Uber might look cheap here but budget discipline around AI is probably about to become normal
Unlimited token burn isn't a strategy and being disciplined about usage doesn't seem insane
Yup. Which is why you need to find an efficiency metric and measure on the unit cost. Once you have that, increase budget for top performers (or remove it altogether) and decrease budget for others.
Relevant posts:
- "Your Most Expensive Developer Might Be Your Most Efficient"
- "Token Budgeting: How To Think About AI Cost Control"
https://t.co/70loZrHJ09
https://t.co/UXnNzggzbG
@daiteskorosti Not sure I understand the message totally...but looking at a cost per PR, cost per bug fixed, cost per Linear issue closed or cost per feature shipped is infinitely better than just looking at token spend or consumption standalone.
*UBER SETS $1,500 MONTHLY CAP ON SOME AI CODING TOOLS FOR STAFF
$UBER officially reeling in the Claude budget after blowing their AI budget earlier this year.
Undoubtedly more companies to follow
3 months from now:
Excited to share we are septupling down on our stake in Anthropic which is raising 10 quadradrillion at a kazillion dollar valuation.
We first invested 13 days ago in their Series N. We're very excited to support them into their Series W raise today.