Patch

The AI and climate debate has shifted. Maybe not so much in the public consciousness (yet), but among the people closest to the levers of power at the intersection of AI and climate, it's in a completely different place. For the past two years, I've been in rooms where people argue about whether AI belongs in sustainability at all. At the end of SF Climate Week, no one's talking about whether, it's all about how. How do you govern AI adoption when your company is deploying faster than any framework can track? How do you right-size it when the energy tradeoffs are real? How do you maintain claim integrity when the data systems underneath you are changing week to week? These are harder questions. They're also the right ones. What I found consistent across every conversation this week (from the CSO summit at Rebalance to the workshops Patch ran midweek) was that the people furthest ahead aren't the ones with the most sophisticated AI stack. They're the ones who've figured out how to coordinate across their organizations. How to make hard calls together, in conversation, rather than siloed off inside their own systems. That's what the climate work requires. And it's what these rooms are for. Watch the video. It's mostly just people. That's the point.

I facilitated a workshop at Patch's Rebalance event on Monday with sustainability leaders from some of the largest companies in the world. Across panels and our workshop, one tension ran through the conversations: AI's promise versus its environmental cost. For sustainability teams, that trade-off is directly tied to our mandate. If the purpose of our work is to preserve resources and protect the environment, using resource-intensive tools to do this work creates a palpable tension. When we run AI models without context constraints, model efficiency decreases and resource intensity increases. So, for those of us working in climate, deploying AI without model constraints runs the risk of consuming more resources than the task will save. As we discussed this tension on Monday, an idea surfaced and echoed: right-sizing. Constrain your inputs to the right datasets, minimize the energy consumed per task, and you'll typically get a stronger result anyway, because the model isn't wading through noise and is focused on what's relevant. At Patch, we've been navigating this balance ourselves. We celebrate how many people on our team are experimenting with AI and finding new use cases, such as building agents that draw on our internal project diligence to surface the strongest matches across thousands of carbon credit options for a given set of buyer preferences. At a time when a lot of companies are still figuring out how to get their people to genuinely engage with AI, we're excited to have a culture of curiosity and experimentation. But, I'd be overstating if I said we've cracked the balance between encouraging experimentation and right-sizing each task. That's a muscle still being built, by us and across the industry. When the whole industry is asking "how much more can we do with AI?", the question I'm reflecting on is a bit different: how do we best right-size?

SFCW is here, and naturally the buzz from the mecca of AI, is the role of this technology - whether friend or foe - in advancing the enterprise climate strategies so many of us are driving forward. At the SF Climate Week Rebalance CSO Executive Summit, we cut straight to the bimodal challenges and opportunities AI presents in tackling our climate crisis. The dialogue was refreshingly specific, focusing on where AI provides immediate leverage and where caution is still needed: AI Strengths: ✅ Accelerating scenario planning ✅ High-level benchmarking ✅ Finding critical patterns in disparate emissions/climate data AI Roadblocks: ⚠️Policy synopsis and analysis of standards ⚠️ Identifying and building reliable data sources ⚠️ Crafting a resonant narrative (more on this later) It was striking to see the vibrant community of Chief Sustainability Officers doing what they do best - cutting through uncertainty to find a path forward despite high pressure, low resource environments. I am always thankful for the openness, humility, and courage climate leaders have - to share what we're experimenting with, how we've failed but quickly learned, and the ambition we're not losing sight of. Grateful to be part of this community and keen to keep the conversation going. Immense thanks to Salesforce & Patch for hosting this important dialogue, and the leaders who shared their hard won insights. Sunya Norman Brennan Spellacy Erik Hansen Joe Speicher Erica Fensom Nat Bullard Jessica Hyman Blair Swedeen Ellen Jackowski Brendan Molony #Sustainability #AI #ClimateAction #SFCW2026

Yesterday, I sat down with Chief Sustainability Officers at the San Francisco Climate Week Rebalance Summit to discuss AI & sustainability. As we spoke, one thing became clear: we are living in tension....it is hard AND sustainability professionals were born for this moment. The tension is between: 1) The enormous potential of AI to advance environmental / social progress as well as scientific discovery and innovation 2) AI's risks, impacts and unintended consequences. But don't forget that our field has been grappling with tension since the very beginning. The tension between what consumers are willing to pay for goods and what it actually costs to manufacture them with integrity. The tension between long-term and short-term growth. The tension between optimizing for the few versus the many. So many of us are already helping our companies navigate this new era and leveraging AI to do so. I was so inspired by the examples the speakers shared as well as their sense of humor and honesty. We're in this together. I’m grateful to leaders from across tech, finance, and sustainability for sharing guidance, solidarity, and inspiration. Thanks to everyone who joined the discussion, and to Patch for co-hosting with Salesforce. Happy Climate Week, everyone. I'm looking forward to seeing you out there! #ClimateWeekSF Special thanks to Bianca Williams-Torres for planning such an incredible gathering!

One number from Rebalance stuck with me: OpenRouter is processing 25 trillion tokens a week. A year ago it was 12 billion. The rate of growth is eye-watering and it's not being solely driven by humans typing queries, it's agents doing work. And it doesn't seem like we'll be slowing down anytime soon. Nat Bullard put it well yesterday: we're ten minutes into the first 24 hours of this paradigm shift. Yesterday, we brought together 40 enterprise CSOs to kick off SF Climate week to explore how the word's leading companies are approaching AI in the sustainability function. What I heard from them is that the hardest part isn't the energy math. It's coordination. Every business unit is writing its own AI playbook. CIOs are onboarding new models, platforms and agents each a week. Governance structures that existed a year ago are being reevaluated. And the sustainability team sitting across all of it is usually a handful of people. We see this at Patch every day. The companies making real progress aren't the ones with the most ambitious targets. They're the ones who've figured out how to align their ambition with the organizational and technological changes their companies are undergoing. Thanks to Sunya Norman, Erik Hansen, Erica Fensom, Joe Speicher, Blair Swedeen, Jessica Hyman, and Ellen Jackowski for a sharp morning of conversation.

I've read a lot of carbon strategies that never made it off the page. Not because the strategy was wrong. Because by the time it was written, the team had already moved on - reacting to a supplier pullout, an internal question from finance, a headline about a major buyer pausing purchases. Carbon markets don't move on a quarterly timeline. The decisions that matter happen faster than that. That's what I'm taking into my new role as Embedded Climate Strategist at Patch. It's not a consultancy model. It's not external advice on request. It's a climate expert working inside your sustainability program — present when it counts, not just available when you reach out. The carbon markets are more complex than they were two years ago. More scrutiny, more volatility, more noise. The companies managing it well aren't the ones with the best-written strategy. They're the ones with the most informed people making calls at the right moment. That's what ECS is built to do. Link in the first comment. #EmbeddedClimateStrategy #CarbonMarkets #Sustainability