Does AI's future depends more on data quality than model size?

[Daz]

A Computer Weekly opinion piece by a data and analytics lead at HSS ProService Marketplace argues that as AI models become more powerful and widely accessible, the technology itself becomes less of a competitive differentiator, and the biggest difference between businesses using similar AI capabilities will come from the quality and relevance of the data supporting those models. The piece pushes back against years of conversations focused purely on collecting more data, arguing that more information doesn't automatically mean better insight, data only creates value when it helps make a better decision, catch a problem in time, or get the right information to the people who can actually act on it

Drawing on the author's own marketplace business, which connects customers, suppliers and services for hire equipment and building materials, the piece describes using customer interaction data, website behavior and supplier performance signals to catch problems like missed delivery expectations early enough to actually intervene, rather than simply measuring what already went wrong after the fact. The author argues good data leadership means judgment about which information is genuinely useful and reliable, not complexity for its own sake, and that adding more sources without consistency or proper structure creates additional confusion rather than better intelligence

The piece concludes that implementing AI properly takes real time specifically because it requires the right data infrastructure underneath it, and that businesses benefiting most from AI will be those that took time to understand their own data first rather than simply adding a new AI tool on top of existing messy systems. Curious what people think about this framing, is data quality genuinely becoming the main competitive differentiator as AI models themselves become commoditized, or does raw model capability still matter more than this piece suggests

The Future Of Ai Depends On The Quality Of The Data Behind Itwww.computerweekly.com

[PulseRider]

Technology itself becoming less of a differentiator as models commoditize feels like the right long term prediction, once everyone has access to similarly capable frontier models, the actual data feeding them becomes the real competitive edge

[Oscar_57]

Catching problems early through data signals rather than just measuring what already went wrong is a genuinely useful distinction, most business reporting historically has been purely retrospective rather than actionable in the moment

[RayOfLight31]

Curious how this plays out for smaller businesses without dedicated data teams, understanding your own data properly sounds sensible in theory but requires real expertise most smaller organizations genuinely don't have in-house