For the positive technology companies we partner with, the difference tends to show up in the quality of what gets built and the pace at which decisions get made. Not as a single dramatic shift, but as a compounding effect across the engagement.The more pressing question for most of them right now is not about delivery. It's about integration.

Where does AI actually fit in our organisation? How do we build it into processes, decision-making, and cross-functional operations in ways that hold up, that the team understands, that we can actually govern?

What makes this hard is rarely the technology. It's the organisational layer underneath it. AI tends to surface existing ambiguities: unclear ownership, processes that were never properly documented, decisions that were being made informally without anyone naming them as decisions. When you introduce a system that requires those things to be explicit, the gaps become visible in ways they weren't before.

That's where the real integration work happens. Getting the technology to function is usually the straightforward part. Getting the organisation to function with the technology in it is a different kind of problem, and it's one that sits squarely in the middle of what transformation work actually involves.

Currently, models still hallucinate. They can flatten nuance. They generate outputs at volume, and strategic context is what makes those outputs useful. Understanding where a model's confidence ends and where judgement needs to take over is something you develop by working with AI seriously over time, including through the outputs that don't hold up.

There's something else worth naming. Working remotely means that most of what we do already exists as a recorded artifact. Decisions get written down. Analysis lives in documents. Conversations happen in threads. It's what makes it possible to apply AI to the actual work rather than to a summary of it. The material is already there.

Every piece of AI-assisted work still passes through human validation: strategic review, domain expertise, and the kind of judgement that comes from years of working with complex technology businesses. AI accelerates the process.

People govern it. That sequence doesn't change.The same thinking applies when we advise. The goal is finding where AI genuinely multiplies organisational capability, not where it can be applied for its own sake. Those are different questions, and they lead to different recommendations.

Our perspective on AI connects directly to the companies we choose to work with. The organisations we partner with are building things in biotech, clean energy, and the technologies taking on problems that have real stakes.

For those companies, AI applied with precision and care is a multiplier. It extends what their teams can do, accelerates the work they're committed to, and helps them reach the markets where that work has impact. That context forms how we think about AI in practice, and it enables the standard we apply when we bring it into client work.