UnTangled with Gozi Egbuonu – Agency in the Age of AI (Season 2 Finale!)

We wrapped up this season of UnTangled on a topic that has been quietly sitting underneath almost every conversation we have had this year. Agency. What does it mean to maintain your own autonomy, self-determination, and critical thinking in a world where AI is increasingly present in every tool you use, every decision you make, and every workflow you have built?

For this final episode of the spring season, I was joined by Gozi Egbuonu, Director of Programs at the Technology Association of Grantmakers (TAG) and self-described Troublemaker, which is frankly a job title I deeply respect. Gozi is also a doctoral student in behavioral health and an AI strategist who co-created a responsible AI framework for philanthropy. She thinks carefully and critically about this space, and this conversation did not disappoint.

The Word “Agency” Has Been Hijacked

Here is something worth sitting with. The same companies that are eliminating jobs, scraping our content without compensation, and building data centers that drain clean water from already underserved communities are also the ones using the word “agency” in their marketing. A lot.

Gozi noticed this too, and it bothered her enough to make it the focus of her thinking. AI agents. Agentic workflows. Giving you back your agency. It is a word that has been co-opted to sell a product that, in many ways, is doing the opposite of what it promises. Real agency means the ability to act on the world around you. To make choices. To shape outcomes. When tools are making decisions for you, when your job disappears because a billboard literally told someone not to hire a human, that is not agency. That is the loss of it.

You Can Be Skeptical and Excited at the Same Time

One of the things I love most about this conversation is that Gozi and I are both genuinely excited about what AI can do and genuinely alarmed by where it is heading. Those two things are not in conflict. You can use these tools, find real value in them, and still ask hard questions about who benefits, who gets hurt, and what we are quietly giving up in the process.

The hype cycle has been loud and relentless. We were told everyone would be replaced by machines by 2023. Now some of those same companies are quietly walking those statements back because it turns out building and running AI infrastructure is extraordinarily expensive and the returns are complicated. The fear was real. The jobs lost were real. And the recalibration is happening without much acknowledgment of the damage already done.

Gozi’s take: a lot of what we are hearing is noise. Technologies that were supposed to change everything come and go. Remember Skype? iPods? The question is not whether AI is powerful. It clearly is. The question is whether you are thinking critically about what you actually need it for, or whether you are just running toward it because you have been scared into believing you have no other choice.

The Wall-E Problem

I will be honest. This is the part of the conversation that keeps me up at night.

If we outsource our thinking too early and too completely, we stop developing the muscles we need to evaluate the output. Gozi put it simply: the whole point of the human in the loop is that you have enough knowledge and expertise to judge whether the answer is good. If you never develop that expertise because you handed it off to AI before you had a chance to build it, you are not using AI as a tool. You are becoming dependent on it.

This is especially urgent when it comes to young people. Kids in K-12 environments are still learning the mechanics of writing, reasoning, and problem-solving. They do not yet have the foundation required to evaluate what AI gives them. Using AI to write your essay when you are still learning what makes a good essay is not a shortcut. It is skipping the whole point. We are watching a live version of the Idiocracy scenario unfold in slow motion and we need to be a lot more intentional about protecting the development of critical thinking before it quietly atrophies.

Where Agency Lives: It Might Be More Local Than You Think

Here is where the conversation got genuinely energizing. Because AI feels like this enormous, untouchable force controlled by a handful of enormously wealthy companies, it is easy to feel like there is nothing you can do. That is not true.

Gozi made a compelling case for local politics and policy as one of the most powerful levers we have. Data centers are coming into communities right now and consuming massive amounts of water and energy. In some of those communities, the water was already compromised. If you do not want that infrastructure in your backyard, you can organize, you can show up, you can fight it at the local level. That is agency in action.

The same logic applies to the regulatory environment. The deregulation happening right now did not come from nowhere, and it will not fix itself. The work that created the EPA, consumer protections, and labor rights happened because people showed up and demanded it. We are going to need that same energy applied to AI policy before the decisions get locked in.

Agency also shows up in how your organization chooses to use these tools. Which platforms you select. What you decide AI should and should not touch. Whether you are asking your team to think first and reach for AI second. These are not small decisions and they are absolutely within your control.

We Are Still in the Jello

As I said on the episode, we are still swimming in unformed jello. It has not hardened yet. Which means there is still time to shape it.

That is actually the hopeful note here. The future of AI is not already written. The tools, the policies, the norms around how we use them are still being figured out. The people who opt out entirely do not get a seat at that table. The people who engage thoughtfully, ask hard questions, and push back when something does not align with their values do.

That is what agency in the age of AI actually looks like.

That Is a Wrap on Season Two

This is our last episode of UnTangled for the spring. We are heading into summer and honestly, Gozi said it best: go touch grass. Reconnect with your human. Let the technology take a break for a minute.

Here’s the full post:

We will be back after Labor Day and I have a feeling we will have no shortage of things to untangle by then.

In the meantime, catch up on any episodes you missed by checking out our UnTangled site, or following us on Spotify and Apple Podcasts.

And if you want to keep the conversation going all summer, subscribe to UnTangling the Week, Firefly’s newsletter, and UnTangling the Month, our LinkedIn newsletter.

And if any of this has you thinking about what AI adoption could or should look like at your organization, I would love to talk it through. Even if you are just trying to sort out where to start or what questions to ask, that is exactly the kind of conversation I am here for. Grab some time on my calendar.

Thanks for a great season. See you in the fall!