AI ROI: Frenzy & Strategy. The Return On/To Intelligence?
- Borrow2Share

- Jun 1
- 1 min read
The Economics Of AI. The Benefits/Costs Of AI Magic. The Compute Capacity Race. General Fast AI For The Rest Of Us; Advanced Premium AI For The Select. May Make Better ROI Common Sense. Because It Becomes Unaffordable For Everyone Just To Have/Use ‘Supercomputers.’
(A)rtifficial Intelligence Do With Immense Costs. It’s Too Compute-Intensive Intelligence. Much More Energy-Intensive Than Our Own Real AI (A)ctual Intelligence ;). A Calculator Computes Numbers. An AI Model ‘Computes’ Inferences Word Tokens. Giving Rise To The Awe-Amazing Emergent Intelligence Abilities! Can Make You A Superman.
It’s A 10X Productivity Tool. And It's A 10X Ability Enhancement/Expansion Tool.
But, An Uber Case Study. 5,000 Engineers. Burned $3.4B AI Year Budget In 4 Months. The Output Value Gained From Burning So Many AI Tokens They Say Is Not So Clear?
Best To Use AI With ROI In Mind & On Your Side As You Go! Calculating & Justifying The AI Use.
So, Some Of The Ways To Measure AI ROI:
(1) Hrs Saved / Compute Cost. - Does Low-Value Work For You. The Heavy-Lifting. Work You Can Do; AI Does It.
(2) Ability(Value) Harnessed / Compute Cost. - Value Created. Does The High-Value Work For you. Work You Can’t Do; AI Does It. But Within The Abilities Limits Of AI.
(3) AI Agent Hrs / Compute Cost. - Autonomous Agents Doing The Mundane Work For You, i.e. Customer Service.
(4) Revenue / Compute Cost. - Actual AI Revenue vs. AI Compute Costs.
Shssh! The Return On Building AI Infrastructure Is Not So Good. But, The Return On Being In The Software/Apps Riding On That AI Infrastructure Is So Much Better!








Comments