Artificial Intelligence – frequently shortened to AI – is transforming every sector of the modern economy. It’s behind everything from social media and eCommerce to the development of autonomous vehicles.

Across all industries, artificial intelligence provides the power to personalize content and engagements, boosting customer satisfaction and improving business performance. 

In the travel and retail sectors, Artificial Intelligence mainly provides the ability to respond to and engage with customers within a live environment, removing delays caused by manually crunching numbers, compiling reports and processing actions. 

The rate of progress is dizzying, and we have only reached the state of so-called Artificial Narrow Intelligence (ANI), designed to perform specific tasks without human intervention. For some in the field, there remains the goal of creating Artificial General Intelligence (AGI) that can apply itself to any situation, and Artificial Super Intelligence (ASI) that surpasses human capabilities. The next years and decades will see astonishing advances as science moves forward.

Having explained the relevance and bandwidth of AI, it is no wonder that almost everything today is ‘AI’ed’. But as with many buzzwords, not everything which has AI as a label is intelligent or makes sense. Here, we look at the big question – what really is Artificial Intelligence? – including:

1. A brief history of Artificial Intelligence

Although the concept of a thinking machine goes back to Greek philosophers, the phrase ‘Artificial Intelligence’ was first used in 1956 by computer scientist John McCarthy at the Dartmouth Conference. McCarthy defined AI as “the science and engineering of making intelligent machines”; the aim of AI was initially to create machines that “imitate the cognitive abilities of a human being.”

“Can machines think?”

– Alan Turing, 1950 paper ‘Computing Machinery and Intelligence’

Turing cracks the code

McCarthy coined the phrase, but English mathematician Alan Turing is considered the father of computer science after his seminal 1950 paper ‘Computing Machinery and Intelligence’. Turing asked “Can machines think?” and proposed the famous Turing Test to distinguish between computer and human responses. The film ‘The Imitation Game’ recounted how Turing’s calculating machine cracked the Nazi’s Enigma Code.

Game over for Kasparov

In 1967, US psychologist Frank Rosenblatt built the Mark 1 Perceptron, the first computer based on a neural network, learning through trial and error. By the 1980s, neural networks – which use algorithms to imitate how human brains learn – were common. Another milestone came in 1997 when IBM’s Deep Blue beat chess world champion Garry Kasparov, an unprecedented feat. Today, chess computers are vastly stronger than any human player.

DeepMind points the way forward

In 2016, there was a dramatic illustration of the emerging AI discipline of machine learning. British company DeepMind – now owned by Google – developed its AlphaGo computer, powered by a deep neural network. It defeated Lee Sodol, world champion at Go – a game with 14.5 trillion possibilities – after only four moves. A year later, an improved version, AlphaZero, taught itself to play chess in a few hours and then crushed Stockfish, previously the strongest chess computer.

AI takes the world by storm

In more recent years, AI has noticeably changed the way we live with innovative technologies. It is increasingly used for our broadly used virtual agents like Amazon’s Alexa or Apple’s Siri. It has made a huge step forward on self-driving cars like the ones from Tesla and even lifted various healthcare disciplines to a new level – for example through AI-driven medical imaging analysis. Today, in every industry, AI has a profound impact on every sector of society.


2. How AI works

Artificial Intelligence combines large datasets with algorithms to learn from patterns in data. It constantly evaluates its performance and develops greater expertise. In an era of big data, the power of Artificial Intelligence increases exponentially. With limitless energy, AI can run quickly through millions of tasks. And AI can carry out multivariate analysis, where it sorts through data from multiple dependent variables.

AI uses several techniques and technologies to think independently. To understand how it works, let’s examine the key concepts.

Data

A collection of raw facts and figures. In the travel sector, for example, it can be everything from behavioral visitor traffic data, booking information to product details. Today’s era of big data increases the power of Artificial Intelligence to learn quickly and enhances its predictive capacity.

Algorithms

A defined set of rules, procedures, calculations and steps to process data automatically. In travel sales, as an example, the algorithms respond to each customer activity on-site, assessing data and applying pre-set rules as determined in the coding.

Machine Learning (ML)

This specific application allows Artificial Intelligence to find patterns in data and, most importantly, improve its ability to solve tasks without explicit instructions. It can be a blend of external rules that have been set, alongside the freedom to continuously refine itself in-line with an overall output or objective. ‘Black box’ models are so complex they are hard for humans to interpret.

Self-learning

Is also known as unsupervised learning and forms a category of machine learning in which the algorithms are not provided with labeled data. Labeled data is “correct” information that has been processed by humans.

Deep Learning (DL)

Deep learning is a subset of machine learning that uses artificial neural networks that mimic biological ones. DeepMind’s owners say unlike other AI systems like IBM’s Deep Blue, their system is not pre-programmed, but learns from experience using deep reinforcement learning.

Neural Networks

Neural networks act like neurons in the human brain. These information messengers allow AI systems to find patterns in big data using algorithms. The networks learn and are trained by processing examples, assessing a likely association between points known as the input and the result, and require supervised learning to improve their output.

Natural Language Processing

Vital for any AI system that interacts with humans, NLP allows computers to understand how we communicate.

Computer Vision

A field of AI that trains computers to capture and interpret data from images and videos.

 

3. Where to apply Artificial Intelligence today

Where once it was the preserve of larger companies, Artificial Intelligence has become affordable and accessible for smaller organizations. Here are some concrete examples from key sectors using AI in business:

Travel

BD4 raised revenue for a leading airline by using machine learning to target customer incentives more efficiently. The requirement was to get away from a generic approach to offering vouchers that triggered unnecessary incentives as some visitors bought tickets anyway.

BD4 implemented its personalization platform for a customer-centric approach to achieve higher conversion rates. It utilized real-time customer-level signals to respond automatically to shopping behaviors. Personalized incentives are just one out of a multitude of interventions that can be provided automatically based on this platform.

Artificial Intelligence optimizes websites to attract direct bookings and review every aspect of the digital user experience. Other applications include AI concierges, hyperdynamic pricing, chatbots, robot check-ins and smart hiring.

EasyJet Holidays Case Study - How to provide relevance at scale with the help of AI-powered customer-centricity

eCommerce

Artificial Intelligence helps retailers optimize forecasts, pricing decisions and product placement, providing the most relevant shopping experiences. It provides personal recommendations on Netflix and Spotify. Amazon’s revolutionary Amazon Go stores use deep learning so customers can grab whatever they want without paying in store. Sensors and cameras track purchases and charge personal accounts.

Healthcare

In healthcare, AI can be used for analyzing scans, identifying cancers, forecasting kidney disease and more. It has been trained to provide personalized medicine, including giving reminders and suggestions for exercises. Recently, DeepMind used AI to reveal the structure of all proteins known to science. The insights are already being used by Oxford University to create a malaria vaccine and scientists predict a new era for medicine. Everyday wearable devices like Fitbits also rely on AI.

Manufacturing

AI analyses big data from sensors, machines and workers in factories. It creates solutions for managers running logistics operations. Another example is the development of AI-driven robots equipped with machine vision, including cobots working alongside employees.

 

4. How to get started with AI

Too many travel and retail companies shower customers with generic offers and content in the hope something will stick. But why provide discounts or incentives to people who won’t value them, or need them in order to buy? Everyone’s needs are unique and with 91% of consumers preferring personalized offers, speaking directly to individual users with tailored messaging must be the pillar your eCommerce strategy is built upon.

A study by Shopify shows online conversion can improve by around 8% when you include personalized consumer experiences on your website, with 74% of consumers growing frustrated by content that isn’t relevant to them. 

Precision Marketing

With BD4, you can optimize your website to take a Precision Marketing approach, communicating with every user individually in real-time.

BD4’s Machine Learning technology analyzes thousands of customer touchpoints and behaviors to form instant profiles. The AI predicts an individual’s willingness to book, and assesses which intervention is required to provide an ideal user experience. Unknown or anonymous users can be tracked and profiled from their first visit using self-learning algorithms developed specifically for travel purchases. Through integrating your product information into the BD4 platform, the algorithms work to match relevant products to each user.

More autonomy

AI can be implemented into existing Marketing and Sales platforms to provide automation to replace manual updates, freeing businesses to focus on strategic and tactical activities and plans. Hereby companies are able to deliver highly targeted marketing campaigns effortlessly and instantaneously without the need for manual web updates.

Technically, AI can be implemented in modular technology stacks. The concept of ‘Composable Commerce’ describes this approach, which enables an easy and efficient combination of industry-leading technology from multiple vendors.

To find out more about how BD4’s AI systems can improve your company’s performance get in tou.

Contact us