Decoding the GenAI hype wave
Chris Kindt's view from the frontlines
Decoding the GenAI hype wave
Chris Kindt's view from the frontlines
Chris Kindt, Hg's Head of Value Creation, delivered a data-rich reality check on generative AI's impact across Hg's portfolio of 50 companies. Drawing from 300+ live GenAI projects, his presentation cut through the hype to reveal what's actually working in enterprise software today.
The platform shift accelerates
"We're moving multiples faster than the previous platform shift," Kindt emphasised, comparing the current AI transformation to the decade-long migration from on-premise to SaaS. While it took 10 years for SaaS to meaningfully displace on-premise solutions, early indicators suggest AI disruption could happen in a fraction of that time.
The evidence is already visible. For instance, Accenture's most recent quarter showed non-AI business lines flat or declining, with all growth coming from AI services.
Real impact in real companies
Kindt's data revealed striking productivity gains across the portfolio. Developer productivity has jumped by more than 30% through AI coding assistants, with some CEOs already experimenting with one or two-person engineering teams. But the truly transformative shift comes from moving beyond productivity enhancement to complete task automation.
One compelling example came from Hg portfolio company, CINC, which provides technology for homeowners associations. Their AI agents have compressed a 4-5 hour board pack preparation process down to just 10-20 minutes of human verification. "This is what people are talking about when they're talking about a multiplication of the market and the value," Kindt explained. "It's that 10 to 15x."
Models are good enough - now what?
A crucial insight: the models have reached sufficient capability for most transformative use cases. On benchmarks where PhD students score around 30% accuracy, current AI models now match or exceed human performance. The bottleneck has shifted from model capability to implementation.
The real work now lies in creating "strong process context" - understanding the intricate workflows, compliance requirements, and domain expertise that separate a demo from a production system. Success requires deep customer knowledge and the ability to navigate what Matthew Brockman had earlier called "real world friction."
Fierce competition from "small but mighty" entrants
The competitive landscape has fundamentally changed. New entrants aren't typical startups - they're "small but mighty businesses" achieving massive scale with minimal headcount. Cursor, with fewer than 50 employees, has reportedly surpassed $300 million in revenue. These companies combine brilliant products with AI-first operations to achieve previously impossible operating leverage.
Traditional advantages of scale and distribution remain valuable, but incumbents need "new levels of product agility and organisational agility" to compete.
The AI-first culture imperative
Perhaps Kindt's most actionable insight concerned organisational transformation.
Companies seeing real impact aren't just deploying AI tools - they're cultivating AI-first cultures from the bottom up. Success stories share common elements: CEO manifestos, protected innovation teams, strategic acquisitions of AI talent, and most importantly, clear expectations that AI adoption is mandatory, not optional.
Surprisingly, some of the highest-impact employees - the "10x employees" who grab AI technology and infect their organisations with change - were previously underperformers.
"What's intriguing is that probably inside of our organisations, we already have our 10x employees, maybe sleeping, ready to be activated," Kindt noted.
"What's intriguing is that probably inside of our organisations, we already have our 10x employees, maybe sleeping, ready to be activated,"
The path forward
Kindt's prescription was clear: embrace the platform shift fully, and above all, spark organisational change. The companies that will thrive are those that combine their existing advantages - customer relationships, domain expertise, and proprietary data - with the agility to experiment and adapt at start-up speed.
