Why AI is more than a passing fad

Whatever your views on Artificial Intelligence (AI), one thing is certain- it’s not going anywhere. AI technology is already deeply embedded in our homes and workplaces, and it will only penetrate further as time goes on. 


This presents a challenge for businesses. While most are keen to embrace new possibilities, many are unsure of the best way to go about it. Changing too much too fast could alienate customers and employees or put data and processes at risk, but taking a cautious approach could mean missing the boat altogether. 


Finding the right balance is vital, and nobody understands this better than our resident AI gurus Greg Duckworth and Damien Duff. It may be a young technology, but these two already have years of experience in helping businesses incorporate Machine Learning (ML) and AI. We caught up with Greg and Damien to talk about the current state of AI, and to find out what the future has in store. 


Hi Greg and Damien, thanks for joining us. Could you start by telling us what you see as the pros and cons of AI for business in general?




It's probably a good idea to start with the cons as these have been getting the most attention lately. The potential harms of AI are weighing heavy on us all to the point that they have become a pressing political issue. Apart from the existential and social worries that we all have, in the business world, the biggest risk being felt is the risk of being left behind. Many businesses will need to constantly adapt to stay ahead of the curve, but there is also a risk of biting off too much newness and creating new problems due to incorporating new technology without the oversight.


In more concrete terms, AI can create problems with both privacy and accuracy. The recent Samsung data leak into ChatGPT showed how careful we have to be about the information we allow AI to access. There have also been many cases of AI simply getting things wrong in a way that alienates customers. A classic example is a feedback loop where customers are recommended products that have only got famous as a result of having been presented by the same algorithm in previous recommendations. These kinds of glitches are still common, and can be detrimental to business outcomes like customer experience. 


That said, these particular issues must be considered alongside the massive potential benefits. AI is creating opportunities that would have been impossible a few years ago. For example, in the world of customer-focused business, businesses can personalise customer service to a much greater extent, improving relationships and winning customer loyalty in the process. I’m also excited about the possibilities for the world at large. Big problems like climate change are going to require big tools, and AI is playing an important role. 




I have to agree with Damien’s optimism. The true possibilities of AI are still unknown, but the technology is already capable of amazing things. Businesses can automate processes that could not be automated before, generating huge gains in efficiency. Today’s algorithms are powerful, robust and accurate, giving businesses access to sophisticated modelling and new opportunities for growth. It’s natural to be afraid of these changes, but I believe that this fear is misplaced. Of course, there are still very real things to worry about, such as those that Damien has already pointed out. But I think a good solution is approaching this technology carefully, and being transparent about its performance.  


Besides these general pros and cons, are there any industry-specific benefits or drawbacks?




Retail is definitely one of the biggest growth areas for AI. Soon we’ll be able to build systems that understand customers as a human might, by incorporating an understanding of their expressed desires along with their behaviour.  This will certainly help with sales, but it will also improve customer experience. AI will be able to interact with customers in an empathetic way, answering their questions quickly and with perfect accuracy. 


Of course, there will be teething troubles here too. Retail scammers can harness the same kinds of up and coming AI to become more convincing, using personalised information to trick customers into revealing sensitive data. Retail is a customer-facing industry built on trust, so anything that undermines this trust could be extremely harmful. To avoid this scenario, automated systems for protecting customer data must be thoroughly vetted and hardened.  


AI will become incredibly efficient and accurate at its designated tasks, but I don't think that AI systems will or should be fully autonomous and operate without human intervention. Rather, I think AI systems will be used to fulfil tasks like automation, personalization and so on, and humans will act in executive and regulatory roles in which they take on the responsibility for automated processes, oversee them and intervene if necessary. For this, design of the interfaces with AI that account for the non-deterministic nature of AI and the ergonomic and cognitive aspects of users is key.




Finance is another industry in which AI will be transformative. Personal/private banking is really an extension of retail in the sense that banks offer products to individuals, so many of the benefits will be the same. Financial service providers will be able to build more accurate customer profiles, allowing them to offer a more personalised range of products. On top of this, repetitive data-driven tasks can be automated, freeing up human workers for more high-level work. Finally, as algorithms become more powerful they will allow more accurate modelling. This will be a huge help in areas such as fraud detection and portfolio management.


The monetary stakes are higher in finance, so the risks are also greater. AI is only as good as the data it’s trained on, so there’s a real danger of predictions being skewed by bad information, or biased datasets. Transparency is also a problem: AI models are often incredibly complex, and it is difficult to understand how they make the predictions they do. This could make it difficult to regulate markets and industries, and may cause some hesitation by some sectors in fully adopting AI. 


How is the general workforce responding to the presence of AI in the workplace?




It’s a mixed bag really. Certain industries are facing huge job losses, and this is justifiably creating a lot of anxiety. Tasks like translation can be almost entirely automated, leaving very little room for human workers. As AI improves, many other jobs are likely to be replaced. 


The flip side of this is that AI has been embraced as a major efficiency tool by a vast number of workers in some industries. Many individuals have incorporated it into their working practices and have found that it makes them vastly more productive. Some of the time, the current AI transformation seems to be moving in the opposite direction to some digital transformation and previous AI fads. While these are often imposed from above on sceptical workers, some aspects of the current AI revolution are being driven from the bottom up. 




ChatGPT has played a huge role in bringing AI to the masses. People are using it in their personal lives, and then bringing this experience into the workplace. They realise that AI can improve their productivity at work too, and this has driven its adoption in many workplaces. The effects vary from industry to industry, but AI can add a lot of value to low level work. Once workers realise how much AI can help them, they tend to be less fearful.




And what about businesses? Do they share their workers’ enthusiasm?




Many of them do, yes. So to pick an example in the cutting edge, we’re beginning to see a lot of visual personalisation. AI can look at shop windows or living rooms and suggest similar or complementary products. Or it can follow a customer’s Instagram or Twitter feed and make suggestions based on their activity. More broadly, personalisation is a key sales driver, and retailers have been quick to recognise the potential for growth as well as improved customer experience. 


In another example applicable to non-retail industries, businesses are getting in on the act by using AI to organise their collaborative platforms. So I envision, for example, information stored in Google Docs, Confluence and Microsoft Sharepoint being checked regularly for contradictions and redundancies and remediated in an automated fashion. AI can do this in a fraction of the time. It can be as effective as hiring a hundred interns! 




Aside from the usual fears about AI taking over the world, my impression is that most businesses are pretty positive about the technology, and are, at the very least, thinking about how AI can help them become more efficient.. The main hurdle at the moment is its perceived complexity. This dissuades some companies from implementing even basic AI applications, even though they could add real value to the organisation. Cost is also a major blocker, and it seems this is mostly due to the real costs of running AI systems to be simply unknown at present to the company that wants to implement AI. It’s then difficult for the business to gauge whether or not adopting AI is a good investment or not. Implementing AI often requires hiring an external company, and many businesses are unwilling to spend this money. But I think in future, costs and difficulty are inevitably going to go down, allowing more and more companies to fully embrace AI. 


Are some sectors embracing AI more than others? 




I think a lot of it comes down to risk. Young start-ups and corporations with deep pockets are both able to take bigger risks than those in sectors with more governance and regulation. These companies want to be seen as forward-thinking, so they’ve been more willing to include AI in their R&D initiatives. 


Retail and marketing have incorporated AI at a much faster rate than other industries. This is because they can act on AI recommendations with relatively little risk. It’s not the end of the world if Netflix recommends the wrong film, or H&M sends a personalised advert for the wrong T-shirt. These kinds of prediction errors would be far more serious in a healthcare or insurance setting, so these sectors have been a bit more cautious around their AI implementation. 


Retail certainly seems to be leading the way in AI adoption. How do you see this playing out over the next few years?




Personalisation will continue to improve. Retailers will use AI to anticipate their customers’ needs, ensuring they get what they need as quickly as possible. Large language models will blur the line between chat and shopping, allowing AI to respond to shoppers' questions and make product recommendations in a way that feels appropriate and natural.


Predictive analytics will also play a vital role. Retailers have access to a myriad of customer data, and they will use this to make ever more sophisticated predictions about the behaviour of individual customers and the market as a whole. This will allow them to spot dangers and opportunities, and to act on these before it's too late. The data owned by retailers is itself a treasure trove that can be monetised in a variety of ways, not just in their retail businesses.


There are also some exciting practical applications making their way into stores. Computer vision will be a big help in preventing shoplifting. AI can monitor security footage and discreetly warn staff of any suspicious behaviour.


Looking further ahead, the number of advanced and experimental AI options is going to increase. Retailers need to be aware of these opportunities, and to start thinking about prototypes and proofs of concept now. This technology is constantly evolving, so getting in on the ground floor will make a huge difference. 


AI Retail Personalisation


How big a strategy decision is it for a business to embrace AI? How can Daemon help with this process?




I suppose it depends on what you mean by AI. If you use the broadest definition, then lots of businesses have already embraced it. Excel, for example, has long used a basic form of AI to calculate data. Things get a bit more complicated when it comes to Generative AI, but even here there are many instances where this technology is already commonplace. Predictive text  is a rudimentary form of Generative AI that we’ve all been using for years. 




For many types of AI, it’s not a big decision at all. The technology is becoming increasingly accessible- not just for experts but for everyone. Costs are coming down and open source AI and AI platforms are making the technology easier to access. 


 A big part of our work is helping to upskill our clients’  in-house engineering teams, and AI tools can play an important role here. A good example of this is our Ignite GenAI proposition. This is a specially designed roadmap that helps businesses take their first steps into Generative AI. We work with clients to decide on the best use cases, and then we build out Generative AI prototypes in a matter of weeks. This lets businesses get a feel for the technology without putting a lot of time or money on the line. This can be combined with our Ignite Machine Learning (ML) offering, which delivers machine learning prototypes in a similar timeframe. 


Find out more about IgniteML


So, it’s safe to say that AI is not a passing fad?




It's here to stay. In the long term we need to not only think about how to deal with the worst consequences of AI - we need to collectively get behind positive social movements to manage the changes that are coming. For businesses, we need to be on our toes in understanding this new technology as it comes through, and before it comes through, to stay ahead of the market and keep our competitive edge, whilst also doing our utmost to understand its ramifications and be at the forefront of good practice.




AI has been around for a while now in various flavours- predictive text, spell checkers, machine learning etc. Generative AI has the potential to transform our working lives, but there’s really no way of knowing at this stage. Whatever happens, AI is bound to play a huge role in the workplace of the future.  It’s just a question of how. 


Whatever your business, it’s never too early to prepare for the future. If you have any questions about AI, machine learning or anything else relating to digital transformation, don’t hesitate to get in touch.

Back to Blog