Artificial Intelligence (AI) is great at generating valuable insights, but what companies end up doing with those insights – that’s where true achievement lies, and that’s what companies should be focusing on.
This is one of the core messages delivered by Ryan Waite, Director of Open “From Open
Waite described the interaction between AI and human expertise as a learning loop, a kind of hill-climbing machine that helps businesses improve, and keep on improving.
The knowledge belongs to the business
AI models are great at processing enormous amounts of information and generating insights at a scale humans cannot match (and sometimes cannot even fathom), but people bring something arguably more important to the table: context and judgement.
Putting insights into wider context, and using that to make conclusions, helps employees turn information into knowledge, and that knowledge later becomes part of the organization’s collective expertise. And with every iteration the knowledge grows, helping businesses climb further up the proverbial mountain.
Sovereignty and ownership seem to be an underlying theme to most of the keynotes here at the Openrates insight, and humans turn that insight to knowledge, it now becomes vital to secure that knowledge and make sure it remains under the organization’s control
“It’s based on your data,” he said. “It’s based on your people. It shouldn’t be given off to an AI model, that’s separate. And that’s why we believe that for organizations to build their own hill-climbing machines, to build their own learning groups, they need to build control with the sort of learning loops and not be beholden to a single AI provider.”
After all, everything AI generates is based (or at least, should be) on proprietary business information, and the decisions made based on those insights depend on employees and their understanding of the company. So, businesses need to make sure ownership of the knowledge accumulated through these interactions are not handed over to an external AI provider – a challenge that will probably grow bigger and bigger as businesses become more dependent on AI tools.
The knowledge generated through these tools could become tied to a particular model, or provider. What happens when businesses want to adopt better technologies as they emerge? To future-proof these processes, Waite argues businesses should build their own learning loops and maintain control over the information, processes, and technologies that make them work.
Choosing the right AI model
One way to do that is to make sure you’re not tied to a single AI model. Many businesses, Waite says, begin their AI journey with proprietary frontier models, especially when tackling problems they don’t fully understand themselves. However, as they become better and understand their tasks more, they might not need expensive frontier models any more – they can move to specialized models that could be smaller, or cheaper. They could even move to open-weight models they can operate themselves.
“Having the ability to move between different models is important, especially as you think about your algorithms. Which models are you going to use? You shouldn’t be trapped with a particular kind of model because it has your knowledge,” he concluded.
So, what’s important here is to retain the freedom to switch and move along the way. That also means businesses need to control where their AI workloads run, whether that’s within sovereign infrastructure governed by local regulations, on-prem systems, or public cloud environments (where they can take advantage of greater scalability and cost advantages).
