AI/Tech · Current
Flowlink
Estimates of energy, carbon, water, and electricity costs for AI workloads.
Overview
Flowlink brings resource use and electricity costs into one estimate so teams can compare AI workload choices. Its outputs are modeled estimates.
My role
I co-founded Flowlink. I lead market research and pitches, review the model's assumptions, and co-write explanations of how the estimates are calculated.
What I worked on
- I started exploring the idea in Grade 10 and worked with my co-founder to develop the estimator.
- I talk with potential users about what they need to compare and how they interpret the estimates.
- I co-wrote a technical white paper and documented four corrections to the model.
Status
Validation and pilot outreach are in progress. We are seeking benchmark runs and independent GPU readings to compare against the model's estimates.
Links
Related writing
- AI Cost Forecasts Need Prices and Discount Rates That Match
The European Business Review