Sick of poor customer service? 'Luna' could be the solution

“This is Luna,” says Kiryl Chykeyuk, motioning me towards a woman dressed in a red suit at the back of the room. 

Luna shifts about awkwardly, occasionally placing her hand on her hip, then holding it limp by her side, before cycling through the motions again, in a way that is sort of but not quite human. 

Luna can be forgiven, of course, because she is not a human, but a hologram.

What I am looking at is an avatar generated entirely by artificial intelligence, projected in 3D by hundreds of rapidly spinning LEDs. 

Chykeyuk explains to me that the hologram machine rotates the lights faster than the human eye can detect, although even a glass case cannot mask the sound of the fan-like apparatus that brings Luna to life whirring in the background.

It’s an image that is hard to process, but it is perhaps something we may all have to get used to. 

Chykeyuk hopes his holograms could soon have pride of place in banks, shops and even schools.

It comes as Britain is in the throes of a customer service crisis – with thousands of bank branches disappearing from high street and companies and government departments forcing callers to wait on the phone or deal with computer-generated chatbots. 

But critics warn that banks and service providers should think twice before replacing human beings on the customer service frontline with holograms. 

Originally from Belarus, Chykeyuk, 37, has been developing this technology for the best part of 11 years. 

After completing a PhD in Oxford, the entrepreneur won investment on BBC’s Dragon’s Den, which he later rejected due to the terms of the offer. 

Investment later arrived after a successful pitch to Richard Branson, enabling Chykeyuk to create HYPERVSN, the company of which he is now CEO.

I am now standing in a showroom full of the company’s efforts: a life-sized parrot shimmers on the wall; the logo from Barbie rotates over my head; and in the corner, a sports car idly spins on a platform. 

Adjusting a tablet mounted next to the hologram causes the car’s interior to change colour, tapping more buttons swaps the body paint. 

It’s not unlike a video game, but Chykeyuk explains the applications are more high-brow than Need For Speed.

“It’s great for car manufacturers,” he says. “Very often in dealerships, you don’t have the cars to hand, so customers can see what’s available. It’s also more entertaining than a screen.”

Versus a simple screen, the appeal of holograms is their added third dimension – the rapidly spinning lights play on the human eye’s depth perception, creating the illusion of a 3D image. 

The technology has rapidly eclipsed Augmented Reality (AR), which requires expensive and complicated headsets to give the impression of a 3D image – and can only be worn by one person at a time.

That being said, holograms themselves are not an especially new technology – and Chykeyuk tells me they have already made their way into British high streets. 

The Sports Direct on Oxford Street, for example, already uses them as part of its décor. 

Scores of other businesses and public bodies – including Network Rail – are enquiring about installing holograms to display messaging, which are cheaper to run than LED screens.

What is new is Luna herself – who marries hologram technology with AI, specifically ChatGPT. 

Speaking aloud to Luna prompts her to respond as the generative software would. 

Chykeyuk asks her to tell him about London, and she obediently rattles off iconic landmarks, her mouth clumsily framing each word, her face displaying no emotion.

As AIs go, ChatGPT is not as intelligent as bespoke systems companies have developed for their own purposes, and HYPERVSN does not develop its own software. 

But companies that purchase a hologram can simply plug in their own, more intelligent chatbots, and let Luna do the rest. 

Banks, Chykeyuk, explains, are the obvious market.

“Luna can be the face of any generative AI,” he says. “Banks have their own chatbots and that entire database can be connected to the hologram. It doesn’t even have to be Luna – it can be a man, a woman, even a cartoon character.”

As anyone who has tried to contact their bank in the last two years will tell you, AI chatbots are already widely used by banks. But they have not exactly gone down well with customers. 

A 2021 Which? study found chatbots and social media were among the worst ways to complain to a bank. 

Retailers, too, have come under fire for overreliance on the technology. 

Which?’s study found 65pc of people thought face-to-face complaints were the most effective method. But what happens when even that involves interacting with AI?

Derek French, campaigner and former bank executive, says customers were already struggling with service provided by computers. 

He says: “Most people who come to their bank with a problem start the conversation with a robot, but they generally seem incapable of dealing with issues you come up with. 

“If you’re at the point where you’re trying to contact your bank in the first place it’s because you’re dealing with an issue you can’t solve yourself online.

“My experience of these chatbots has been very frustrating with the level of intelligence these chatbots currently have. I’m very for progress but sometimes we go too fast and we don’t get it right.”

Chykeyuk maintains there should always be someone on hand in case Luna cannot handle a question. 

Holographic AI assistants are set to be installed in Neom, a 150km line of urban development planned in Saudi Arabia, but a “back office” of humans will be ready to step in if needed.

But the army of Lunas will not stop at customer service: holographic AI assistants are being lined up for schools, too. 

Case in point, an AI teaching bot is already being implemented into Harvard’s flagship coding course.

Naturally, the expansion of AI will have a knock-on effect on the job market. In some ways this is already happening. 

This month, The Telegraph reported that workers in the technology sector were already seeing fewer vacancies than last year, and lower salaries for previously lucrative roles. 

Teachers, too, are nervous about the encroaching threat of AI. 

Two in five teachers admitted they were worried their job would be at risk if AI continues to advance at its current pace, according to a survey by learning platform Quizlet. 

Over two-thirds called for schools to establish a code of conduct or an advisory body for the use of generative AI or similar technology.

In the absence of legislation, this may be the only defence schools have. The law simply cannot keep up with the evolving technology – and unions are clamouring for action. 

Less than half of those surveyed by tech union Prospect said they were confident their employer would consult them about introducing the technology at work. 

Andrew Pakes, Prospect’s deputy general secretary, said: “Advances in technology have the potential to bring huge benefits to both employers and workers. 

But there needs to be proper consultation with workers about what is being introduced and how before anything is implemented.”

Previously it seemed job roles with a human-facing element were safe from the AI revolution. 

Consultants, therapists and life coaches were all thought to be in the clear, and continued to pay high salaries, according to research by job website Adzuna.

Typically, AI’s defenders argue that jobs lost to the technology will be replaced by jobs to do with manipulating it. But asking retail customer service staff to retrain as AI software engineers seems a little naïve. 

I ask Chykeyuk if he is concerned that low-skilled workers will lose jobs to this technology, and the tech whizzes behind it (Luna would cost a company £35,000, a drop in the ocean for most businesses). 

“This is the future, with or without us,” Chykeyuk says. “Low qualification jobs to a certain degree are going to be replaced. But this is how the world works. It’s human development, basically.”

Looking at Luna from the corner of my eye while talking to Chykeyuk is giving me a headache. I doubt she is ready to step into a classroom in the state she is in. But soon she will be. 

“If you look at the state of AI five years ago, we didn’t have anything,” Chykeyuk says. 

“Within five years we have got something that is extraordinary and is only going to improve. We are expecting that in three to five years’ time – less than that, even – you won’t be able to see the difference between Luna and a real human.”

The hologram has a different function I’m keen to test while I’m here. 

Sitting on a stool in front of a green screen at the other end of the room, my entire body is projected in convincing 3D in the glass case where Luna stood just minutes before. 

Angling my head slightly, it almost looks as if I am in conversation with Chykeyuk. 

It is, and I cannot stress this enough, really weird.

But done well, such technology can be pulled off without anyone caring they are looking at what are essentially hyperfast windmills of tiny lights. 

ABBA’s Voyage residency in London, performed entirely by holograms of the band, shifted one million tickets – enough for Universal Music Group to send the show on a global tour.

Chykeyuk says holograms are the next logical step for post-Zoom work events. Rather than project conference speakers via video link on grainy cameras in home offices, they will be beamed in as holograms. 

Daft though it may sound, it is already happening. 

Ukraine President Zelensky appeared at The Next Web conference in Amsterdam via hologram just last year. Keynote speakers routinely speak via hologram – and can be booked either live or virtually for future events.

“It gives you this feeling of presence more than Zoom,” Chykeyuk says. “You can stream yourself from London to New York, appearing almost as if you were there.”

In the meantime, for the banks and companies wanting to save on staff costs, Luna might soon become the AI-generated face of Britain’s customer service decline.

The Telegraph’s technology editor, James Titcomb, has answered the most pressing of readers’ queries about AI and what it holds for the future, here