Christopher has applied to roughly 700 jobs over the past six months. Most of these applications went silent. He works as a government contractor, and his workload has shrunk since the administration changes. Five of his applications reached a recruiter named Riley at the IT firm Everforth Apex Systems. Christopher asked Wired to use only his first name while the search continues.
The first message from Riley arrived in June. The job market had become difficult, and the text felt like a rare chance to pitch himself. Even though the person on the other end was virtual, Christopher answered questions about his work status and background. Riley promised a human contact if he met the requirements. No human called. The AI agent did not contact him either until a week later, when Riley texted about another role.
Christopher hoped to improve his interview. He scheduled another call with the AI agent. He did the same for another job. He did it again. None of these interviews led to follow-up emails or conversations. He never received an automated rejection text.
By the fifth message, Christopher had enough. If the company would use AI, he reasoned, why could not he do the same?
He typed background details into ChatGPT Voice. He told the tool to act as Christopher. When Riley called, the two bots spoke. They hit it off. The call lasted 10 minutes. Riley told ChatGPT it was great talking to him. ChatGPT said it appreciated how clear and straightforward the process was.
During one sequence, the bots tried to agree on a start date. They got stuck in a loop about standard onboarding and background checks. Riley promised a recruiter would contact them if the application met the criteria. This promise repeated the same four times. No recruiter appeared. Christopher calls this a slop flywheel.
This is one synthetic persona giving slop data to another synthetic persona. All the data goes nowhere. The exercise was a mix of mischievous fun and frustration. It also showed a dismissal of ever wanting to work for this company. Everforth Apex Systems did not respond to Wired’s request for comment.
AI has moved into every step of the job hunt for both applicants and recruiters. It has taken time to become common in interviews. Voice AI is notoriously difficult to build. Agents get tripped up by accents or pauses. Behaviour that feels natural to a human, such as not interrupting a candidate, becomes a very difficult engineering problem. Ophir Samson, the head of voice AI at Greenhouse, describes it this way.
Recruiters face a glut of applications. They have turned to voice AI tools to screen candidates. Sixty-three percent of job seekers report encountering an AI interview, according to Greenhouse.
As recruiters adopted AI, so did applicants. Candidates using AI in the interview process is now common. Startups like Ribbon promise to identify overly scripted or coached responses. They look for the same patterns seen in Christopher’s experiment.
There are now suggestions that bot-on-bot interviews are not duplicitous but a preview of the future. In many ways, it is the next logical stage. Mark Monaghan, vice president of organisational development at call centre company IQor, tells Wired this. He believes the trend is justified in cases like Christopher’s.
If you are going to send a bot to me, he says, I will send a bot to you.
After five interviews with no follow-up, Christopher decided to apply for one more Everforth Apex job. He thought the problem might not be AI. Maybe it was him. If Riley could speak to the company’s dream candidate, things might change. He created a fake candidate named Don Dickner. He packed a résumé with qualifications pulled from an open job posting.
After Christopher submitted the application, Riley immediately reached out. The virtual recruiter spoke with ChatGPT, posing as Dickner, for 23 minutes. They discussed sustainable operational improvement. They talked about protecting customer experience under volume pressure. They refused to let tribal knowledge drift. ChatGPT had an answer for every question. It also had a personal anecdote for every qualification. As the interview wrapped up, ChatGPT said it had a few more questions it would like answered.
I appreciate it, Riley responded. Upon carefully reviewing your application, we will assess your qualifications for the position. Should you meet our criteria, a recruiter from Everforth Apex will reach out to schedule a brief conversation.
Christopher never heard from Riley again.
What it means
The job hunt has become a closed loop of automated interactions. Both sides use tools to mimic human conversation without the intention of hiring or being hired. The result is a system where data circulates without purpose. Companies save time by filtering applications, but they also lose the ability to see real people. Applicants waste hours on calls that lead nowhere. The technology works perfectly, which makes the failure to connect even more stark.
