Ringier's Nils Körber on ArtificiaI Intelligence, Training the Machine, and Talking about Problems

"What would you do with one million interns?"

The award for the best question at the PropTech and Portal Watch conference in Bangkok went to Nils Körber who posed a fascinating (if surreal) question to the audience during his keynote presentation about Ringier's journey to the AI-powered marketplace.

Nils is the Managing Director of Ringier South Africa, a software development hub providing tech solutions for many well-known classifieds platforms and property portals around the world including Imobiliare in Romania, BuyRentKenya and Reality in Slovakia.

I wanted to understand more of the motivations behind Nils' presentation, fight vs flight in the face of emerging technologies, the industry's seeming unwillingness to discuss its problems, and preparing for the next generation of workers, so I asked him...

Nils, give us a quick outline of your keynote presentation in Bangkok.

I called it our journey to the AI-powered marketplace.

[At Ringer] We are asking ourselves what we as a marketplace software provider can do with artificial intelligence, just like everybody else. We've done machine learning, recommendation, and personalization, which is a good start. But where do we go next?

We need to find a strategy.

I see that in three layers. The first is 'how can we improve the efficiency and performance of our own employees?' Not only software engineers, but also the employees across our marketplace businesses. In other words, engineering, sales, marketing, customer care and HR—how can we with AI support them to do a better job?

The next layer is the back end. So from a platform point of view, how can we enable engineers to build better AI-driven software? What is needed for that?

And then the last layer is more the consumer-facing part.

By dividing the strategic and tactical uses of AI into three digestible parts has made it much easier for us to parse this topic without getting overwhelmed, and my talk was about sharing these insights.

Below: Nils Körber on stage at PropTech & Portal Watch Bangkok 2024

If I operate a marketplace and want to start using AI, where should I look first: the back end or the front end?

The back end, for sure. Why? Because we all love to text. Let me explain.

How do we use our laptops? How do we use our phones? How do we interact with apps? We type, right?

Texting is ingrained in us. and it is fairly difficult to change this innate behaviour in one day. This will take some time. We should focus our efforts on building your product for internal use rather than on bells and whistles that our users aren't used to yet.

A user behaviour shift needs to happen before we can revolutionise the front-end experience for consumers. We are not there yet so it is better to make incremental improvements in the back-end for now.

Are you predicting an evolution from text to image or video? Where is the search experience headed?

Yes, this is a good question.

It's pretty easy to fall into the hype of Apple Vision Pro, AR, and the like. I believe it will take some time until that tech is really there for us.

But I think LLMs (large language models) will fundamentally change how we interact with the machine.

I also think voice is going to be reborn. Siri and Alexa almost messed it up.

Why did they mess it up? Because they have always been able to process what you say pretty well, but at the time when they were released—without LLMs that didn't exist—Alexa and Siri couldn't give you a satisfying answer.

It was a one-way conversation that limited the usefulness of the early Siri/Alexa product experience.

In the recent past, for me to use my voice to trigger a search in the marketplace hasn't made sense. If I use my voice, I want that voice experience to be rich, not just replicating what you can type. We must overcome this hurdle before we can truly extract value from a voice search.

What skills have I inherited thanks to AI?

It is difficult to comprehend, but with AI you have inherited all human knowledge—and a workforce to match.

This was always the case with Google but in the past, you had to process it yourself. But now the machine is writing for you, generating for you what you want to know about human knowledge.

This means we can now parse market-related data and formulate a strategy out of that, just by asking the question.

This is where human interference is needed. When it comes to strategic stuff, the machine can't do that yet, but collaboratively that potential is there.

As a businessman, today you have infinitely more support. You no longer only have your normal teams available. And you know, like, managers of sales and marketing will report back to you and you can give them commands. Suddenly you have of these resources available across the board, and they can work for you.

Your presentation asked, 'What would you do with one million interns?'

That's enough to trigger my fight-or-flight response! Where are the quick wins for a technophobe like me?

I think there is a rather tremendous mind shift that needs to happen on this topic. It is therefore a difficult question to answer without getting too high-level and philosophical, but I will try to answer it!

The short answer is mastering prompt engineering.

I think we are quite stuck within this business-focused structure whereby we have teams, onboarding processes, and knowledge transfer from person A to B that needs to happen for the business to succeed.

If a new hire joins, there is a bedding-in period for them to learn the ropes before they can start meaningfully contributing to the business—even for a CEO. The same goes for AI.

The key is to break down your strategy into achievable tasks and divvy them as if the AI were your team member, then put the AI to work once it has been trained according to your processes.

And because AI is getting smarter, one million interns quickly become one million junior-level people, then mid-level people and at some point even senior people—either way, it is about getting accustomed to, and mastering, prompt engineering.

We can now tell a machine to perform all types of tasks. It's just about pushing our creativity and moulding the AI to what we need.

What isn't the industry talking about that it should be talking about?

Budgets, and the transparency of problem-sharing.

It isn't so much the budget itself that we need to talk about, but I want to know how businesses are extracting revenue from budget investments. I would like to see more transparency in which products and services are performing well for other businesses and watch someone get up on stage and be more open about how 'hey, I messed this up and this is what we will do about it and this is the return we can get from this level of investment.'

I would like to see anonymous submissions from attendees at a conference being brutal about their failures and then we can work through them together as a roundtable discussion. If someone wants to stand up and say 'That was me and it's the biggest mistake I made in the past 12 months!' then maybe we can continue that conversation at dinner later.

What has your biggest failing been in the past 12 months?

You're going to laugh, but it's being four to six months late on generative AI.

Of course, we knew it was going to happen, and we knew about the impact, but until we came up with our little strategy it took us a while to realise that wow, we need to do something here. It feels like we came to that conclusion a little late. Not years late, thankfully, just months late.

What's the next big problem?

From my responsibility, which is predominantly tech teams, we have to increase the efficiency of our developers.

The big guys are doing this already and that's why we are seeing all these tech layoffs. From my point of view, The Big Four and McKinsey are being a bit aggressive, but in essence what they are saying is that in the next five to ten years, you have 60%-80% efficiencies, and efficiencies in that vocabulary mean you can let 60%-80% of the people go because the machine takes care of their jobs.

For me it is a problem because I like my team! And this cost pressure will only come to mind more and more. Tech salaries are high and it's a big line item on our total budgets.

We will need to be 20% more efficient but have the same output. My mandate as a manager and as a business owner is to understand how to make that happen.

One more question: Digital Natives Vs. AI Natives. Is the workplace going to place a higher value on creative thinkers who are more in sync with AI and can think laterally in tandem with emerging technologies?

The 18-24 bracket is going to be the first generation that is not only internet native but will be used to having everything available to them, all the time, 24/7.

This will be the first generation of AI natives, and they are already being underestimated. People lazily brand them as "the generation that doesn't want to work". They may be ten years old today but they will be used to having a whole life experience that is generated just for them at their fingertips.

We will need people like that and we need to take that generation more seriously.

Can we get a testimonial for the conference?

I don't think there's a better conference for the marketplace industry. I attended some other conferences more local to me in Europe and Africa. For me, the set of great people, such as real business leaders and industry leaders for online marketplaces, make you unbeaten.