not analytics

By Chris Pitt
AI in travel: 3 ways to build customer loyalty

AI in travel: 3 ways to build customer loyalty

By Chris Pitt
Face made up of code

4 practical uses of artificial intelligence in marketing

By George Karapalidis