The next step in AI cost saving: treating frontier models like expensive consultants
Letting frontier models do the complex thinking and planning, and none of the execution, is the key to saving on your AI budgets.
AI consultants are advocating for a new approach to managing the cost of frontier models, likening them to expensive consultants. The key to saving on AI spending, according to these experts, is to treat these powerful models as advisors or consultants, delegating complex thinking and planning while leaving execution to cheaper alternatives.
Ameya Kanitkar, CTO of San Francisco-based AI measurement platform Larridin, suggests that expensive models like Anthropic's Fable 5 should be used to create roadmaps and plan workflows, while cheaper models should execute these plans. Michael Murphy, a partner at Sydney-based AI transformation consultancy Adaptovate, echoes this sentiment, stating that powerful models should be reserved for strategic tasks and initial app or website builds, while lightweight models should handle day-to-day tasks such as transcribing meetings or creating creative briefs.
This approach aligns with the growing concern among executives about achieving proportional returns on AI spending. Companies are moving away from token-maxxing, a trend of unrestricted AI experimentation, and instead focusing on cost-saving tactics like model routing and open-source models. AI Routing startups, which help developers select and monitor AI models, are gaining traction as a C-suite priority.
Written by urgent.news from Business Insider's reporting — not their text. Machine-written — may contain errors; check the original before relying on it.
