Jerry and George Discuss Worlds Colliding
If you’re a Seinfeld fan, you probably remember George Costanza’s fear of his worlds colliding.
George had Relationship George. Independent George. And under no circumstances were those worlds supposed to meet.
Recently, AI caused two of my professional worlds to collide: market research and accessibility.
And unlike George, I’m actually pretty happy about it.
World #1: Research
I recently worked with a series of in-depth interviews with physicians.
The subject matter was highly technical. These weren’t simple interviews where you could quickly scan the transcripts and say, “I see three themes.”
There was a lot of information. A lot of nuance. And I needed to categorize responses in a way that would allow for meaningful comparisons across interviews.
This is the kind of work researchers have been doing forever. Read. Code. Categorize. Compare. Go back and check. Reconsider a category. Move something. Read it again.
It works.
It can also take a very long time.
So, I brought AI into the process.
I used it to assist with content analysis, helping organize and categorize responses so I could more efficiently compare what physicians were saying across interviews.
The important word there is assist.
I wasn’t asking AI to tell me what the research meant. I still needed to understand the context, question whether the categories made sense, recognize nuance and determine what was actually important.
AI helped with the heavy lifting.
The researcher still had to do the thinking.
Then Another World Appeared
Not long afterward, I was working with a PDF presentation that needed to be made accessible.
Completely different assignment.
Completely different skill set.
Or so I thought.
Again, AI became part of the process.
This time, I leveraged it to help remediate the document, including restructuring and tagging the PDF so that it could be better navigated and understood by someone using assistive technology.
And somewhere during that process, I had my George Costanza moment.
My worlds were colliding.
Here I was using the same underlying technology for two very different things.
In one case, AI was helping me make sense of complex qualitative research.
In the other, it was helping make information more accessible.
There Was Also an Interesting Similarity
The more I thought about it, the more I realized these two examples weren’t as different as they first appeared.
Neither was really about AI doing the job for me.
In research, having AI categorize responses doesn’t mean the analysis is finished.
In accessibility, having a PDF tagged doesn’t automatically mean you’ve created a good accessible experience.
Both still require judgment.
You need to ask questions.
Does this make sense?
Is something being missed?
Is the structure right?
Would a person actually understand this?
That’s where I believe we sometimes get the AI conversation wrong.
We spend a lot of time debating what AI can replace.
I’m becoming much more interested in what AI can enable.
Connecting Things That Didn’t Used to Connect
I’ve spent much of my career trying to understand people.
What are they experiencing?
What are they telling us?
What aren’t we hearing?
Accessibility has led me to ask many of those same questions from another direction.
Who might be excluded?
Whose experience aren’t we considering?
Who might not be able to participate in the first place?
AI didn’t create either of those interests for me.
What it did was give me another way to bring them together.
And that’s the part I find exciting.
Not because AI magically solves research.
Not because AI magically solves accessibility.
It doesn’t.
But it can give us new tools to work across disciplines, remove some of the tedious work, and spend more of our time on the things that require human judgment.
Understanding.
Questioning.
Interpreting.
And acting.
So, George Costanza may have been terrified of his worlds colliding.
I’m starting to think that when it comes to AI, the collision may be where some of the most interesting things happen.