How AI Has Helped Me Feel Psychologically Safe

One major criticism of AI is how much water its data centres use. Do you know what else uses water? People. Including me, staying up late crying at a spreadsheet and plugging my laptop back in because I’m still not done.

If we’re counting the cost of getting the work done, can we count that too?

I went to Toast Summit yesterday, and it was amazing to connect with so many other women in tech. I also heard concerns about AI, our data, the resources it uses, and the harm people can cause with it. What surprised me was how easily that became a conversation about AI itself being the problem.

AI has changed my life. So hearing it demonized is jarring, especially when the conversation leaves so little room for what people are actually doing with it.

It’s a tool. Yes, some people are building bad things with it. Yes, companies are going too far. We need to be able to talk about those things without treating every use of AI as the same decision.

I spent my summer trying to pitch an AI startup, so if I know one thing, it’s how little most people know about AI. I know that very, very, very well. The AI somebody deploys to hack ten thousand websites and the out-of-the-box chatbot the average person knows how to use keep getting thrown into the same conversation. Those uses have different purposes, different capabilities, and different consequences. Calling both of them “AI” doesn’t tell us enough to judge what’s happening.

We understand this distinction elsewhere. Water can keep someone alive. It can also drown them. We know that what someone does with it matters. I want that same attention to purpose and consequences in the AI conversation.

The data argument has its own version of this. I’m not saying throw your SIN into a chatbot. But the sudden alarm about AI and our data, while we’re already on Instagram and other platforms that make a business out of knowing us, is a little funny to me. Our data didn’t start being valuable to companies when we learned how to prompt. That ship sailed a long time ago.

We should care about what companies collect and how they use it across all of these tools. Being familiar with a platform doesn’t make its practices better, just as calling something AI doesn’t tell us everything about its risks.

It reminds me of the Family Guy joke about the Amish having decided that humanity reached just the right amount of technology in the nineteenth century.

The wheel, fine. Fire, okay. Light using electricity, absolutely not.

Sometimes it feels like we’ve done the same thing. An automated workflow was the perfectly acceptable amount of technology. An agentic process? Now we’ve gone too far.

“But Paisley, this one’s worse for the environment.”

I wondered about that too. Using published estimates, processing 60 basic ChatGPT prompts comes out to roughly a quarter of the electricity estimated for an hour of Netflix.*

Have we actually examined the data before deciding which uses of technology deserve our disapproval? And have we considered what someone gets out of using it?

If you care about the environment, find the actions that will make the biggest difference. I’m not convinced that making someone feel guilty for getting help with a spreadsheet is where we should start.

Because while I was hearing how harmful AI is, I kept thinking about how much it has helped me. AI has made me feel psychologically safe.

I have ADHD, and getting to a deep thought takes scaffolding: examples, connections, exceptions, years of things I’ve learned. Losing the thought can mean losing the route I took to get there. Being able to talk it through and save the context somewhere I can return to takes some of that pressure off. I can leave a thought without feeling like I’m abandoning it.

A hearing aid uses technology to help someone hear. We don’t call that cheating. We understand that the person is getting support for something that’s harder for them. So why does help with thinking, organizing, or getting a thought into words become a question of laziness? You might not need that support. I do. And finding it has made me realise how much effort I was spending just trying to function without it.

Using a tool that helps me function doesn’t mean I approve of everything the company behind it does. Tech billionaires and the investors funding them are choosing to build more data centres and race toward bigger systems. I was fine with the AI we had six months ago. I was fine a year ago. Why am I the one expected to pay the inconvenience tax by boycotting a tool that helps me get through my day?

We should be holding these companies accountable for what they collect, what they build, and what they consume. We can demand better from them while recognizing the good the technology is doing.

AI has changed my life. I want more people to have the chance to find out what useful support could look like for them.

Let’s use AI for good.