TTE2018: Getting GDPR ready is first step towards AI adoption, says ICE

Getting control of data is ‘will prepare firms’ for roll out of fast-emerging tech Continue reading →

Becoming compliant with new EU data rules is a first step towards firms adopting Artificial Intelligence, visitors to Travel Technology Europe were told.

Ian Richardson, co-founder and principal consultant at consultancy TheICEWay, said looming GDPR rules due to come in on May 25 require firms to get control of their data.

Presenting on “How to spice up your business with AI”, he said getting control of data is also an enabler for AI, for which data is the most important ingredient, and will prepare firms for the roll out of this fast-emerging technology.

“AI needs to learn and it needs data to learn,” he said, “it needs a combination of both good and bad data.

“Data is personal and it’s regulated, and it’s also not in one place. How do you get control of all of that data?”

Richardson said becoming GDPR compliant requires firms to know where their data is and what they hold and so was “an enabler towards getting ready for AI.

“Because you have to know where your data is and control it. That’s the law. As part of that [GDPR] you are getting control of your data, you are taking tour first steps towards AI,” he said.

Richardson set out a 90 day plan for firms to adopt an AI solution into their businesses.

First firms must start with a “data knowledge exercise” to understand where their data is and then work out what challenges they need to solve. “The amount of times our consultants go into companies and they have not done the thinking about what type of problem they are trying to solve. They are just wasting a load of money,” said Richardson.

“Once you’ve worked out what the problems are, ask can an AI product by adapted to solve them.”

At this point firms should research AI vendors, re-imagine their business processes and consider how the new technology will be integrated into their business.

Following this firms should do due diligence on their AI vendor and develop proofs of concept. Richardson advised companies to budget for not every proof of concept to work out as hoped.

He set out the four key questions AI vendors should be asked in the due diligence phase:

• What does AI mean to you and does this product fulfil that definition?
• How is your product superior to a current option that does not use AI?
• How will the performance of your product improve with AI?
• How much time and staff will I need to dedicate to your product?

Among the key applications for AI in travel Richardson cited review and sentiment analysis, chatbots and virtual assistants, making sense of images and emotion and facial recognition.

He said chatbots was probably the area travel firms could most easily adopt today because the technology is relatively mature. But he warned: “A bad bot is bad for business. If you do not get it right it can really affect your brand. But consumers do expect to be conversing now with machines.”

Richardson said the low cost of computing today means AI is enjoying a “summer” in terms of hype and the promise companies see in the technology and as a result there is plenty of funding available for AI projects including from new forms of investment like crowdfunding. He said this is prompting entire industries to see AI as an enabler for complete digital transformation but that it was “shocking” that there is hardly anyone in travel involved in this “rethinking” of the sector.