FutureWeek sat down for a conversation with Peter van der Putten, Director of the AI Lab and Lead Scientist at automation software firm Pega to delve into what left and right brain AI is, the influence this is likely to have on marketing, and how agentic AI works under the hood.
What do you do in your role?
I head up our AI Lab, which means I’m responsible for AI innovation within our client-base and internally. I’m always thinking about how I can utilise AI so our clients can create more meaningful and relevant interactions with their customers. I’m also thinking about how they can optimise their business to become more effective.
Can you describe what Pega does?
We help our clients to have more meaningful interactions with their customers – whether in marketing, business process management, or the automation of workflows. We do that with a platform that combines both AI and automation.
It’s about being efficient through automation but also having the intelligence to do the right things. We often talk about the platform being both intelligence and muscle – the intelligence knows the right thing to do but the muscle performs the action.
That’s a great analogy of AI. What is a useful way to think about AI?
Nowadays, when people hear about AI, they think of a Corgi running on a beach made with text-to-video. That’s an example of Generative AI, but there’s lots of other types of AI.
To chart that complex landscape, I often use the metaphor of the left and right brain. You get creative forms of AI which are like the right-hand side of the brain – this links to how AI is used in a marketing context.
The left brain is where we make more optimal and rational decisions – in a marketing context, this could be analytics, automated decisions or predictive decisions.
Why is it useful to think about AI in this way?
When you compartmentalise AI into these two different roles, you can apply it to how you want to apply AI.
For example, let’s say you’re spamming your customers’ emails. To achieve a closer one-to-one approach, you’ll probably want a creative, right-brain AI to decide that we need to engage with an audience through interesting content – but then you need the left brain to rule out what content to serve audiences based on classical business rules, for example.
What are some of the ways AI is influencing audience personalisation and engagement?
A more hardcore left-brain AI can help target what’s the right thing for that customer to see in the moment. If you take generative AI and give it more autonomy, then you get agentic AI. It uses generative AI and vast language models to understand what a marketer’s goal is, and then decide what actions they need to achieve that goal.
So when an ‘agent’ works, it comes up with ideas but then also has the capability to challenge those ideas and start to negotiate.
An agent is constantly talking to a large language model (LLM) or search engine to reach its goals. Agents are quite chatty in that sense, and it’s constantly talking with its environment.
What impact can agentic AI have on the marketing industry?
ChatGPT is built on the world’s knowledge and it’s very generic. You need to know how to prompt carefully and what questions to ask with chatbots. Many of the enterprise use cases where there is real value comes from when you build features that leverage these models in the back-end.
In this sense, you’re hard-coding the AI with best practice, variations, persuasions and thoughts – basically with whatever you want. That’s agentic AI.
What does the future of agentic AI, in relation to marketing, look like?
I think we’re just scratching the surface. The best way to understand agentic AI is to get hands on with it. I think people need to actually start to use agentic AI. When we start to use them then we can develop them to do better jobs at reasoning, planning, and reflection on their own – they get better over time.
Eventually you’ll have different agents who are collaborating and coordinating with one another. – so that will be ‘swarm intelligence’.
I think in the future, marketing will become a lot more evidence-driven as a result. There’s so much to gain from leveraging AI to make closed loop evidence-based marketing a reality.
