NEWS & VIEWS
Written by Nuno Godinho, CEO of Industrial Thought Group
As is happening with almost all industries right now, it’s fair to first enumerate the potential advantages that will come from technology like Chat GPT (Chat GPT being one of a multitude of Generative AI solutions available, including but not limited to Google’s BARD, LLaMA from Meta, etc), or any other Generative AI technology, for the wealth management industry. This is what happens with any revolutionary technology that is sure to upend the paradigm of the day.
It is important however to highlight that this is the day of the AI large language model, and whilst it’s a subsection of AI, it’s not AI in its entirety (our industry has been using AI and machine learning for a very long time in fact, but that is a chat for another day). This particular AI spinout is only as intelligent (for now) as the data with which it is trained on. We could take it that everything on the internet is good, honest, factually correct and brimming with purpose and reason… But that would be our downfall, and seemingly is a trap that others are occasionally falling into.
Tech scions and industry pioneers like Geoffrey Hinton, Elon Musk and Sam Altman, with financial stakes in the technology, are sounding warning calls, and industries attempting to adopt Generative AI into the fold encounter new problems with each implementation of it. That is why governance on an industry-by-industry basis is critical. In the wealth and asset management industry, trust is paramount, and the power of Generative AI needs to be reckoned with thoughtfully. The first issue is the data that is used to train the technology – where does it come from? Is it verified? Is it fraudulent? Until it is trained on the right data, among all players, all the data put in and all the data verified, the sorts of data that are currently held in the incumbent companies such as ours (FSL & Raw Knowledge) and our partners – it will forever be biased if it isn’t ALL the data. With bias comes a lack of trust, and we, the wealth management industry are custodians of peoples’ earnings, their livelihoods.
Trust and transparency must come first.
Crucially, governments must, and certainly will, impose some kind of regulations, yet we do not know when and how that will be. Industries and corporations can do their own due diligence by creating a marketplace to share their respective data so that the technology can be trained appropriately. Think of it as a central point that amasses the data from different sources and normalises it to be one, coherent, accessible and comparable library.
One marketplace can act similarly to how products such as Apple Health do for access to medical and wellness data – a central platform that pulls the data from numerous sources, devices and specialists, making it comparable and accessible for other Apps and healthcare providers whilst making it understandable and personal to the end user, it’s the hub API for millions of users and use cases. AI applications across wealth management will never be trustworthy or truly useful without the access, the open banking equivalent if you will, for the data – it always comes back to training the AI fairly and accurately.
The large language models are fascinating and useful however, as mentioned they can speed up front-of-office tasks a million times over, automating the fluff and allowing those that know how to develop and work alongside the technology for the good of the industry. And this is key, “those that know how”, as it stands today, Generative AI is liable to generate biased results unless you know exactly what to do in the prompt, despite the proliferation of specialists sharing the secrets of prompting it. One of the biggest learning curves for the majority is not what Generative AI is capable of, it is how to ask it the right thing.
Generative AI will undoubtedly accelerate wealth and asset management because it will enable much better personalisation, from understanding the customer better, to even being able to better communicate with the customer, it will strengthen and multiply connections and relationships. From a wealth manager’s perspective, the relationship is going to be much closer and much more personalised because you will now have the ability to search and discover what you previously could not, at speeds and depths not possible through human effort alone. AI will also speed up the creation of new advisory companies. Some of the things that you can do with Generative AI is create a new business model, build a business plan, and generate a new webpage, quicker, focusing more time on the value-added elements of wealth management – but it will take time before it can build portfolios and actually know market positions. Think about it, with AI as your sidekicks, a CEO, Product Manager/lead and a COO are capable, the three of them, in building out in a matter of days, functions and activities that prior to this tech would take an established team of 20+ months…
Done right, Generative AI is already an opportunity for positive change, but with great power comes… All I am trying to say is that the industry should test, learn, and collaborate and with guidance and a level head, it will better serve us all.