Last week, I did a Button watch party about sustainable content design. In the course of the Q&A, Kristina Halvorson asked about AI: "But what if I'm planning a trip and I have ChatGPT pick my hotel, flight, and the top five things for me to do. Is that one search better than me looking at 100 different pages over the span of weeks?"

My in-the-moment answer was "I don't know," because if we're talking direct metrics it really depends on the size of the pages, if you're watching videos, or how you go about it. Honestly, there are a lot of variables at play, so I can't necessarily do a 1:1 emissions analysis on the fly in a webinar Q&A.

But having had time to think about it, I have three answers.

1. Oversimplification. ChatGPT is rarely a single search. Sure, it will give you an answer, but then you want it to refine the answer, or answer peripheral questions. So while it's quick, it's probably more than just one search. I'm guessing that a trip plan would take five or 10 searches or more to accomplish.
2. Reliance. If you start to rely on ChatGPT for "just this one thing," it's a slippery slope to using it for all the things. I mean, honestly, I've seen this with Google over the years: if I know I can find the information pretty quickly online, there's less reason for me to make an effort to remember it. If ChatGPT will do the thing for me, why would I bother to do it — or other related things — myself? After all, my time is valuable. It outsources my thinking.
3. Nuance and serendipity. If I'm not doing the work, I'm losing the nuance, the joy of serendipity. I've seen people using "summarize this article" to get the broad strokes. And yes, that will give a high-level overview. But that article-writer didn't write a few thousand words for a summary that could fit in 140 characters. Summaries lack depth and detail.
   ​
   Same with the travel example. If I ask ChatGPT to plan a trip to Philadelphia, for example, it's going to tell me to stay near Reading Terminal Market, visit the Liberty Bell, or eat a cheesesteak at Pat's or Geno's, or visit the excellent Philadelphia Museum of Art. And that's all valid. But it's also just a superficial look at the city. It's not going to give you insights into the best coffee experience, the best pizza, the best cheesesteak (spoiler: it's neither Pat's or Geno's), the lesser-known museums, the fascinating speakeasies, or the obscure corners of the city.
   ​
   That's why I want to do the research myself: to gain the nuance to understand and weigh my options so I don't just exit the plane and hit the tourist traps.

And I think that the last item, the nuance, is really the heart of the AI conundrum (emissions impacts notwithstanding). How much do we want to know — really know — about a thing? Do we want a surface analysis, or do we want to learn as we go?

In the process of writing my book, I did so much research. I learned things and made connections that I would never have made if I'd asked a service to summarize it for me. I can see connections and interdependencies that a summary would never reveal. And as we rely more and more on headlines vs depth, or social media vs long-form content, or summaries vs understanding, we lose something valuable. We lose our curiosity.

Al Gore, the former U.S. vice president, gave ideas for how to get the climate back on track. And yes, he recommends the use of AI. But he uses it as a modeling tool, not as a content generator.

​South Africa is struggling with a just transition away from coal-fueled power plants. "Just" in this context means "morally right and fair," and not "barely/only." A just transition protects workers' rights and livelihoods as the marketplace shifts from an extractive economy to a sustainable one.

With the U.S. withdrawing from international climate funding (again), bajillionaire and former New York mayor Michael Bloomberg is stepping in to cover the economic gap. Bloomberg also serves as a U.N. special envoy on climate change.

Female-led companies outperform their male counterparts. So why are those companies so underfunded compared to their male-led counterparts? "Of the $33.5 billion invested in climate tech startups in the United States through three quarters in 2024, only $135.8 million went to female-founded startups."​

​

🗒️

Consulting

​Reduce your Scope 3 emissions​

Alisa Bonsignore

Founder, Strategist, and Author