Originally published as an Advisory Hour Substack post on 2026-09-28. This first-party site copy preserves the text, localizes public supporting media, and connects the record to its project trail.
Original post
I tasked Claude Opus5.5 to create a video based on, well, me and found one oddball future for the resume, LinkedIn and WorkDay. It’ll be something like “HireTok”. A TikTok for jobs and your resume. (Sorry, I know.) Why is because of the resume application matchmaking problem. It’s 1900 coded. I should know, I built a couple of the HR systems last century.
The way those HR systems worked is based on a fundamental mismatch that plagued early social networks. The mechanism for the resume to job application system is simple enough: the job you’re hiring for has keywords. The applicant submits a resume with their keywords. Putting A plus B together means there’s a match. Where there’s a match, it shows up in a report on the hiring manager’s desk, sorted by filter criteria like “college degree”, “time between employers”, or even “credit score”. It made sense in the 1900s because it was the extent of what algorithms did and, by sheer industrial laziness, no one has bothered to really change the formula.
The matchmaking system appeared in early social networks.
In the early days of Facebook, for example, you had to declare your interests and these then helped inform the social graph and matchmaking algorithm who all you should see in your feed. The most common UX pattern for this was a dizzying array of checkboxes.
If you seeded the content well, you’d see more content that matched your interests. And if you didn’t, then you’d see a glimpse of the deeper problem in these kinds of systems. Interests change and in a world where job titles are increasingly ephemeral (how long did Prompt Engineer last as a job title?) it’s clear that the Employer and Applicant match making system are overdue for a change. I still see job boards spring up, but they’re the wrong form factor as we swiftly move into 2030.
The future of hiring likely need two things. First, something that help people make a compelling video that explains who they are. Second, an employer that has certain interests. There must be a mechanism for that employer to tag those interests without explicitly stating them-the same way one does while swiping TikTok.
A hiring manager today on say a job board will post a job and then get back hundreds, if not thousands of job applicant hopefuls. The problem is this is still twisted up in the old system of keyword match making.
What would make more sense is a behavior change coupled to a direct use of AI technology. On the employer side it’d make more sense to feed signals into the system in a way that the interests are properly modeled capability that you’re looking for, and also tracking an engagement of the kinds of hires that culturally make sense for your tribe. There’s two layers to every hire: can they do the job and can they live the job.
When I asked AI to create a video about me, I wasn’t sure what to expect. If you’ve seen my other AI generated videos, then you know we can get just about anything out of a prompt like that. Instead, I got a gift. What came back was quite a personal delight.
And in that gift I realized that if AI could just whip something like this up for me, then surely it could for you. Further, if you can imagine LinkedIn filled with “who am I” videos, then that’s where it’s heading. I can so easily imagine a LinkedIn feed just slopped up with Claude generated “about me” videos of applicants trying to get the upperhand on the job application process. That could lead into someone realizing that there’s a better way, and it likely involves the interest graph.
And if you can imagine that, and connect with my thought on why this is linked to the hiring process, then you’ll also see why something like a “HireTok” is likely the future of how firms find not just the next candidate-but the best candidate possible.
And you know? It’ll probably largely be powered by AI, if not one of the larger labs. After all, they already know so much about us.
Connected work
- HireTok in VideoLab: Watch the AI-generated introduction that prompted the essay.
- HomeTok: A TikTok Clone at Home: Explore another experiment with video discovery and interest graphs.