Inside the Corner - December 2025
AI Risk - Consultants vs Software? Our Portfolio, Mistakes this year & Highlights of best performers of 2025.
This Sunday, I listened to Christoffer Callesen in the StockUp podcast (link) on my way home from a week of work in the snow-covered Norwegian mountains.
His insights made me think deeply around AI risk. In particular, his example illustrated the power of being Counter-positioned — a situation where one player benefits from adopting new solutions whereas someone else (often the incumbent) would cannibalize their current profits doing so. The business graveyard is full of such examples, with Blockbuster vs Netflix perhaps being the most known. Callesen used a firm we (used to) own, one we’ve now decided to part ways with.
1. Parting ways with our Consultants
20 months ago, we presented an optimistic view on IT consultancy Bouvet. Today, after listening to Callesen, we’ve changed our mind.
Callesen illustrated AI risk for consultant firms through Consigli, a firm with AI-driven project-planning tools for construction developers at roughly 1/10 the cost of human consultants. Junior consultants — typically handling repeatable, simpler tasks yet invoicing expensively — are particularly vulnerable. Senior consultants, dealing with complex problems with their extensive know-how, seem less at risk.
The CEO of Bouvet frames AI as another leg for growth, while Callesen points out the vulnerability within pockets of these firms. Maybe both are right — but as investors in a competetive world, we have to consider who is incentivized to attack them?
We believe the tech giants are.
Microsoft and Google have deep pockets and no internal risk of cannibalization here. And even in the more specialized fields, players like Nemetschek, already entrenched in architecture software, could integrate AI solutions reducing demand for traditional consultant services. In short, these firms are counter-positioned against the consultants (in our opinion). And whereas you typically see startups being counter-positioned, it’s not calming seeing some of the best capitalized and talented companies in the world being incentivised to disrupt you.
Further, at 16-17x earnings, Bouvet stock does not adequately compensate for this risk. Opportunity cost matters. For a similar multiple, we could for instance buy convenient store giant Cirkle K — recession-proof, stable, and free from such threats.
2. Could I Be Disrupted Too?
When thinking through this topic, I appreciated how relevant it is for my own practise to. My service business focuses on exercise physiology and endurance sport, helping athletes measure and improve their performance.
Just as AI is reshaping consulting, it is beginning to impact my own field. Athletes increasingly use tools like ChatGPT to generate training plans for free, rather than paying a coach. I’ve tried it myself — it works surprisingly well. But it works best for strength, running, and cycling, where abundant, high-quality data allows AI to produce effective plans as long as the prompt contains sufficient relevant information.
The key insight: in these fields, a slightly lower-quality solution can disrupt a higher-quality offering if it is cheaper and available instantly. At least for the recreational athlete and in large sports (also the places where the big money lies).
We think this is a similar dynamic to consultant firms and their junior consultants invoicing for repeatable, “simpler” tasks.
By contrast, more niche focus — like those I do within physiological testing, skiing, and polar expeditions — is far less exposed. These are small markets with limited data, where expertise cannot be easily codified. Also, we believe just the nature of operating within a niche, make something less disruptable.
The interesting thing for our portfolio, is that we have a similar view on Vertical Market Software (VMS). These markets are too small to attract large competitors, but technical and complex enough that proprietary data is essential to deliver effective solutions. And with Mr.Market starting to price in this AI risk being real, we’ve decided to make a VMS company our Top Idea, surpassing Investor for the first time.