Involve your people: AI is reshaping work for over 1.3 billion people worldwide. And it’s scary for workers to think about how or when their job will change — will they need training or a completely new career path? To get buy-in from employees who will work with or delegate work to an AI, it’s important to involve them as early as possible. Involving them with the implementation process and educating them about AI will develop their trust and understanding of AI and improve the rate of adoption.
With technological advancements, particularly in the AI space, an increasing number of tasks can be either fully or partially automated. In this series, we are talking to People experts about how they’re utilizing new technologies to make their jobs easier and provide greater strategic value. As a part of this series, we had the pleasure of interviewing Casper Wilstrup.
Casper is the CEO and founder of Abzu®, the Danish/Spanish startup that builds trustworthy AI that tackles challenges for the world’s leading companies. Casper is the inventor of the QLattice® symbolic AI algorithm, an explainable AI that rationally reasons and makes evidence-based decisions. He has 20+ years of experience building large scale systems for data processing, analysis, and AI, and is passionate about the impact of AI on knowledge work and the intersection of AI with philosophy and ethics.
Thank you so much for your time! I know that you are a very busy person. Before we drive in, our readers would love to “get to know you” a bit better. Can you tell us a bit about your ‘backstory’ and how you got started?
My adventure with computers began in 1981 when my buddy, Allan, showed up with a Sinclair ZX81. It was super basic compared to today’s tech, but for me, it was revolutionary. We hooked it up to a TV and started coding — pure magic.
I soon got my own computer, a ZX Spectrum. It had some colors and 48 kilobytes of RAM. I got it to chat with me, asking questions like, “Who are you?” and “How old are you?” I soon figured out that to really get the most of it I had to code it in Assembly, and I moved into developing arcade style games. That’s when I got hooked on the blend of code, computers, and human interaction.
In the late 80s during my grade school years, I started to think about what I wanted to do when I grew up. Even though I was still fascinated with coding and computers, it was such a new concept that it wasn’t considered something you could choose as a career. Even today it’s a little misplaced, right? “Computer science” is something you use for something else. You could consider yourself a computer scientist, but aren’t you really an insurance analyst or a meteorologist who uses a computer as a tool? So, I ended up in Physics at the Niels Bohr Institute here in Copenhagen.
But as one of the relatively few who were good at coding, I got pulled into a lot of things outside of class like writing code for the tandem accelerator laboratory down at Risø. Which is an unthinkable opportunity today for a first- or second-year student in university! But such things happen when you’re suddenly identified as an expert in a new and growing field. These opportunities significantly shaped my direction, because although I received good grades in my undergrad classes, these opportunities to build computers and code to power physicists inspired me a lot more than being a physicist myself. So I dropped out of university to build supercomputing clusters and develop the software designed to operate and optimize these machines.
This was the start of my career as an entrepreneur and active member of the global tech startup community. My first company survived the dot-com bubble burst, and since then I’ve built several more and worked as a technical due diligence consultant for VCs. But over these last 25 years, I never forgot an idea that came to me back when I was at the Niels Bohr institute: How to build a new approach to artificial intelligence.
It has been said that our mistakes can be our greatest teachers. Can you share a story about the funniest mistake you made when you were first starting? Can you tell us what lesson you learned from that?
I’ve been in tech for 30 years, and there are only few mistakes I have made ;) Here’s a list:
- Not backing up for days and losing a physics report the day before it was due.
- Turning off a server and inadvertently bringing a system with thousands of users to a standstill.
- Shipping code to be burned into a hardware device with obvious bugs, which I could have easily found if I hadn’t been coding late into the night.
- Showing up at a venture pitch with code that I had finished the night before. I had convinced myself that I absolutely needed that feature to impress, only to have the entire demo fail.
- Bricking an expensive network router by configuring it to accept network access only from its own IP.
What have I learned from this?
Life goes on.
None of us are able to achieve success without some help along the way. Is there a particular person who you are grateful for who helped get you to where you are? Can you share a story?
One person who’s been a game-changer for me is Lars Thorup, a computer scientist from Copenhagen. I met him when I was a student worker at Scanjour, a software company. Lars was super smart and a no-nonsense kind of guy. Everyone looked up to him, especially when it came to anything software-related. He was an early fan of Extreme Programming, an approach to coding that was all the rage in the late ’90s. He even translated a famous book on the subject into Danish.
So, Lars left to start his own consulting firm. Inspired by what I’d learned from him, I started my first tech company. After selling that business, I teamed up with Lars and worked closely with him for six more years. Then his love for adventure took him to Silicon Valley, where he started another consulting business.
Can you please give us your favorite “Life Lesson Quote”? Can you share how that was relevant to you in your life?
“If liberty means anything at all, it means the right to tell people what they do not want to hear.” — George Orwell.
Although there is much I disagree with Orwell about, I admire his stance on letting everyone, even his enemies, speak their mind. He fought for this principle both with words and in action, notably in the Spanish Civil War.
I am adamant about freedom of thought and speech, and this view is super relevant to how I lead. In any team or project, you’re gonna have a mix of opinions. Sometimes these opinions might seem odd or even rub people the wrong way. But that’s no reason to shut someone up. I actively seek out different viewpoints, both inside and outside the company. I want people to feel safe sharing their thoughts with me, even if they think I won’t agree.
Because shutting down dialogue just because you don’t like what you’re hearing is a quick way to stifle innovation and progress. It’s like saying you’ve got all the answers — which nobody does. I especially hate it when someone tries to silence others just because they’re offended. Being offended is a reaction, not an argument.
So, in my leadership style, I go out of my way to listen. I want to understand why people think the way they do. It’s made me a better leader and helped create a culture where everyone feels their voice matters.
Thinking back on your own career, what would you tell your younger self?
Sure, some choices can make you more likely to succeed, but it’s not just about the end game — it’s also about enjoying the ride. Back in my VC days, I told people to build companies around real problems that needed fixing, rather than cooking up solutions in search of a problem.
But you know what?
The coolest thing I ever did was ignore my own advice and start Abzu, which is all about making AI we can trust. The thrill of creating something new and the pure creative buzz has more than made up for not having a super clear business plan at the start. My take? Do what makes you happiest at work. That happiness is what’ll keep you going and might even lead to better results. And even if it doesn’t, you’ll have had a fun and interesting ride, and that’s what life’s all about.
Let’s now move to the central part of our interview. How have recent technological advancements such as AI made your job easier?
It’s made my job as an inventor of artificial intelligence easier and harder! Five years ago, our conversations centered on educating people about AI, and we absolutely had to convince people that it would significantly change the way humans work. We were still in research and development mode back then with our algorithms, and we found some thoughtful, forward-thinking individuals — in industries with complex, highly-regulated problems to solve like pharma — that saw the value of a trustworthy and explainable AI.
Today, everyone has embraced the fact that AI is significantly changing the way humans work. So, our conversations have progressed from that. But the conversation is fundamentally more complex in how this will (or should) come about. Will it be because of regulation? Will it be because we, as humans, will define what work should be human work? What is creativity? And even: What is thought? What is consciousness?
These conversations are harder because we have arrived at the intersection of AI with philosophy and ethics, and even government. But I believe the answer is the same today as it was five years ago: There is immense value in an AI that is trustworthy and explainable. If you are working with anything that matters — personal information like someone’s genetic profile or a complex, high-impact process like aircraft traffic — you simply cannot trust a model that makes a prediction and cannot explain its reasoning.
In which processes do you utilize automation the most?
I use automation mainly for data processing tasks, from formatting to analyzing data. Especially with AI in the picture, these tasks can be done much faster and with fewer errors. Asking a human to process data is like the tech world’s version of sticking labels on soda bottles all day — a job I actually did for an entire summer in high school.
What should people bear in mind when automating processes?
I think we — humans, organizations, civilizations — have only scratched the surface of identifying what work should be human work.
AI can do so many things, but we aren’t often asking should it. For example: Should an AI care for our children? It certainly could. But I think we as a civilization would agree that it is a fundamentally human activity to care for our young humans, and I daresay that everyone would find the idea of automating childcare with an AI repulsive.
But should an AI teach our children? It certainly already is. Does that mean AI will replace teachers?
I hope not. I hope we see AI assisting and augmenting teachers, not replacing teachers. But depending on the level of augmentation, we are entering a gray area.
So how about in organization: Should an AI hire a person? Should it fire a person?
Should it automate customer service or assist a customer service representative? Again, we are entering a gray area. AI surely can do all these things, but let’s have a conversation about whether it should. And not solely from the business perspective of “increasing shareholder equity,” because in many cases an AI can do the job better under that criterion.
What are your “Top Five Tips For Harnessing AI Technology to Propel People Operations”?
1 . Define your objective: When you’re implementing an AI, the first question you want to ask yourself is, “What do I want to achieve?” It’s fun and trendy to use AI in a piecemeal capacity, but if you’re preparing to implement an AI to augment or automate a task, start with defining what you want to achieve. Do you want to be more efficient? Produce a higher-quality product or service? Generate more profits? Provide a better user experience? These things are all certainly related, but setting a primary goal with your implementation will determine to a large degree whether you achieve your goal.
2 . Involve your people: AI is reshaping work for over 1.3 billion people worldwide. And it’s scary for workers to think about how or when their job will change — will they need training or a completely new career path? To get buy-in from employees who will work with or delegate work to an AI, it’s important to involve them as early as possible. Involving them with the implementation process and educating them about AI will develop their trust and understanding of AI and improve the rate of adoption.
3 . Know your data: As the saying goes: “Garbage in, garbage out!” But with AI applications, there is an additional layer of data bias that could result in a veritable dumpster fire. Before starting a project, know where your data comes from, understand any biases or how your data might be skewed. If you don’t consider these possibilities, it may contribute to a system that behaves in ways that are hard to defend. Biased results — especially with personal or sensitive data — will ultimately erode the trust in your AI and possibly your reputation as a company.
4 . Be customer-centric: Even if you aren’t customer-facing, the ultimate beneficiary of your work is your customers. Especially in these early days of AI, you need to be particularly mindful to establish trust — and the people whose trust you need the most are your customers. No matter where in the business you implement AI (directly in product or service delivery, or indirectly in organizational processes), consider the impact on your customers. Implementation of AI (especially a poor one!) will have significant consequences on how they perceive your products or services and you as a company.
5 . Choose trustworthy AI: One of the most important criteria for humans in trusting machines is transparency. If you can see under the hood — understand what is actually happening — that builds trust. When there are so many types of AI to choose from, make sure to choose a trustworthy AI: An AI where you can see and study the reasons for why the system made a recommendation or decided on a certain action. In these early days of AI, and especially in business processes that are critical and complex, you need an AI you can understand, believe, and trust.
What are your favorite “I couldn’t live without these” tools?
Even though I’m the CEO of a growing company, I still find time to code — often during the evenings and nights. My go-to tools for coding are:
- Linux: I’ve been using this OS since the early ‘90s.
- VS Code: The code editor that finally won me over from EMACS.
- GCC: The GNU Compiler Collection.
But since this is a productivity interview, let me also share some non-coding tools that have had a huge impact on how I work:
- ChatGPT: Well, that’s a given nowadays, isn’t it?
- Google Scholar: Though it’s increasingly being challenged by various AI-powered science search tools.
- Overleaf: An academic writing tool
- Anki: A spaced repetition tool that helps with memory.
- My hiking shoes: I can’t truly engage in deep thought without going on long hikes.
How do you see technology impacting the HR space in the future?
There have already been some disastrous AI implementations in the HR space which have all involved biased models acting on personal or sensitive data. Some of these instances are from quite a long time ago when we saw AI as just a tool that could summarize, extract, or locate data, and at the time people weren’t even considering the implications of using biased models.
I think it’s quite tiring to only talk about “operational efficiencies” — of course we’ll see processes sharpened so that time and costs are reduced — but in the future, I’d like to see technology really addressing the future of work. I think that can go beyond making virtual work or hybrid work more accessible; I hope the future of work means making work more human.
The HR space in the future should be better at understanding what humans are best at, what makes humans happy to work, and how to challenge the right people in the right ways. With AI assisting and augmenting work, we should have more time for employees to be trained and upskilled so that they can create and do the work that matters to them. It would be a disaster if humans just did the same kinds and same amount of work, or if humans weren’t given an opportunity to grow and expand into more human work and were made unemployed or unemployable.
We are very blessed to have some of the biggest names in Business, VC funding, Sports, and Entertainment read this column. Is there a person in the world whom you would love to have a private lunch with, and why? He or she might just see this.
If I could choose anyone in the world for a private lunch, it’d be Marc Andreessen. We were born the same year, and I’ve got this feeling that he gets as nostalgic about his first Apple II as I do about my ZX Spectrum. I remember playing around with the WWW when it was still a new thing, seeing its massive potential as a game-changer that could replace gopher, archie, and usenet, and I was blown away by the early versions of Mosaic and Andreesen’s work on Netscape.
Beyond tech, I’m pretty sure we’d have more to chat about. For instance, I’d bet he also spent years of his youth fascinated by Alex P. Keaton from “Family Ties,” portrayed by Michael J. Fox.
When it comes to philosophy and worldview, I think we’re kindred spirits in a way. I consider myself a strange blend of a European liberal and a free-market enthusiast. I’m also a dreamer and a technologist at heart. Sure, he’s taken his tech career a bit further than I have, but that would just make the conversation even more enriching for me.
So yeah, a lunch with Marc Andreessen would be pretty amazing. We’d have a ton to talk about, and I’m sure the time would just fly by.
How can our readers further follow your work?
I’m very active on X — you can find me at @cwilstrup. I also do a lot of writing on Medium, and you can of course always learn more about my work in trustworthy AI at www.abzu.ai.
Thank you so much for sharing these important insights. We wish you continued success and good health!