A year ago, the use of AI for many white-collar workers was speculative. Now, it’s an everyday reality in our lives. 

Consider this piece of evidence. We asked our guests at our first AI event six months ago to raise their hands if they had used Chat GPT – only a few hands went up. But things were different at our event last week. The people were mostly the same, but they had a different outlook on how integral AI is becoming to their overall marketing and communications strategy.   

We’re particularly interested in use cases for generative AI in financial services and fintech. We present here a few reflections, incorporating insight from industry experts who gathered in our offices recently. 

McKinsey research shows that the industries which rely most heavily on knowledge work will see the highest impact from Gen AI. GenAI should add 5% of global industry revenue to the banking sector – one of the largest gains of any part of the economy. (A similar study from Accenture this spring suggested 6% growth in just three years.) For an industry estimated to have a global worth of more than $10 trillion, this represents a significant shift. 

It’s all in the data 

With machine learning, it’s “garbage in, garbage out”. Chat GPT scraped and trained itself on information from some sources that were not always reliable. The result is that even though OpenAI has pumped billions into developing algorithms, results can be painfully stupid. Someone recently found that due to the way large language models parse information, they incorrectly will say that the word “strawberry” has two “r”s. 

There’s no room for error when managing money. That’s why our CEO Tom Coombes has brought together a multi-disciplinary group of experts to make sure that when we deploy AI, it makes sense. It means that Cognito’s been a bit slower to roll out market-facing products, but that stems from a commitment to appropriately using technology. Cognito expects to launch new tools soon that place an emphasis on time-tested methods to ensure accurate data and proper data sourcing. 

Keeping the human in the loop 

Humans have a critical role to play. Across financial firms, from customer representatives to marketing and communications, the expectation can never be that GenAI systems will simply come up with final, client-ready products. Humans are a vital safeguard against AI hallucinations and provide an extra layer of judgment. They also need to use their expertise to provide insight beyond what’s already available through online records to go where the models have not yet been trained. 

By formulating an AI strategy that prioritizes the concept of “human-in-the-loop”, businesses can establish powerful systems that allow both humans and technology to interact continuously and seamlessly. While GenAI possesses remarkable capabilities, businesses will thrive more when AI works in conjunction with human expertise.   

What does the industry need to consider and how can teams prepare for the AI revolution?  

Firstly, organizations must focus on building a strong foundation for AI implementation. Develop robust data infrastructure and governance frameworks to ensure the availability and quality of data required for AI models. Additionally, businesses must invest in talent acquisition and upskilling to cultivate a workforce that can effectively leverage AI technologies. This may mean looking beyond traditional recruiting pools and bringing in talent from people outside of the sector.  Collaborating with AI experts and partnering with technology vendors can provide valuable insights and support during scaling

Moreover, teams should prioritize addressing ethical and regulatory considerations associated with AI. As AI becomes more pervasive, ensuring transparency, fairness, and accountability in AI systems is crucial. Organizations must proactively assess potential biases, privacy concerns, and the impact of AI on individuals and society. Regulatory guidance is steadily arriving – although Asia is slightly behind the European Union right now. This will provide useful guard rails and best practices going forward. 

We’ve passed the awareness phase of the GenAI revolution. Now organizations, from innovative startups to global monoliths, must demonstrate results. To democratize access, AI must be accessible, affordable, and user-friendly. By considering these factors and adequately preparing, organizations can navigate the transition from AI adoption to scaling effectively. This will enable them to unlock the full potential of AI, drive operational efficiency, and stay competitive in the ever-evolving business landscape of 2024. 

Kinki To is an account executive in Hong Kong. This article also uses material from Zoe Forbes-Pyfrom.