Cutting AI budgets won't fix token shock - Neo4j's Jim Webber on graph RAG and the price of accuracy
A small caveat before I begin - this is not a token-cost story. Is throttling AI budgets actually fixing anything? Enterprise controls such as prompt-length limits and per-seat token caps can help to cap the bill, but it's a short-term prescription for a case of token shock.
In a video call recently I spoke with Jim Webber, Chief Scientist at Neo4j and Visiting Professor at Newcastle University, about where the money goes when you swap poor context for good context, and what he's learned about the price of accuracy. He described enterprise conversations about AI spend as being in a moment of recalibration:
Everyone's rushed for AI, because it seems so appealing... And I think now that rush is past. We've all dived in, found what it seems to be good at, what it seems to be poor at, and now I think a more measured mindset is taking over....
Copyright of this story solely belongs to diginomica.com. To see the full text click HERE
