Ai
Everybody wants to say ‘we’re running generative AI’
logistics of operations and experiment digitally before trying things. “Another use case is in IT to
auto-generate code. How many travel companies are sitting on creaking technology developed over decades, with the tech not well-documented? With generative AI, you can regenerate code in modern [coding] language and migrate off a legacy platform to run a more efficient system. “Chief information officers are excited
because there is a big productivity angle for them. You can deploy AI to help with testing, reducing some of the risk and security concerns, and get it to generate documentation and training. “These types of use cases can give
organisations a route into experimenting with the technology and identifying some of the risks and shortcomings without going straight to consumers.”
FIRST STEPS Asked how she would recommend businesses select a use case, Ronca- Thompson advised: “Go back to the problem you’re trying to solve. If you use Gen AI, or AI more broadly, what can you do with it that you couldn’t do before? There is going to be a learning journey with this. Start small with an internal use case, something that isn’t market-facing and that won’t cause you reputational damage should it go wrong. Start with
The more we can develop the right use cases, the right skills, the better we’ll be able to navigate and get around some of the inevitable problems
a team willing to experiment and see how it behaves. Learn what ‘prompt engineering’ is about, develop skills in- house and partner with an organisation that knows how to do these things. “There is a temptation to go after
things without a clear rationale. So, the rationale is important. Once it is clear, move through the ethics and risks. What data are you going to use? Is the data of decent quality? How do you know there isn’t any bias inherent in the data set? How are you going to monitor it for ‘model drift’ – so, in three months or a year it’s still doing what you expected? How do you set up the right governance mechanisms? “Then there are the technology
costs. These things come at the price of processing vast amounts of data. Get clarity on the costs early and decide whether you want to push into production. How far across the enterprise or your markets do you want to push this functionality? Costs can run away from you if you’re not clear about the potential uses, particularly if you’re going to push it out to consumers. It’s hard to predict how consumers are going to behave. “Do this with a multifunctional
A MAJORITY of UK businesses are not using AI, but almost half of large businesses
have explored use cases, with business operations
the most common area of application (Figure 39)
FIGURE 39: BUSINESS USE OF AI
UK business use of AI: % by type of use case
Data processing using machine learning
Image processing using machine learning
Text generation using LLMs*
Visual content creation Robotics 0 1 1.9% 2 3 4 5 6 7 0 All 0-9 staff 10-49 50-99 100-249 250+ 4.3% 3.1% 5.8% 6.4% 100
20 40 60 80
85% 86% 80% 73% 68% 54%
Businesses not using AI at all yet: by size of workforce
10 15 20 25 30 35
0 5
Proposed business use cases for AI: all UK businesses
33.5%
team. If the use case sits in marketing, have a multifunctional team of people who speak different languages with diversity of thought. Have your finance business partner to hand and your
20% 14%
New products/ services
Business operations
Personalised products/services Source: ONS, December 2023 30 Travel Weekly Insight Report 2024
Page 1 |
Page 2 |
Page 3 |
Page 4 |
Page 5 |
Page 6 |
Page 7 |
Page 8 |
Page 9 |
Page 10 |
Page 11 |
Page 12 |
Page 13 |
Page 14 |
Page 15 |
Page 16 |
Page 17 |
Page 18 |
Page 19 |
Page 20 |
Page 21 |
Page 22 |
Page 23 |
Page 24 |
Page 25 |
Page 26 |
Page 27 |
Page 28 |
Page 29 |
Page 30 |
Page 31 |
Page 32 |
Page 33 |
Page 34 |
Page 35 |
Page 36 |
Page 37 |
Page 38 |
Page 39 |
Page 40 |
Page 41 |
Page 42 |
Page 43 |
Page 44 |
Page 45 |
Page 46 |
Page 47 |
Page 48 |
Page 49 |
Page 50 |
Page 51 |
Page 52 |
Page 53 |
Page 54 |
Page 55 |
Page 56 |
Page 57 |
Page 58 |
Page 59 |
Page 60
