Apr 12, 2023 · 55m · green-blueprint
Live: How AI could supercharge climate tech
Stephen Lacey · 14m spoken Amy Francetic · 13m spoken Priya Donti · 11m spoken Jesse Morris · 9m spoken
0:00 / 0:00
gold bands on the timeline = statements, start to end. Hover to read, click to jump. CC turns on captions
Recorded live at Greentown Labs, host Stephen Lacey and an expert panel examine how artificial intelligence and digital technologies can realistically accelerate clean energy deployment while cutting through industry hype and overcoming analog grid bottlenecks.
How this conversation actually went
Every chapter scored 0–10 on four independent dynamics. Hover any point for the reasoning behind the score. How this is scored →
The hosts as informed peer 4.3 Guest teaching 3.2 Guest disagreement 0.7 The hosts pushing back 0.3
speaking balance: gold is the hosts, purple is the guest (3 minute bins)
Sharpest disagreement ▶ 26:02 Public interest reframe on AI commercialization
Priya politely counters Amy and Jesse's venture framing that solutions must be paid for by enterprises, arguing vital accountability tools like Climate Trace serve public good rather than commercial profit.
Hardest push from the hosts ▶ 14:22 Calling out AI pitch repackagingLacey directly presses Amy on whether climate tech founders are slapping AI onto old, rejected decks just to chase fresh venture capital.
Biggest teaching moment ▶ 35:45 Heuristic on generative AI limitationsPriya educates the panel on the fundamental probabilistic nature of LLMs, articulating that generative AI fails whenever tasks are difficult or costly to independently verify.
The host holds their own ▶ 7:00 Deconstructing AI gold rush vs infrastructure realityLacey demonstrates domain authority by steering the conversation away from superficial LLM enthusiasm toward the foundational digital layer required for grid decentralization.
the scores for every segment, with the reasoning behind each
| Chapter | Topic | The hosts as informed peer | Guest teaching | Guest disagreement | The hosts pushing back | Why |
|---|---|---|---|---|---|---|
| Live Show Opening & AI-Generated Introductions | 3 | 0 | 0 | 0 | Host Stephen Lacey opens the live show using live ChatGPT-generated prompts in different movie genres to introduce the panelists in a lighthearted, playful tone. | |
| Setting the Stage for AI in Climate Tech | 5 | 0 | 0 | 0 | Lacey provides the thematic setup, contrasting generative AI consumer hype with the practical digital layer needed for grid decarbonization. | |
| Machine Learning Evolution and Practical Climate Applications | 4 | 4 | 0 | 0 | Priya Donti outlines how data availability, computing power, and attribution science merged to enable machine learning use cases like solar satellite mapping and battery synthesis. | |
| Investing in Digital Climate Tech and AI Tools | 5 | 3 | 1 | 1 | Lacey asks about AI buzzword pitching, prompting Amy Francetic to explain Buoyant's focus on software and sensor solutions like drone inspection of solar assets rather than AI for its own sake. | |
| Digital Infrastructure and Lessons from Web3 Hype | 4 | 4 | 1 | 0 | Jesse Morris offers a candid reflection on the Web3 hype cycle, noting that only 3 of 200 energy blockchain startups survived, contrasting that bubble with AI's practical utility. | |
| Practical Enterprise AI vs. Public Interest Solutions | 5 | 5 | 2 | 1 | The panel discusses monetization and greenwashing; Priya introduces a nuanced reframe to Amy and Jesse's enterprise-first focus by highlighting critical public-interest, non-profit data tools like Climate Trace. | |
| Mid-Roll Sponsor Message: Nexamp | 5 | 5 | 1 | 1 | After the sponsor read, the conversation explores enterprise LLM applications; Priya provides a foundational framework that LLMs only work well when outputs are hard to create but easy to verify. | |
| AI Workforce Readiness and Grid Data Challenges | 4 | 5 | 1 | 0 | Amy and Priya discuss internal corporate upskilling and data engineering, while Jesse grounds the panel by highlighting that utilities lack even basic digital data on home DERs. | |
| Future Breakthroughs and Institutional Bottlenecks | 5 | 4 | 1 | 0 | Lacey guides the panel toward future projections and bottlenecks; Priya notes grid blackout fears restrict testing, Amy advocates for satellite verification of carbon offsets, and Jesse cites institutional inertia. | |
| Real or Robot Trivia Game | 3 | 2 | 0 | 0 | Lacey leads the guests through a trivia game generated by ChatGPT called 'Real or Robot,' testing whether AI climate concepts are genuine or hallucinations. |
Statements from this episode (15)
Lacey: AI rarely ranked top-five in climate tech before late 2022
“Up until about six months ago, if you'd asked people in the energy business or in climate tech about the hottest areas of tech innovation, artificial intelligence might have topped the list for some people, but probably wouldn't have gotten into the top five f…”
Donti: Attribution Science Galvanized Public Climate Conversation
“And I actually attribute a lot of that to
improvements in attribution science, right? When we were able to really start saying this specific event is due to climate change, I think that really kind of galvanized the public conversation.”
Francetic: Buoyant Ventures targets digital climate software over hard tech
“We're looking for digital technologies that we define as software, including AI, and simple hardware, so we're not investing in the hard tech like a lot of the great things that are happening here at Greentown Labs. We're investing in simple sensors that are c…”
Morris: Climate Action Will Fail Without Sector-Wide Digital Infrastructure
“From my opinion, if we don't, and the language I use about this, deploy digital infrastructure across all of these different sectors of the economy to help accelerate action against climate change, we will fail. We will not address climate change in a time fra…”
Donti: Open Climate Fix cut National Grid forecasting error by 33-50%
“For example, the non-profit Open Climate Fix, they used AI and machine learning in combination, you know, with satellite imagery, historical data, to cut the error of the forecast that National Grid, the UK power system operator, is creating by a third to a ha…”
Morris: Very few Web3 or crypto startups sold real enterprise solutions
“In my experience, there were very few web three or crypto companies that were selling a solution to enterprises to solve a specific problem.”
Francetic: Mandatory climate disclosure regulation is still coming to the US
“We knew these disclosures and this regulation was coming, because we could see what was happening in Europe, and we could see that this pressure that was being placed globally on big investors to try to disclose this was gaining momentum in the U.S., even thou…”
Morris: Building chatbot tool for electric utilities to configure DER software
“My, I was on the line with my CTO the other day, and we are now making a tool so that some of our customers, which are these electric utilities, Can use chatbots to configure software that lets them communicate with DER and procure services from them.”
Morris: National lab report shows two terawatts in interconnection backlogs
“There's a report out today it was one of the national labs, I believe, published it. Two terawatts of backlogs, solar, wind, hydro, batteries, In interconnection backlogs. Two terawatts. And I think that's just transmission level.”
Donti: GenAI works when generation is hard but verification is easy
“As maybe thinking back to this barometer to evaluate whether an application kind of makes sense, I think a lot of these make sense because they satisfy this key property, which is something that is hard to do, but easy to verify. So ChatGPT and other generativ…”
Morris: Edge assets will reach 30-50% of grid capacity within 10 years
“And if you look at all of those assets, so all of these small scale assets, Those are really going to be the future of how we balance the grid in the next 10 years. They're going to account for anywhere between, like, 30 and 50% of total system capacity.”
Morris: Utilities cannot collect data on customers' EVs and solar systems
“I live in California. I'm one of those customers. I have a electric vehicle. We have a heat pump. We have solar system. My utility has no idea I have those assets. They don't have the ability to collect data on them, let alone run some AI based algorithm on th…”
Donti: Power grids lack viable deployment pathways for critical AI applications
“There is something really critical holding that back, which is that there aren't really good pathways to deployment for a lot of techniques, AI machine learning or otherwise, that are actually interacting with critical operations and, you know, critical functi…”