AI has come a long way since the first chatbots were used on business websites, where they were of some but limited use. Now, AI agents can do much more. An old-school chatbot can retrieve the most popular holiday destinations; an AI agent can book your flights and hotels and make restaurant reservations.

In short, while human labor isn’t going anywhere anytime soon, it will increasingly be supplemented with ‘digital labor’.

For instance, real estate businesses are currently using AI agents to interact with tenants. The Real Estate Institute of New South Wales has even appointed an AI agent board advisor to deliver ethics advice.

And it’s still early days. There’s not much hard data about AI agents’ productivity enhancing or revenue-boosting impacts. However, the initial anecdotal evidence suggests the efficiency gains are both real and substantial. In fact, a recent survey of 13,000 generative AI-using workers by Boston Consulting Group found that 58 percent of them were saving at least five hours per week.

Working with AI

Treasury Wine Estates (TWE) counts Penfolds and Lindeman’s wineries among its brands, employs 2,500 staff across 70 countries and has a market cap of almost US$6.63 billion.

Unsurprisingly, it also has an enormous amount of data, which wasn’t being leveraged as fully as it might be. As TWE’s Managing Director, Angus Lilley, has noted: “We have more data than we have ever had before – from a various range of insights across the company, the business – and we really needed a way in which we could organize that data into one point that the teams could access, and even more importantly and powerfully, had a way in which we could actually distill that data down into useful insights.”

That may sound abstract, but the results were concrete and immediate once TWE staff accessed their ‘virtual’ coworkers. TWE’s Head of Global Category, Adam Smith, recently even used TWE’s AI agents to handle an unexpected client query about sangria.

“I don’t know how I would have answered that question,” he admits. “I would have said, ‘Leave it with me’, and gone off and done something for a week, and then come back with it. But I was able to answer the question in five or 10 minutes… that was a great example where it was a bit of a game-changer.”

TWE’s Global Consumer Insights Manager, Kate Duggan, is also a convert.

“Bringing consumers to life through AI has really helped build the visibility and simplicity of our consumer segments,” she explains. “And I think this is like a little bit counterintuitive because, for a lot of people, AI is a little bit scary and daunting.

“But instead, what AI is doing for us in this space is it’s bringing what we previously had as 100-page decks of consumer insight presentations, and making them more visible and able for people to interpret them. And, in that way, we’re making this information more accessible. Not only to our business, but also the way that we share this information with our customers.”

AI agents in workforce

AI agents are an emerging technology, and I don’t want to tempt fate by making definitive long-term predictions. However, I’m confident they will have the following impacts, probably sooner rather than later.

Firstly, organizational knowledge will be deployed differently. AI agents will be the interface through which organizational knowledge is deployed. Secondly, the playing field will be leveled – businesses with limited resources can access AI agents for research, analysis and customer engagement.

This will allow them to compete globally and against larger and slower-moving incumbents.

Finally, AI agents will put a (much-needed) rocket under productivity. Productivity gains will be experienced by organizations in two phases. The first will come from doing what they currently do faster.

The second (much more exciting and potentially lucrative) phase will arrive when the ‘nature of work’ itself starts to change profoundly.