HPE argues enterprise AI costs should be measured per useful outcome, not per token
Overview
HPE argues that enterprise AI economics depend on effective cost per useful outcome, such as cost per merged pull request, rather than per-token pricing.
Per-token pricing makes agentic coding costs rise with usage, according to a MIT Technology Review article that reports HPE's position.
The article recommends routing routine coding work to open-weight models running on owned capacity and reserving frontier models for difficult tasks. It cites HPE's reported 30 to 50% reduction in token costs from running routine workloads on premises.
Written by AI from the articles below · updated Nov 2, 9:00 AM ET
Check the sources:
Developments
2 developments
- Nov 2, 8:10 AM ET · 1 articleStop counting tokens, start counting what shipsMIT Technology Review · AI: Stop counting tokens, start counting what ships
- Oct 20, 11:20 AM ET · 1 articleHPE argues enterprise AI economics hinge on effective cost per useful outcomeMIT Technology Review · AI: HPE argues enterprise AI economics hinge on effective cost per useful outcome
Article timeline
The articles in this story. Times are ET.
- MIT Technology Review · AINewsStop counting tokens, start counting what ships
AIPer-token pricing makes agentic coding costs rise with usage, which HPE says can be offset by tracking cost per merged pull request. The article recommends routing routine coding work to open-weight models on owned capacity and reserving frontier models for difficult tasks. It cites HPE's reported 30 to 50% token cost reduction from running routine workloads on premises.
- MIT Technology Review · AINewsHPE argues enterprise AI economics hinge on effective cost per useful outcome
AIHPE says enterprises should measure AI economics by effective cost per useful outcome, not token or GPU-hour price. The sponsored piece says an integrated private AI platform, HPE Private Cloud AI, can beat repeated rental of equivalent capacity on cost per token in HPE-commissioned testing at high utilization. The article notes results will vary by workload mix and utilization.
Heat Trend
- Current comparable heat
- 9
- Peak comparable heat
- 10Nov 2, 9:00 AM ET
- Change in past 24 hours
- –
The trend compares only continuously observed participants, so its scope may be smaller than the current heat total. Hover or tap for hourly heat; use arrow keys with a keyboard.