Power to the people
John O’Malia, founder and CEO of VAIX.AI
, is captivated by the potential for artificial intelligence (AI) to not only transform online gaming and betting, but also usher in a new era of human empowerment.
“I couldn’t not do it,” says O’Malia of his decision to launch VAIX – a business on a mission to use AI and AI agents to transform ecommerce, human capability and business opportunity. A serial entrepreneur experienced in starting and building online spreadbetting services, live betting interfaces and online gaming companies, O’Malia had in fact had no intention of founding another start-up – not until he became captivated by the potential offered by AI, triggered in part by the birth of his son and a new interest in organic brain development versus AI.
O’Malia therefore launched VAIX in early 2016 and now leads a team of nine people operating from the UK and Athens.
“We provide consultancy and the operational deployment of algorithms that improve business effectiveness,” he says.
The business is currently working with clients in the online casino and sports betting spaces, some of whom are completely new entrants.
These “AI first” disruptors “will create a real shake-up in the landscape”, O’Malia says. Through AI, VAIX can help such businesses analyse customers’ behaviours and give them a much more tailored service.
"Gaming is a great lab for developing certain types of products, especially relating to CRM and customer loyalty,” O’Malia says. “Because it’s so data rich, you can very quickly validate theses about what’s going to be effective. We are going to take those insights about communicating with players and making deeper connections with players and push those algorithms and insights out to other industries.”
The approach has applications for online marketing more widely, potentially in any sector where there are enough data points to enable rich analysis.
“I started this company wanting to do something easy so that I can do something hard,” O’Malia says. “The something easy is deploying existing AI technologies. The something hard is trying to create something transformative in AI that really moves the needle in terms of human empowerment.”
Personalised offers through AI
In terms of the “something easy”, this translates into helping betting companies offer truly personal betting opportunities or odds to individuals. For online gaming and casino companies, it could mean identifying which of their many games are most likely to appeal to a particular player.
“You make sure that, from the first game they choose, you start to learn a bit about how they play and what else they might like,” O’Malia says. “The idea is not to offer them 1,200 games, but to offer them eight or 10 games that they are definitely going to love. That’s the magic we are looking to bring to operators. It’s empowering to the player and it will generate loyalty, brand connection and turnover.” In online bingo, AI can also ensure that chat rooms, which are an important contributor to players’ enjoyment, are kept interesting. “You can make the chat and community vastly better by using AI to pull interesting topics that are bingo-friendly out of Reddit [a social news website and forum] or elsewhere and push them across to the chat hosts, who can then feed them over to players,” O’Malia says. “The hosts get a constantly changing feed of content that they can drop them into the chat room to spice things up.”
Applying AI in this way signals the end of long-established marketing theories, O’Malia believes.
“In my university marketing classes we talked about ‘bucketing’, ‘cohorts’ and ‘clustering people’,” he says. “With AI you can bin that entire conceptual framework. Why jam people into a bucket if the algorithms are so capable of dealing with high dimensional data that they can learn to understand the subtle differences between players and so really customise messages?”
O’Malia sees a “pretty exciting roadmap” in gaming over the next three years or so.
“Things will be responsive and personal and connected to you, rather than generic and ‘spammy’ with the same messages for everyone,” he says. “To become more personal is a great opportunity for brands, and I think that’s certainly where gaming is going.” Perhaps further ahead, video gaming could evolve further. AI could become able to generate “novel, completely new visual elements, increasingly realistically”, O’Malia says: “Generative Adversarial Networks could lead to some interesting new visual representations of randomness and chance. Maybe that leads to some interesting game dynamics.”
Time to experiment
The emergence of AI cannot be ignored, O’Malia feels.
“Because you can do it, you must do it,” he says. “Otherwise your competitors will do it. These are the advantages that craft competitive shifts in the landscape. The people who become personal and relevant are going to outperform those who don’t.” He therefore encourages businesses to “start playing around” with AI now. “Spend a bit of time and resource on it, because it’s probably going to be transformative for the way you do business and you have got to figure out how. If you don’t start exploring it now, you are going to get disrupted by the people who are.” One of the encouraging things about AI is that so much is open source. “In many cases the moment an algorithm is finished and tested, it’s pushed out into open source,” O’Malia says. “There’s lots to be explored.”
O’Malia believes VAIX is not only helping its clients to disrupt their sectors by applying AI, but also disrupting the whole CRM sector by focusing so intensely on deep AI personalisation.
“Our approach to this move away from bucketing to entirely personalised algorithms is that for each player we look for the optimal configuration – that’s certainly disruptive,” O’Malia says.
Returning to the “something hard” that O’Malia wants to achieve through VAIX, he says:
“Right now AI is mostly high dimensional computational power and pattern recognition. Our view is, why not see what people can do if you put this at their disposal? That’s the technology that we call Wingman, that we are starting to develop. We’ve just patented an architecture. Wingman is all about giving people the ability to direct AI algorithms, to do what they want to do. If we’re talking about predictions for the future, I think that will be the big story.”
For example, rather than simply replacing drug researchers with algorithms that crunch through data, the Wingman approach would combine the power of the algorithms with the insights of the human researchers.
“I don’t want the next drug breakthrough to depend upon us creating artificial general intelligence,” O’Malia says. “I would rather say to the drug researchers, here’s an amazing new computational tool that should make you vastly more effective – if we customise it to your domain and give you the power to direct it using natural, intuitive communication.”
Wingman plays to the strengths of people.
“Humans have this extraordinary capacity, which is the ability to synthesise and associate information in a very intuitive and non-structured way,” O’Malia explains. “Even if there isn’t a direct correlation between things, we can draw an analogy or we can use one thing as a metaphor for another and use that to understand what’s happening. AI doesn’t have that capability.” This isn’t a problem if all that’s required of AI is the ability to draw conclusions in data rich environments, but often there is a need to apply other insights, judgment or experience. Stockmarket traders, for example, could potentially have real value in the Brexit era when so much is uncertain. “If a person has been trading in a market for 20 years, they have a fair bit of insight,” O’Malia says. “What happens in a Brexit scenario where there is no playbook algorithmically that you can really look to? You can monitor the price action, but if you have to formulate a theory about what’s going on and your market’s correlation with other markets, that’s still certainly the domain of human operators. And we empower those operators with a powerful new AI toolset.”
For this reason, the Wingman approach is less “scary” than pure AI, O’Malia suggests. It may replace data crunching activity, but it wouldn’t replace people completely.
“It’s about creating a world where we have incredible tools and are able to do things we could never do before,” he says. “That era of empowerment through AI – that’s what gets me up in the morning. It makes AI less threatening and more aligned with humanity. People are starting to realise that’s important.”
This view drives O’Malia’s business goals for VAIX.
“We would like to transform gaming and use that momentum to drive into online marketing and CRM more generally,” he says. “For Wingman, the goal is nothing less than to shift the approach of AI research away from considering humans in the loop to be a disadvantage, and really looking at humans in the loop as the future. Half of AI research should be focused on human empowerment – on human-directed AI. If we achieve that, then the world will be a very different place in 10 years’ time.”
There will still be challenges from increased automation and AI, of course, even where it enables higher levels of human performance. For example, if more junior type of activity become automated, how do you cultivate the next generation of experts?
“Societally, that’s a big challenge,” O’Malia notes.
About VAIX AI
Industry: Information Technology and Services
Company size: 2-10 employees
Ownership: Privately held
Although an American by birth, O’Malia has worked in Europe since the start of his career in the early 1990s. He sees the UK as a particularly strong AI centre, with the likes of Deep Mind (an AI leader acquired by Google) and Apple (which is investing in AI and building its new headquarters in Battersea). He considers the UK to be ahead of Germany and France, encouraged partly by the UK government’s support in the form of R&D tax credits.
“There’s also been a lot of activity by the UK government to seek guidance from the [AI] industry,” O’Malia says. “I would encourage the UK government to continue to really focus on this, Brexit or no Brexit. There is such a vibrant community in the UK for AI. The government should continue to talk it up and encourage investment in it.”
AI itself is evolving.
“The stacking of networks has become hugely important,” O’Malia says. “We are moving from a single algorithm that tries to master a situation, to stacks of algorithms that communicate and optimise things.” For example, the stacked algorithms could be part of a chatbot and each be contributing different suggestions for the conversation. “You could have a hub at the middle that selects the most appropriate one,” O’Malia says. “That could give you a higher quality result because you are not trying to generate one perfect sentence; you are trying to generate a bunch of different sentences and then match the most appropriate ones for your conversational partner. That could lead to some more dynamic interaction.”
“We challenge ourselves to read all the relevant research that comes out,” O’Malia says. “You might get 40 new papers that come out on AI in a day – before a conference there’s a blizzard. To ensure that one is aware of what’s going on and where it’s moving, one has to digest and process a fair amount of information about new research. That’s what sets oneself apart – the willingness to continually re-evaluate one’s approach in the light of new algorithms because the pace of change is shocking.”
One of the challenges of working in the AI space is that progress is being made so fast, so future-proofing a business like VAIX requires staying right up to date.
We asked: Away from VAIX, what disruptive tech has most impressed you?
O’Malia vividly remembers his uncle, who designed computer chips in Silicon Valley, turning up with a “big briefcase computer” with a mouse and a small, fold-down, green screen.
“That moment changed everything for me,” O’Malia says. “You could create and erase and do stuff. It was magic. When I see AI now, it’s like seeing that computer again. It changes everything about the way one lives one’s life."