Organizing Center by TectonicaBook a demo
Chapter 14 of 30
Other chapters

Act III — The AI Question

There is much excitement about generative AI, and innovators are hunting left and right for the good applications of it. It is impressive, and every week seems to bring another installment of the science fiction we spent the last century dreaming about becoming reality. A good deal of that is hype, which makes the real thing harder to find rather than easier.

My excitement hasn't been found in just the general potential of generative AI so much as the unique potential role I see for the systems issues I've struggled with. For years I have been stuck on a problem in my own field with no way out from the inside. Building power takes people. Giving people real work takes the capacity to develop and support them. Building that capacity takes the time and resources you do not yet have. What results is a failure to scale beyond the usual roomful of people planning on post-it notes, the failure of our progressive causes, and the decline of democracy. This pattern has been a part of my work and a large part of my writing

Our political knowledge is centralized inside the heads of too few. The organizer's judgment. The strategy you rent from consultants. The unspoken norms of the organization that is lost when someone leaves their position. The working know-how of how a relationship can become a strategic asset. Scarce, expensive, and the reason the people closest to this work are often not equipped as they should be.

With this new technology knowledge is already starting to work differently. (Not really faster as some assume.) Before it took specific years of training to learn best practices in fundraising. Now a project can house the expert information and help build a plan with you in real time - combining best practices with community-level wisdom. Brand guidelines can automate copywriting. I think that this difference opens a systemic leverage point, not just about what we produce, but who has the power and ability to do it. And when it comes to organizing, I don’t think we’re talking enough about it. 

You should be skeptical of generative AI. I am. The costs are real on a variety of levels. What does it do to democracy? The speed at which it accelerates power differentials that were already indefensible. Of course, the environmental destruction behind how many Big Tech AI companies operate. Privacy questions. It’s application in war. And quietly, from a cultural perspective, the unknowns about what it might do to our collective political imagination with LLMs using predictive people pleasing to decide what answers it shows you.

In relation to my points above, generic AI, pointed at this work and powered by market logic rather than intention, will do what the systems around us already do, faster. Replace people. Replace relationships. Make the transaction cheaper.

It does not have to go that way. This is not just a thesis. It's also what my team has been working on building.

Chapter 14

Like it or not, you are already using AI

August 31, 2026
by Ned Howey

I keep seeing posts like this one. Somebody announces something like a doubling of their following on Instagram, or Facebook, or LinkedIn. Without using AI.

If you think you did that without AI, friend, I have some pretty bad news for you about how social media works and what it is built on.

You did not double a following without using AI. Something decided which of your posts got pushed out to strangers and which one died quietly in the feeds of people who already followed you. Something decided that the third video deserved a second life on day four. Something worked out which strangers resembled your existing audience closely enough to be worth the impression. That something was a model, and building better ones is the entire business those companies are in. You learned how to make content that drives followers within what the models prioritize. And you made content in your voice, with your goals and your story. That is admirable, strategic, and when it’s done best very authentic as well. But it was not without AI.

Maybe the content was made without AI. Yet your phone did not record that photograph so much as compute it. It stacked several exposures, decided which region was a face and put the focus there, estimated depth in order to blur a background that it deemed was less relevant, and made a few hundred decisions about noise and color as you clicked the shutter. If you edited it, you used tools that have been shipping trained models for years under the word smart, because AI was not yet a selling point to the public. One-click subject selection. Background removal. Sky replacement. Spot healing that invents the pixels it needs. Denoise. Upscaling. The photo library that let you type protest and find the right picture is running object recognition across every other photo in your library.

What about the words? Autocorrect. Predictive text. The grammar suggestion you accepted without thinking about it. If anything was translated, Google Translate has been running on neural networks since 2016, which is nearly ten years before the advent of AI as a buzz word. The automatic captions on the video are speech recognition. The alt text the platform wrote for you is object recognition. The email announcing the campaign or platform got past a spam filter that has been doing statistical classification since the early 2000s.

None of that is generative AI (rather predictive AI, machine learning, etc). That is the thing worth stopping on, because this category has been under the floorboards of everything we use for well over a decade. What arrived in the last few years is an evolution of all of this technology. It’s shiny because it writes and draws and does other things. This is a space that has jumped to immediate commercial value to the public. And that is just the direct use. Forget even looking down the supply chain.

The lunch that you ordered while you were working was routed by an optimizer. The driver's shift was assigned by one. The restaurant's stock order was forecast by one. The tomatoes were graded by machine vision somewhere along the way, the payment was fraud-checked by a model in about forty milliseconds. And as you sit down to work through your lunch break the electricity running your laptop is balanced across a grid by systems nobody would describe as human judgment.

None of that was made without AI. So when we say something was “without AI” we really mean that no large language model created the image or wrote the captions. Which in today's world is actually worth stating and valuing. The people making these posts are asking a real question about what this technology is doing to creative work, to the environment, to people whose labor trained it, and to the value of human creation. Yet people have been scandalized about technology whenever new technology happens. TV in the 50’s. The printing press in the 1500’s. Every technology advance has had its skeptics. And we know that on all sides of politics, abstinence and purity, as a political stance, can be more dangerous than the thing they are protesting.

Abstinence from AI by progressive organizations is about as likely to achieve our goals as recycling our way to carbon neutrality. Sorting plastic is admirable, and I do it. However, if anything, it gives people the sense that choices with marginal impact can produce the change we need in the world as it actually is, instead of sending them after the system underneath all of it. Thirty years of that produced excellent municipal sorting and a hotter planet.

If you run an organization and you believe your staff are not using AI, I have a second piece of news. They are. Right now. Tonight, somebody on your team is behind on a deadline for a key project, and your strategy document is going into a chatbot that your organization has no relationship with, no agreement with, and no visibility into. Almost certainly one of the big commercial products. ChatGPT, or Grok, or whatever arrived bundled inside a tool they already had. These consumer terms were not decided by your organization and your data might be in the hands of companies whose politics your organization would not choose. And because its use is covert and in secret the output produced is generic. Your staff is not doing this because they are reckless. They have too much to do, no help, and a tool that gives them a competent collaborator at eleven at night when nobody else is awake. Sometimes the work comes out better. And they will not openly share why.

A policy that says we do not use AI here does not stop any of that. It just relocates it, off your systems and out of your sight, and without any safeguards in place.


Everybody reaches for the Terminator when they talk about this rising age of AI, and they always reach for the wrong film.

The one worth thinking about is Terminator 2. Remember this scene? Sarah Connor has been locked in a psychiatric hospital for saying, accurately, that the world is going to end. She breaks out. She is running for the exit when a figure steps into the corridor ahead of her, and it is the T-800, the old model. The same model that hunted her seven years earlier and killed everyone she loved.

She turns and runs the other way, just as the newer model sent by the machines closes in from the other side.

She is running from the thing that has come to protect her. Her son is with it, screaming at her that it is on their side. When she finally stops, the machine holds out its hand and says the line: Come with me if you want to live.

Now imagine how the rest of that story goes if she folds her arms in the corridor and says, sorry, we have a policy against AI.

The whole plot of that film is a woman, a boy and a machine working together to stop the rise of the machines. The moral was never that AI is bad. Everything depends on which machine is in front of you, on who sent it, and on what you are aiming to achieve.

A movie is not an argument, I know. We have been quoting this one at each other for decades, and there is a more complex lesson about technology and about us humans.


The fights are real and they are coming. Data centers are destroying water systems. The industry is consolidating extraordinary power in a very small number of hands, at a speed no regulator has matched. The labor question is not abstract to anybody who has watched a client cut a design or copywriting team this year. The bias that comes out of these systems, which lands, as it always does, on the people who were already carrying the most. Our sector is going to be on the front line of every one of those fights.

Plenty of us are working on the alternatives. Ethical development practices, real guardrails, honest measurement and control of environmental impact. There are more people doing that than you would guess from the discourse. People and organizations that have earned my trust and taught me ways that this technology can be a force for good. That you can utilize a technology while also fighting against the parts of it that stand between us and a better world. This is not hypocritical.

The whole social and political sector could adopt a no-AI policy tomorrow and it would not stop one data center from being built. We are not going to change the labor practices of a hundred-billion-dollar industry through the purchasing decisions of the nonprofit sector, which is, and I say this with love, not a market anybody is optimizing for. Refusing to touch it removes us from operating at the same pace as everyone else.

Meanwhile the people who want to consolidate power are not abstaining. They are counting their cash and setting up PACs to use AI in whatever way they can to trick voters in every upcoming election.


Here is the reframe I would offer.

Nobody in our position needs a no-AI policy. What we need is an answer to a harder question: how do we use this to curb the harms of it, and to fight the people holding the reins from choosing for us?

This is not only a more exciting question, but also one that gets us closer to the work we do as progressives. Some examples include:

  • Building and funding alternatives, so that the ethical options exist

  • Making the companies accountable for what they are producing rather than for what they are promising

  • Being in the room where the standards get written

  • Considering the potential unintended consequences to our democracy and our movements when we do adopt these tools

  • Considering how this can be used to rebalance power, putting the capability in the hands of the people it has always been kept from, instead of ceding it to the people who are already ahead.

Abstinence is not a position, because it was never on offer. You have been using it all along, in the ad platform and the spellcheck and the delivery route and the grid. The only thing actually available is the choice on how you use it. Being purposeful with all the concerns you may have taken into account.

I am not neutral here. Nor would I claim to be. My team is building a platform, based on these new opportunities provided by generative AI.

I did not get here by looking for something to build. I got here by spending fourteen years watching a cycle nobody could break from the inside, and then noticing that the thing holding it shut finally has a place to move.

There is no altitude at which you get to stop being part of how the world runs.There is only the possibility to work with the world the way it is and to be clever about how to upend it if we seek it to be different.

This is a reflection on how he got here, and why we need innovations built for people power. Ned Howey has spent fourteen years at Tectonica building organizing infrastructure for progressive movements.

Start from the beginning