Why Job Search Feels Broken for Tech Today
Latch Editorial
2025-12-18 • 4 min read
Hiring didn’t break because there aren’t enough jobs or talent. It broke because everyone automated the wrong things. What once moved through human judgment now moves through AI automation. Bots talk to bots, nuance disappears, and the system quietly falls apart.
ATS: The First Gatekeeper
ATS systems were meant to organize resumes. Instead, they became AI-powered bouncers filtering humans by keywords.
- Strong engineers get rejected because they didn’t phrase “REST APIs” the right way
- Career switchers never make it past filters designed for linear paths
- ATS doesn’t understand impact, learning speed, or real-world problem solving
So candidates optimize for parsing, not truth. The result? Resumes written for machines, not people.
Resume Analyzers: Training People to Game the System
Resume analyzers promise: Increase your ATS score. What they actually do:
- Standardize everyone into the same template.
- Same bullet formulas
- Same action verbs
- Same metrics, sometimes completely made up
When everyone has a 90% match score, no one stands out. The resume stops being a signal and becomes a compliance document.
Resume Builders: Manufacturing “Perfect” Candidates
AI resume builders don’t ask who you are. They ask what job you want to trick. So candidates:
- Inflate experience
- Borrow language they’ve never used
- Mirror job descriptions line by line
Now recruiters complain: “Resumes look great but candidates can’t back them up.” Of course they can’t. You trained them to fake it.
AI-Written Job Descriptions: Vague, Unrealistic, and Detached
Recruiters now use AI to write job descriptions. Sounds efficient, until you read them.
- 5 years of experience in tools released 2 years ago
- “Fast-paced environment” as a warning label
- Responsibilities that belong to 3 different roles
These JDs are optimized for volume, not clarity. Candidates don’t apply because they’re a fit, they apply because they think they might survive the filter.
AI Interviews: Cheating Becomes a Feature
Candidates use AI to cheat during interviews:
- To generate answers of coding rounds
- To explain concepts live
- To sound confident in areas they don’t fully understand
Results companies use AI to track and reject later
AI to Detect AI: The Most Absurd Loop
Recruiters respond by using AI to detect AI usage. So now we have:
- AI answering questions
- AI detecting AI answers
Humans stuck in the middle pretending this is normal. It’s no longer hiring. It’s surveillance theatre.
AI Shortlisting Thousands of Applications
Companies receive thousands of applications, not because everyone is qualified, but because applying costs nothing. So they deploy AI agents to shortlist. These systems:
- Favor keyword density
- Penalize non-traditional backgrounds
- Reward over-optimization
The best candidates don’t always survive the first pass. The best optimized ones do.
Auto-Apply Bots: Spam Disguised as Ambition
Candidates now auto-apply to hundreds of jobs a day using AI agents.
From the candidate’s side: “It’s a numbers game.”
From the recruiter’s side: “We’re drowning.” Everyone is exhausted. No one feels chosen.
Everyone Is “Perfect”: So No One Is Trusted
This is the core problem.
- Perfect resumes
- Perfect cover letters
- Perfect interview answers
But perfection create zero trust. Recruiters don’t trust candidates. Candidates don’t trust job posts. Everyone assumes deception by default. That’s not a hiring system. That’s a cold war.
The Real Cost: Wasted Time, Burnout, and Mismatch
What gets lost in all this automation?
- People spend weeks tailoring resumes that never get read
- Recruiters screen endlessly and still hire wrong
- Great talent ends up underemployed or invisible
- The system rewards who can game it best, not who fits best.
The Truth No One Wants to Admit
Hiring didn’t fail because of AI. It failed because we used AI to amplify broken incentives.
Instead of asking: “Who is this person really good for?”
We asked: “How fast can we process them?”
Speed replaced alignment. Volume replaced intent. Automation replaced understanding.
Why This Matters (And Why It Has to Change)
People shouldn’t have to pretend to be someone else to get hired. They should find roles that align with who they already are. When alignment improves:
- Interviews become conversations
- Resumes become summaries, not lies
- Hiring becomes faster and more human
Until then, tech hiring will keep feeling broken, because it is.