It has been impossible to avoid the buzz surrounding Gen AI. Nearly a quarter of all articles from the global consultancies, IT industry and our own sector are extolling the potential of Gen AI, explaining how it has been created, how it can be used, and how it will change the way we all work. We are indeed now on the cusp of experiencing the biggest leap forward since the adoption of the internet.
But trying to work out what it means for your financing organisation is decidedly not straightforward. This article may help.
The latest crop of Gen AI models differs from previous AI applications in that they have been able to process gigantic amounts of data from the internet and use that to answer complex questions in different styles with differing levels of summarisation.
But, startingly, they have also started to address with surprising success some tricky logic problems, responding to hints to improve answers as if in a conversation. The new models can analyse data without direction, help developers by proposing code, and create genuinely new and unique content for music, images, text, and even poetry.
More prosaically, these new models can be used to collate, analyse and summarise new content and then use what is has learnt from all the other content to generate the output.
It is the sudden breadth and depth of usable capability, and the pace at which this is expanding and deepening, that has led to a consensus that projections of what AI will be able to do should be substantially brought forward.
This in turn means that organisations, if they haven’t already, must now consider the broader implications for their business. Which is fine, but what does that mean right now?
Firstly, if your organisation has already embarked on a digital transformation including ‘traditional AI’ then carry-on doing it. You are ahead of the game. You will be constantly considering how your business model will operate in a digital economy, will have moved or be moving to an agile model both in IT and across the business. You will have built (or be building) a hybrid multi-cloud architecture to efficiently and effectively support your significant data and application requirements.
Your platform will be able to safely access partner APIs, information and applications across different clouds and on-premises environments. You will have a focus on education to develop new skills and be majoring on data and cyber security. You will be bringing the entire organisation with you on your transformation journey.
For those earlier in your journey, perhaps starting to explore how to leverage the potential of AI, understanding your AI readiness is crucial.
To support businesses in this, Finativ offers an AI Readiness Assessment, which gives you the foundation for preparing your business for the next steps in AI integration, ensuring your strategies are both robust and forward-thinking.
There are many caveats surrounding the real-world application of Gen AI, including from academics who point out that the current crop of models and applications uses a narrow form of AI that does not really understand the data on which it appears to operate so powerfully.
Similarly, they argue that by design this method of Gen AI can lead to inaccurate results, and that it can suffer from invisible bias and may yield results which are difficult to explain, or indeed made-up.
Whilst all this may currently be true to a greater or lesser extent, there is so much investment being poured in, and the pace and inventiveness so intense that the consensus is that the current limitations will be successfully addressed, probably with hybrid AI approaches leveraging the recent advances.
Fears over wider control, model owner concentration and copyright may remain, but such is the momentum and the potential, that acceptable paths through those challenges will be found.
If you are already implementing AI, or just about to, then here are some considerations that come with Gen AI:
The traditional AI approaches may well be using machine learning or existing models that were created by but not necessarily using Gen AI. These are still valid and valuable. However, going forward, and in the not-too-distant future, it is likely that there will be a proliferation of Gen AI capabilities complemented by foundation models trained on financing. It is likely that these will be made very easy to access, in the way that many traditional AI models and applications have been packaged up and made easy to implement in the last few years. Therefore, it will be critical to stay in touch with the leading edge across the industry, or consider collaboration, to avoid inventing something which might be available off the shelf a short while later.
With new Gen AI models, there will be a greater reliance on having a hybrid platform to support the more onerous information and application requirements. This journey should if anything be accelerated.
Gen AI is presenting opportunities for disruption. For example, if a typical customer journey currently starts on a search engine, a user clicks through either directly to your website or to articles which have links embedded.
If that search now starts in a Gen AI application, consideration has to be given to what that means for SEO (Search Engine Optimisation) and it's likely that marketers will have to be prepared to optimise content for AI.
Once a customer engages with your website, leverage of the technology is maximised if the customer journey is supported by intelligent workflow and complementary messaging for guidance and support.
This matches the expectations of a generation raised on smart phones: a recent survey cited by Ofcom identified that 60% of mobile phone users in the UK prefer messaging apps such as WhatsApp rather than texts or voice calls. So indeed, the days of the traditional call centre may well be numbered.
Gen AI has already started to appear in day-to-day applications such as Microsoft Office, and this trend will continue, enabling rapid creation of data analysis, reports, marketing content and presentations. This will free up time, but likely lead to reductions in headcount in certain areas.
Gen AI will likely accelerate automation by simply making it easier to implement AI and, therefore, lead to fewer administrative staff being needed. However, operational subject matter experts will become even more valuable as guardians and enablers of the AI running the business.
So, it is important to identify those key individuals in the business now and start migrating them to be ‘AI ready’ if they are not already. Going forward, it will be even more important for the business and IT to work as one.
These are just a few pointers to start the conversation in your business. With the latest Gen AI applications, the expectations, demand and investment in AI has reached a tipping point.
Whilst the pace will only increase going forward, the good news is that it will become easier to implement AI. However, this could be both an opportunity and a threat: if your business has not started to modernise then it may be left behind. The potential, though, is exciting - soon Gen AI will just become ‘AI’, it will be everywhere, and a well-trodden path for all of us.
