Welcome to part 1 of 2 on our blog series about AI, written by our friends at Yes Consulting. These blogs will cover the following:

Part 1: What is AI, with a semi-technical explanation with examples of use cases and tools used across this industry and others.

Part 2: In the Hospitality sector, the more widespread use cases of AI tools this sector with examples, including content, imagery and guest communication.

Read Part 2 here.

WHAT IS AI?

In the context of this article, it made sense to use ChatGpt 4 (openai.com) to self-define its own space rather than take a human approach!

Artificial Intelligence (AI) is a broad area of computer science referring to the simulation of human intelligence in machines. Essentially, it’s about creating algorithms that allow computers to perform tasks that typically require human intelligence. These tasks include problem-solving, understanding natural language, recognising patterns, and making decisions. AI encompasses various subfields, including machine learning (ML), neural networks, robotics, natural language processing (NLP), and more, such as LLMs (Large Language Models like GPT-4)

Back to a human narrative: In many cases, AI requires a human interface and the use of language to make it worthwhile. To make this even more confusing there are both LLMs and CAIs (Conversational AI). 

LLMs: are a subset of AI models designed to understand, generate, and manipulate human language. They can be used for text generation, translation, summarisation, and question-answering tasks. LLMs are a type of neural network, typically using architectures like transformers (e.g., GPT or BART). Their focus is specifically on language-related tasks.

Conversational AI: refers to technologies that allow machines to understand and generate human-like conversations. This can be in the form of chatbots, voice assistants, and other interactive platforms. While conversational AI often leverages LLMs (especially for state-of-the-art systems), it also involves other components, such as intent recognition, dialogue management, and user context understanding. Conversational AI can be built using rule-based systems, machine learning-based systems, or a combination of both. 

LLMs are concerned primarily with understanding and generating language. In contrast, Conversational AI aims to facilitate a two-way interaction, mimicking a human-like conversation, but all fall under the AI umbrella for the average individual.

Bots are a commonly used word and a piece of software that uses artificial intelligence and machine learning to mimic human interaction. In doing so, it can respond to natural questioning. 

A few typical AI applications

• Virtual Assistants: Like Siri, Alexa, and Google Assistant.
• Recommendation Systems: Used by platforms like Netflix or Amazon to suggest content or products based on user behaviour and in the selection of accommodation!
• Autonomous Vehicles: Cars or drones that navigate without human input.
• Medical Diagnosis: AI systems can analyse images, genetic information, or other data to detect diseases.
• Finance: AI is used for fraud detection, robo-advisors, and algorithmic trading.
• Travel: The next wave of travel bots is expected to predict booking requirements, build traveller profiles, assist with disruption avoidance and increase the uptake of loyalty programs. Plus, all the usual chatbot assistance on bookings, cancellations etc. 

According to Booking.com, over 45% of consumers prefer messaging platforms over email when contacting businesses, but how useful is this technology in the travel industry?  In July this year, Booking.com launched a new travel tool, the “AI Trip Planner”. This quote explains their direction.

The AI Trip Planner takes the trip planning process one step further by providing travellers with a visual list of destinations and properties, including Booking.com’s pricing information, with deep links to view more details. Travellers can go back and forth between their chat with the AI Trip Planner and the Booking.com app interface as they consider options for their trip. With just a tap on any accommodation they are interested in, they can complete the reservation, as the AI Trip Planner is directly integrated into the booking experience in the Booking.com app.

Airbnb’s CEO Brian Chesky was recently quoted as saying:

“Airbnb’s plans to incorporate AI into the app and make it the centre of how the app runs, with some launches coming later this year but major changes coming by mid next year, focusing on personalised recommendations and improved customer service.”

Airbnb already uses AI to improve searches, manage prices, classify in-app messages and more. As Airbnb and Booking push toward mobile and app-based business models, diluting Google search, AI becomes even more essential to improve the traveller’s experience. 

—————————————————————————-

All AI relies on data; the bigger, the better, hence the term “Big Data”. This refers to enormous datasets that traditional data processing systems can’t handle effectively. These datasets are characterised by their massive volume, high velocity, variety, and complexity. In this context, ‘velocity’ means the speed at which new data is generated and collected. 

We can witness data streams at an unprecedented rate from sources like social media, physical sensors, and online transactions. This big data is vital in AI, as it provides the extensive examples needed for machine learning models to refine algorithms, recognise patterns, and make accurate predictions.

The narrative above explains and gives examples of large corporations’ capacity to use large amounts of data to accelerate their operations and booking opportunities. These large companies dominate many accommodation booking market sectors, increasingly targeting guest influence.

What about smaller and localised businesses, do they have an opportunity to leverage AI in their operations, marketing or business activities? Smaller companies rely on third-party tools to compete with guest management, marketing and business practices.  A few of these are listed below that are commonly used but by no means comprehensive: