FAQ
Questions, answered honestly.
Launch timing, devices, data, and cost, including the questions where the honest answer is still “we don't know yet.”
For households
Which devices and platforms will GreenAI support?
Compatibility is not finalized. GreenAI is being designed for flexible loads such as EV chargers, HVAC systems, water heaters, dishwashers and washers, with home-automation platforms such as SmartThings providing the connection across compatible brands. We have not published or field-tested a production compatibility list yet.
Will GreenAI change my routine or run devices when I need them?
The intended product works around limits you set, such as an EV's ready-by time or an HVAC comfort target, and shifts only the flexible part of a task. We have not tested GreenAI in occupied homes, so we don't know yet how often people would notice or override those shifts.
When does GreenAI launch?
GreenAI will launch in late September.
What will GreenAI cost?
Pricing will be determined in the future, but GreenAI is expected to have a free tier.
How does GreenAI handle my data?
GreenAI does not process household device data because no live product exists today. The mailing-list form stores the email address and optional context you submit in Google Forms, while prototype state stays in your browser. The site runs no third-party analytics or advertising cookies. The privacy policy will be updated before GreenAI reads data from a home.
For researchers
How has the scheduling model been validated?
The current calculation applies scheduling rules to simulated NREL ResStock household loads and published grid-carbon data, then compares baseline and shifted emissions. That tests the calculation against historical inputs; it does not tell us yet how accurately the model predicts results in occupied homes.
What do the 35% and $350 figures actually represent?
The 35% household carbon reduction is a modeled upper bound for shifting flexible loads in a simulated household. The roughly $350 per year figure is a modeled average bill-savings estimate for New Jersey. They are not guarantees, they do not come from the same regional model, and no GreenAI field deployment has measured either result.
Should I expect the same result outside California or New Jersey?
No. The current carbon illustration uses California grid data, while the bill-savings estimate is for New Jersey. Results should vary with the local generation mix, electricity rates, weather, home type and flexible devices. Region-specific modeling is required before making a claim for another place.
What does the current model leave out?
The current estimate uses one simulated household archetype and climate zone, historical snapshots rather than live or year-round data, and one fixed set of device deadlines and comfort assumptions. It does not yet include measured occupant behavior, real installation constraints or field performance. We don't know yet how much those omissions will change the result.