Think Ahead - AI-powered growth in uncertain times
Strategy first, clarity over hype: how to turn AI into meaningful, measurable value
In a fast-moving Think Ahead session on 27 November, London Business School convened Keyvan Vakili, Associate Professor of Strategy and Entrepreneurship; Academic Director, Data Science and AI Initiative, Rebecca Dykema, Technology and Media Business Builder, and Matt Altass, VP Strategic Growth, Array, with journalist-moderator Charlotte Kan to cut through AI hype and focus on value.
A live poll set the tone: the biggest blockers to AI impact aren’t algorithms but organisations' lack of strategy and vision, plus unclear ownership and fragmented efforts. Talent shortages and ethics concerns trailed, suggesting the conversation has matured from “can we?” to “how do we, coherently and responsibly?”
Vakili drew a sharp line between predictive AI (decision support) and generative AI (content), with “agentic” systems now chaining tasks and triggering actions. His core message: AI serves strategy, not the other way around. Start with a business problem, map processes, and deploy narrow, high-return applications; experiment cheaply, then scale what works, while managing stakeholders, incentives and revenue-model knock-ons.
Dykema warned against “tools-first” theatre and pointed to systematic, measurable wins: when firms embed AI into operating systems (not one-off campaigns), ROI follows, citing Nestlé’s global, AI-driven creative standards that materially lifted media returns. She also showcased AI for inclusion, using models to audit ad diversity at scale and debias workflows when organisations define clear standards.
From a legal tech perspective, Altass showed how AI is already compressing cost and cycle time in document review and contract workflows, shifting skilled work rather than erasing it, while underscoring the need for verification in a world of hallucinations and model poisoning. His mantra: augmentation beats automation, and training is non-negotiable.
Across regulation and risk, the panel contrasted US “move fast” pragmatism with the EU’s precautionary stance, agreed that data security and governance are table stakes, and were sceptical that self-regulation alone will suffice.
The through line: strategy first, focus over frenzy, and human judgment at the centre proved to be a central message. Use AI to sharpen decisions, raise the floor on skills, and scale inclusive, reliable systems, then double down on what demonstrably moves the needle.