I just returned from the 2024 Information Architecture Conference. I sensed high levels of engagement and energy, especially from new folks. But if asked to summarize the conference’s main theme, I’d say it was “Information architects are deeply concerned about AI.” It’s a mistake.

To be fair, I wasn’t able to hear all presentations. But several struck me as being overly pessimistic while ignoring AI’s obvious advantages. The same (negative) case studies came up in several presentations, with very few mentions of positive use cases. One audience member asked for advice on how to push back on AI initiatives at work. Another mentioned being ‘terrified’ at the prospect of using AI for translation.

The negativity is understandable. Over the last two years, we’ve been fed a continuous stream of FUD, mostly from people who stand to lose status, opportunists inflating expectations for profit, and cynics and ideologues who oppose new technologies on principle. Of course, big changes can also be scary without the hype.

But it’s a mistake for information architects to reject generative AI out of hand. While the technology has downsides, it can also be put to productive use — particularly by people who work with language, such as information architects.

Let’s not mince words: AI is hugely disruptive. Disruptive technologies change things. Every once in a while, one comes along that spins societies toward a different direction. Generative AI is one of those — and it’s not going away. So we must engage with it.

People living through radical tech-driven disruptions have a choice: they can either embrace or reject new technologies. Only those who embrace change have a say on future directions. I’ve always understood this idea in theory, but only grokked it viscerally recently when I stumbled across a 102-year-old movie called THE BLACKSMITH. It’s on YouTube:

In this 25-minute silent comedy, Buster Keaton plays a bumbling assistant blacksmith. The humor lies both in his incompetence and in surrealistic sight gags such as showing a horse a choice of horseshoes as though it was shopping at a fancy shoe boutique. This must have been hilarious at the time. Now, it just feels quaint.

1922 is a very distant time from our own, chronologically and psychologically. The film opens with a title card that says, “Under a spreading chestnut tree the village smithy stands.” The director isn’t compelled to explain what a “village smithy” does. Not surprising, since many people still used horses as a power source back then. Horses needed shoes, so villages needed blacksmiths.

I doubt you know anyone today whose full-time occupation is blacksmith. The internal combustion engine mostly did away with blacksmithing as an occupation. But it didn’t do away with the people doing these jobs. Instead, many became the first generation of auto mechanics.

It’s obvious why: blacksmiths had skills that were highly relevant to the new technology. Also, customers already thought of them as the ‘transportation fixers.’ The transition was underway when the THE BLACKSMITH was made: the film shows Keaton destroying customers’ automobiles, including a Model T.

You could argue the internal combustion engine has been a net negative on the world. But it’s undeniable that it changed society. And the change happened despite the expense, unfamiliarity, and dangers posed by early jalopies. Why? Because people loved the convenience and independence afforded by affordable automobiles.

Something similar is happening with generative AI. Yes, the technology is problematic: it can up-scale misinformation, sometimes gives out wrong answers, and consumes a lot of energy. It can reflect and amplify biases and will likely displace a lot of jobs.

Also, many organizations are also following a “ready, fire, aim” approach to implementing AI, leading to the sort of spectacular fails so prominently featured at this years’ IAC. (Replacing humans service agents with chatbots seems particularly egregious.)

Which is to say, it’s early days, and we’re in the midst of a hype cycle driven by greed, ignorance, and FOMO. There will inevitably be disillusionment as reality (read: lawsuits) force organizations to move on from magical thinking.

But claims that generative AI is either useless or inevitably toxic are patently absurd. There are many people (including me) using these tools productively. You don’t have to squint hard to see how they could drive better experiences more cheaply and efficiently.

Mastodon post from Greg Wilson that says: 'Starting to wonder if Al is like spreadsheets: for every programmer pointing out flaws and deficiencies, a double dozen people are using 'em to do something they find useful. 1/4 Added: please see mastodon.social/@gwilson/1122... for clarification.'

It’s important to have thoughtful voices in the rooms where decisions about how to use AI are being made. Given the obvious upsides, outright naysayers will be excluded. It’s much better to be constructive participants in these discussions — especially if we have relevant skills.

Information architects are well-positioned to help steer these conversations in positive directions. After all, we are the language systems people — i.e., the team members who look after conceptual integrity across contexts. Putting LLMs to work toward humane ends is in our wheelhouse.

One way to start is by defining principles for the appropriate use of AI in architecting information systems. In her presentation at IAC24, Kelsey Thomson laid out several. One that stood out to me is that we should use AI as a tool, not as a decision-maker. This aligns with how I’ve approached AI: as a way to augment my work — a design tool, not a production one.

We should do more of this: engage with the technology to help define principles and processes for its productive use. It’s essential that we roll up our sleeves and work with the technologies directly, rather than learning about them second hand from blog posts, podcasts, academic papers, and conference presentations.

Blacksmiths who saw their job as shoeing horses didn’t have a future. Only those who understood they were transportation enablers and engaged with the technology managed to thrive through the disruption.

It’s now our turn. Generative AI isn’t going away. IAs have an important role to play in ensuring it’s used productively towards humane ends. Our mission isn’t to organize taxonomies, structure navigation bars, or draw wireframes. It’s to improve peoples’ lives by making the information they need more understandable and easier to find. AI changes how we do it. We’d be foolish to reject it.