Storia in 2 fonti

Microsoft Tells Engineers ‘Tokenmaxxing Is Not What We Are Optimizing For’

Microsoft is introducing budget limits for AI use but says it still wants to be an ‘AI-first’ company.

Raccontata da

404media.co

thenextweb.com

Confronto fonti

2 prospettive sulla stessa storia

AI · summaries

404media.coStai leggendo11 h fa

Microsoft Tells Engineers ‘Tokenmaxxing Is Not What We Are Optimizing For’

Microsoft caps AI spending, sets per-team budgets, defaults to GPT-5.6; engineers spend hundreds to thousands monthly. Signal: ROI on AI isn't automatic. Governance now critical—shifts focus from token maximization to outcome-per-token for budget and stack.

originale

thenextweb.com9 h fa

Microsoft tells employees to stop tokenmaxxing, sets division-level AI budgets

Microsoft set division-level AI budgets, switched to cheaper defaults, and capped engineer spending. Industry shift: despite 98% price cuts, enterprise bills tripled as agentic tools demand far more tokens per task—forcing budget discipline at Meta, Amazon, Walmart.

Leggi questa versione →

Timeline cronologica

martedì 4 agosto 2026·404media.co
Microsoft Tells Engineers ‘Tokenmaxxing Is Not What We Are Optimizing For’
Microsoft is introducing budget limits for AI use but says it still wants to be an ‘AI-first’ company.
martedì 4 agosto 2026·thenextweb.com
Microsoft tells employees to stop tokenmaxxing, sets division-level AI budgets
Microsoft EVP Jay Parikh told staff to curb AI token use, switched the default model to a cheaper one, and gave each division a formal AI spending cap