Scaled Machine Learning, not Generative AI (GenAI) being adopted by big business

Following Cumulo9 GM David Allen’s attendance at the recent AI & innovations conference in Auckland, the overriding observation was that a focus on scaled machine learning, rather than GenAI, is being adopted by big business. To follow is a summary of some of his takeaways. 

Attending  

When you have conference sponsors like AWS, Dell, Google and Spark you know there is something hot being discussed, and this conference did not disappoint. The line-up of speakers was impressive. Given that AI and process automation is all about creating competitive advantage, I was surprised that so much detail was presented by the likes of AA Insurance, BNZ, One NZ, Westpac and Spark. 

Chat GPT was always going to be a hot topic of conversation because the release of the technology (particularly R4) has amplified and stimulated the whole AI conversation within businesses. It’s no longer a fringe activity, and “C” level executives across all roles are now seeing the relevance of the technology, opening the way for meaningful budgets. There was a view that 2023 might be seen as a watershed moment in time, like the arrival of the PC, Microsoft Windows desktop, the internet or the mobile phone.  

What were the key takeaways

Westpac’s Matt Haigh spoke about the bank adopting “scaled machine learning” rather than generative AI. He presented a detailed description of the work the bank is doing, which has involved addressing the very real cybercrime issues being faced by all in financial services. This area was chosen to be their first cab off the rank. Matt talked about the complexities of taking use cases from models into full production. Complete and clean data was also seen as a critical element that must be tackled first. But Matt acknowledged that this area is always a hard place to start. 

It was no surprise that for Westpac, working through the organisation’s standards, policies, and rules created the greatest complexity, a key theme throughout the day. New roles are emerging within organisations, and Matt spoke of the key role of the Machine Learning Engineer. 

The bank sees machine learning, supplemented by digital humans (my words, not Matt’s), taking over the mundane, less complex front-line tasks.  New Zealand-based Soul Machines was mentioned as a world-leading player in this rapidly evolving area. 

Ethical Panels 

Some businesses are creating Ethical Panels so that there is formal governance over the use of scaled machine learning and AI, as there are ethical issues to be dealt with. The collection of data and deeper use of it is one area where boundaries need to be monitored and rules and laws upheld. However, this new technology is generally heavily reliant on personal data, which does create a dilemma within the organisations adopting the technology.  

Where to next with these technologies? 

The role of the digital agent is seen as a background one, where their activity will be “just in time”, and that they will assume the less complex tasks where efficiency of response overrides the need for personal touch. The reality is that this is already happening in some businesses; it will just become the norm moving forward. 

The format of the day was mainly panel discussions, where organisations like The AI Forum, Auckland Transport, BNZ, Dell, Fonterra, Gallagher, Spark, F&P Healthcare and Xero all had a say. 

This technology is moving fast, with the use cases adopted being highly relevant, with immediate returns. AI is now being widely used to supplement skills so that more complex tasks can be performed without bringing in a subject matter expert. There is a view that we are just a click away from having a “universal language” where AI will be used to generate real-time speech translation services. A big upside for international meetings as well as general global connectivity. 

There was an interesting conversation on “skills for the future” and a view that roles within business are quickly evolving and that within three years, many workers will be in a different job or that the current role description will be substantially different.  

Being AI ready 

The “AI-ready economy” concept was discussed, and businesses not adopting the new technology would get left behind. Culture was seen as a big issue, and successful businesses must create a “culture of diversity of thought.” There was general agreement that workers needed to spend more time back in the office as the “water cooler conversations” were an important part of effective, collaborative working and innovation. Most of the businesses presenting now mandate a certain number of days in the office. Generally, at least three. 

In Summary 

Whether it is scaled machine learning or pure AI being adopted, Chat GPT's arrival has certainly created one of those moments in time. “That was so 2023” could well be a conversation reference to the way we ‘used to’ do things. The team at Cumulo9 are embracing the changing landscape and working together to determine the role AI could play in our Solutions and will be closely watching developments in this fast-moving space. Contact Cumulo9 if you have any questions or suggestions.