THE TECHNOLOGY INTERVIEW
The company is trying to compete across the entire stack, from computing infrastructure through to ready-made AI applications.“ We have infrastructure for people that want to build on our AI infrastructure,” Oliver explains.“ We then have models: our Gemini models, other models like Lyria and Chirp that are very specific to certain tasks. We have a platform so developers can build on that platform but take advantage of not just our models, but anybody else’ s models.”
Google’ s newest product, Agentspace, takes a different approach to enterprise AI adoption. Instead of requiring companies to rework their entire technology stack, it tries to embed AI capabilities into the tools people already use. It’ s a recognition that most successful AI implementations happen when they solve immediate problems for real users, rather than requiring wholesale organisational change.
The companies getting it right The best way to understand what works is to look at companies that have moved beyond pilots to genuine business impact. Oliver points to Screwfix, the British trade tools retailer, which built an AI-powered visual search system for its website.“ I sat down with them nine months ago at the London Summit, and they were just trialling it out. Now, 50 % of the users on the B & Q website are using Screwfix Lens.”
technologymagazine. com 31