HPE proposes measuring AI coding spend by cost per merged PR
What happened
On 2 November 2026, MIT Technology Review reported that the industry is starting to question measuring AI coding spend by token consumption. An agent can use 10 to 50 times more tokens to complete a coding task than a normal chat prompt, the report said, so token-based billing makes the most valuable workflows the most expensive; Uber reportedly used up its entire 2026 AI budget in April. The report argues for counting delivered software instead of tokens. HPE had already suggested measuring AI coding spend by the cost of each merged PR. The idea of measuring by delivery has now been echoed in the media, but the report gives no specific figures or implementation results for HPE's approach.
Written by AI from the reports · updated 3 h ago
Timeline
Follow the reports to see the event from each side.
- MIT Technology Review · AIStop counting tokens, start counting what ships
When a developer gives an AI agent a coding task, the agent does more than answer once. It plans, generates code, runs tests, and repeats the process to correct errors. A single task can use 10 to 50 times more tokens than a basic chat prompt.
Heat of this event
Current heat 9·Peak in the comparable range 9（Nov 3 00:00）·Change over 24 hours in the comparable range –
The trend compares only the same accounts observed without a break, so its range can be narrower than the current heat count. Move the pointer or tap the chart to see each hour; with a keyboard, use the left and right arrow keys.