Your secret weapon for climate action isn't human (and that's precisely the point)
UnsplashUnsplashUnsplash
· 10 min read
This article is the third part of a three-piece series. Here is part 2
Article 1: Why climate messaging fails. I spent five years learning how my friend Kim's neurodivergent brain works differently. It showed me where ALL our communications fail.
Article 2: Specific structural recommendations for talking TO people instead of AT them.
This article: How to use AI to test whether you're talking AT or TO people. BEFORE you hit publish.
I watch businesses hollow out their futures by substituting AI for real humans at entry-level positions.
Smart companies are eliminating the very roles that build institutional knowledge and develop judgment.
What I see is short-term thinking wrapped in the language of efficiency.
Cost savings today. Organizational brittleness tomorrow. That's the pattern I've watched play out repeatedly.
For three decades, I've signed documents that implemented these trade-offs. I know the rationale. I know the outcomes.
However, here's what I learned from my friend Kim, who taught me the difference between talking AT people and talking TO them:
AI's highest use isn't replacing humans. It's helping us figure out whether our human messages actually reach other humans.
In my last two articles, I shared how it took me five years to realize I'd been talking AT my friend Kim instead of TO her.
Once I understood how her neurodivergent brain processes information differently, she gave me specific structural recommendations: clear frameworks, generous white space, emotional connection, and interactive elements.
Then I proposed an experiment:
Take content you're proud of. Feed it to AI with this prompt: "Take this piece and tweak it so it's engaging for a neurodivergent person with ADHD and dyslexia. Then summarize where the differences are and why."
I ran this experiment myself before suggesting it to anyone.
The results stopped me cold.
The AI didn't dumb down my content. It elevated it.
Where I had buried the lead three paragraphs deep -- talking AT people and making them hunt for my point -- I pulled it to the surface.
Where I had used dense paragraphs that assumed shared context, again talking AT people who already thought like me, AI built explicit frameworks.
The AI version wasn't "simplified." It was clarified. Strengthened. More direct.
And here's what hit hardest.
The changes AI suggested weren't just better for neurodivergent readers. They showed me everywhere I was talking AT people instead of TO them.
My neurotypical brain had filled in gaps I didn't know I'd left.
I'd assumed a structure that wasn't there. I'd followed patterns so familiar to me that I couldn't see where they failed others.
AI, with no neurotypical assumptions, showed me exactly where my own communications were optimized for people like me and failing everyone else.
Here's what I think might be AI's superpower in climate communications.
It has no bias toward any particular way of processing information.
It doesn't assume readers will follow non-linear narratives.
It doesn't expect them to hold multiple concepts in short-term memory.
It doesn't rely on shared context.
AI examines information architecture neutrally and asks:
Does this structure actually communicate TO the reader? Or is it just broadcasting AT them?
Think about what this means for the climate movement!
We've spent decades crafting messages that make perfect sense to us. We've assumed that if the data is solid and the arguments are logical, people will respond.
They haven't. Not at the scale or speed we need.
Not because they don't care.
Because we've been talking AT them, using that pre-1980 playbook, instead of talking TO them in ways that actually work with how their brains function.
When AI restructures content for neurodivergent accessibility, it's showing us everywhere we were talking AT people:
• Buried leads (making them hunt for your point)
• Dense paragraphs without visual breaks (assuming their brains work like yours)
• Missing frameworks (expecting them to build structure from chaos)
• Assumed context (talking to people who already agree with you)
• Abstract before concrete (starting where you are, not where they are)
Every single one of these is a symptom of talking AT rather than talking TO.
AI can see this because it's not swimming in our assumptions.
It doesn't know what "everyone knows." It doesn't assume shared context. It simply examines whether the structure effectively serves the reader.
This is incredibly valuable.
Because we can't see our own blind spots. We need something outside our assumptions to reveal where we're falling short.
I'm not suggesting we hand climate communications over to AI.
That would be a disaster.
AI can't bring lived experience. It can't share an authentic struggle. It can't build trust through vulnerability. It can't offer the hard-won wisdom that comes from making mistakes and learning from them.
Those are human contributions. Essential human contributions.
The only way to communicate with people is on a human-to-human level.
What AI can do is help us test whether we're actually talking TO people or just talking AT them in new formats.
Feed it your content. Ask where you're talking AT people, not TO them. Study the gaps.
Let it show you where your information architecture assumes too much. Learn from those patterns.
Use it to create multiple versions optimized for different processing styles. Let humans choose what works for them.
Experiment with different approaches quickly. See what actually communicates before deploying.
• The source of authentic voice
• The creator of strategy
• The replacement for human judgment
AI should test whether we're talking TO people.
It must never do the talking for us.
Here's the pattern I keep seeing:
We build systems optimized for narrow parameters. Those systems work for some people. They fail everyone else.
Decades later, we discover the unintended consequences.
The pre-1980 marketing playbook worked for talking AT customers about products.
It fails catastrophically when discussing planetary survival with diverse populations.
The ability to test whether we're talking TO people or AT them before we deploy at scale.
We don't have to guess whether a message will land with neurodivergent audiences. We can check.
We don't have to assume our structure works for everyone. We can verify.
For the first time, we can catch ourselves talking AT people before it becomes embedded in our systems.
I see too many businesses using AI to eliminate humans.
Chatbots are replacing conversation. Automated systems are replacing judgment.
This is the old playbook applied to new technology. Talk AT people more efficiently.
What if we flipped it?
What if we used AI to test whether we're talking TO people effectively?
What if we deployed AI to make human expertise more accessible rather than replacing human connection?
The sustainability movement has an opportunity here.
We can model a different relationship with AI. One that serves human communication rather than substituting for it.
Here's what this looks like in practice:
You write from your experience, your expertise, your authentic voice.
That's the human contribution. The only way to talk TO people is through a genuine human connection.
Then you use AI to test:
Am I actually talking TO people? Or am I talking AT them? Where am I assuming shared context? Where am I burying my lead? Where am I making people work too hard to understand?
You revise based on those insights.
Not letting AI write for you, but learning from AI's neutral analysis of where you're talking AT instead of TO.
You publish in multiple formats.
Detailed for those who want depth, visual for those who need frameworks, and audio for those who process through listening.
AI handles the translation between formats. You maintain the authentic voice.
Here's what I've seen result from this approach.
Your human expertise reaches more humans more effectively. Because you're talking TO them, not AT them.
Climate change demands that we talk TO everyone.
Scientists and farmers. Policymakers and business leaders. Activists and skeptics.
We need communication that works for both neurotypical and neurodivergent minds.
For visual learners and audio processors. For people who want data and people who need stories.
We can't afford to optimize for talking AT one group and hope others adapt.
The stakes are too high. The timeline is too compressed.
AI gives us the capacity to test whether we're talking TO people before we deploy.
To catch our AT-ness before it embeds. To translate human expertise into formats that actually reach everyone.
This isn't about replacing human communication.
It's about making sure human messages actually communicate. Effectively.
I keep saying: We're living with the unintended consequences of actions we collectively took over the last century.
Had we known the destructiveness, we'd ask for a do-over.
But do-overs are a fantasy.
Except here's where I might be wrong.
AI gives us something close to a do-over for communication systems.
We can test whether we're talking TO people or AT them.
We can design for actual human diversity rather than assuming uniformity. We can catch failures before deployment, rather than discovering them decades later.
We can do better.
Not by abandoning human connection. By using technology to test whether our human messages actually connect.
Run the experiment.
Take your proudest recent content. Feed it to AI with that accessibility prompt. Study where it shows you talking AT people instead of TO them.
Not to let AI rewrite your work. To learn where your communication architecture assumes people think like you.
Then make your own revisions. Guided by your human judgment. Enhanced by technology's neutral analysis and authored by you.
Share multiple formats. Let people engage in whatever way works for their brains.
Watch what happens when you shift from talking AT to talking TO.
AI, used thoughtfully, helps us test whether we're communicating in the highest good for all.
Or just for people whose brains work like ours.
It shows us where we're talking AT people instead of TO them.
It tests accessibility before deployment. It translates between formats, allowing everyone to engage.
This is technology in the service of actual communication.
Not broadcasting more efficiently. Connecting more effectively.
The climate fight needs every tool at our disposal. AI is one of them.
Let's use it to ensure we're talking TO everyone who needs to hear us.
Because the world we're building has to work for all of us.
And communications that talk AT some of us won't get us there.
By the way, all three articles in this series were optimized for the neurodivergent brain. Did you notice?
illuminem Voices is a democratic space presenting the thoughts and opinions of leading Sustainability & Energy writers, their opinions do not necessarily represent those of illuminem.
Sustainability needs facts, not just promises. illuminem’s Data Hub™ gives you transparent emissions data, corporate climate targets, and performance benchmarks for thousands of companies worldwide.
Alex Hong
Cities · Agriculture
illuminem briefings
Sustainable Lifestyle · Sustainable Living
illuminem briefings
Cities · Sustainable Living
Deutsche Welle
Climate Change · Architecture
Grist
Climate Change · Adaptation
CNN
Cities · Sustainable Living