I wrote a piece in August 2023 with a deliberately bold title: AI will save Recruitment. At the time, the take landed as either too optimistic or borderline naive. The headlines were about AI co-pilots writing 80% of code and synthetic Drake tracks going viral. The mood in talent was: AI is coming for our jobs.

Three years on, the prediction has stopped being a prediction. The data is in. Agentic AI showed up. Recruiters who leaned in are running circles around the ones who didn't. And the gap is widening every quarter, not closing.

So I want to revisit the original thesis with the benefit of evidence, with what I've learned running Metaview through this shift, and with the numbers from our 2026 AI & Hiring Alignment Report sitting right next to me. The original four-minute manifesto was correct in direction. It just understated the magnitude.

The drudgery thesis, three years later

The core argument in the 2023 piece was simple. Recruitment is one of the most human-centric professions on the planet. The work that matters most (nuanced judgement, persuading candidates, advising hiring managers) is the work furthest from automation. What sits between planning and performance is the drudgery: scheduling, transcribing, chasing feedback, drafting outreach, screening 800 applications for one role. That's where AI lands first, and that's the layer it removes.

What I underestimated in 2023 was how fast that layer would collapse. I expected the shift to take five to seven years. Agentic AI compressed it to two. The recruiters I talk to today are not asking "will AI take notes for me?" They are asking which of their seven workflows they should hand to agents next, and how to keep their hiring managers calibrated on the output.

The original four-minute read covered three claims: AI means less drudgery and more zone of genius, increased efficiency means increased demand for recruiters, and avoidance is not a winning strategy. All three held. The first one held with room to spare.

What recruiters actually do now

If you walked into a recruiting org in 2023 and asked a sourcer to describe their day, you got a list of mechanical tasks. Boolean strings, linkedIn searches, spreadsheet imports, sequencing in Gem or Lever. Manually scoring 50 profiles a day against a job description nobody had read end to end.

If you walk into that same org in 2026, the job has been rewritten from the inside. The sourcer is reviewing the output of an AI sourcing agent that already ran the search, ranked the slate, and flagged the top 15 against the ICP. The interview notes get written by the AI during the conversation, leaving the recruiter free to listen and ask the next question. The application screening that used to take six hours a day takes 45 minutes of triage on a ranked pipeline.

This is not a productivity story. It's a job-redefinition story. The recruiter's hands move less, their judgement gets exercised more. The conversations with hiring managers are deeper. The work the recruiter takes home at night is the work that needs a human (calibration calls, persuasion, market intel), not the work an LLM can do in 30 seconds. That's the shift I called in 2023, and it landed harder than I framed it.

Jevons paradox in recruiting

The 2023 piece spent a paragraph on Jevons paradox. Make a resource more efficient, total usage of that resource goes up, not down. Light bulbs got more efficient, and we put more bulbs in more rooms. I argued the same thing would happen with recruiters: efficient recruiters would become more valuable, not less, and demand for them would rise.

This is exactly what's happened. The recruiters running multi-agent workflows are getting promoted, getting raises, and getting pulled into more strategic work. They're hiring for senior roles their peers can't fill. They're advising founders on org design. The best ones operate at a scope that didn't exist three years ago: solo recruiters running pipelines that used to need teams of four, but doing it with higher-quality slates and faster cycle times.

The companies that figured this out earliest are running leaner talent teams that hit higher attainment numbers. According to Metaview's 2026 AI & Hiring Alignment Report, surveying 505 recruiting leaders and hiring managers across North America and EMEA, the teams that report AI as core to their hiring process are 5x more likely to rate the recruiter-hiring manager partnership as excellent than teams where AI is incidental. The use compounds. Better partnerships make for better hires, better hires make for better referrals, better referrals reduce the load on sourcing, and on it goes.

The bimodal recruiter

The other thing the original piece predicted, and that I want to put a sharper edge on, is the bimodal distribution in the profession. There are now two kinds of recruiters: the ones who built AI fluency between 2023 and 2026, and the ones who didn't. The middle of the distribution is hollowing out.

The fluent group is unrecognizable from their 2023 selves. They run a stack. They know which agent to point at which problem. They have opinions about prompt patterns. They review LLM outputs the way a senior editor reviews a junior writer: trust but verify, then ship. They are, by every measurable signal, better at the job than they were three years ago.

The other group is still doing the same job they did in 2023, but the bar moved underneath them. Their work feels harder. Candidates are slipping away to faster competitors. Hiring managers are quietly comparing their cycle times to peers at other companies and finding them slow. The grind that AI promised to remove is still there, because they never picked up the tools that remove it. The penalty for avoidance is paid in lost reqs, slower hires, and shrinking influence.

This is the operating environment I want every recruiting leader to internalize: the fluent group is widening their lead every quarter. Catching up gets harder, not easier, with each passing month. If you can read how the best recruiters use Claude and not feel a pang of "I need to start doing this tomorrow," the gap is already wider than you think.

Where AI actually removes drudgery

Three years of building Metaview taught me where the drudgery actually lives. It is not where most recruiters expect. It hides in the seams between workflows: the handoff from sourcer to recruiter, the gap between interview and scorecard, the silent decay of application piles that never get reviewed, the meetings about hiring metrics nobody bothered to instrument. Agents kill drudgery one seam at a time. Here's where the cuts land hardest.

Sourcing

Removes the drudgery of Boolean strings and profile scoring. The agent runs the search, ranks the slate, and flags the top candidates against the ICP so the recruiter spends time on judgement, not on filtering.

Application Review

Removes the drudgery of triaging 800-application piles. The agent reads every applicant against the ICP, surfaces the strongest fits at the top, and lets the recruiter focus the 45 minutes that matter on real conversations.

Notes

Removes the drudgery of frantic interview typing. The agent captures verbatim, summarizes the key moments, and fills scorecards so the recruiter listens to the candidate instead of racing the keyboard.

Reports

Removes the drudgery of stitching together hiring metrics across systems. The agent assembles the dashboards, surfaces the funnel gaps, and gives leaders the data to advise founders on org design.

The pattern across all four surfaces is the same: the agent does the work that does not require human nuance, and the recruiter spends the saved hours on the work that does. The data backs the framing. Speed of process is now the single biggest predictor of whether a team wins or loses on closing rates.

Read those four numbers as one sentence: the cost of being slow is candidate loss, the cost of being AI-laggard is being slow, and the cost of being AI-core is the lowest candidate loss rate in the report. The drudgery thesis from 2023 was about quality of work. The 2026 update is about quality of outcomes. Faster cycles, better partnerships, higher conversion. The recruiters who picked up the tools are landing the candidates the holdouts are losing.

Avoidance is still not a winning strategy

I closed the 2023 piece with this line: "avoidance is never a winning strategy in the face of technological revolutions like the one we're in now." Three years later, I'd refine it to something blunter. Avoidance is a confirmed losing strategy. Not theoretical, not a hedge, not a probability statement. The data is in. The teams that avoided AI between 2023 and 2026 are losing candidates at 80% rates, while the teams that built fluency are losing them at 50% rates and pulling away.

If you're a recruiter reading this and you still haven't adopted an interview-notes agent, a sourcing agent, or an application-review agent: this is the moment. The argument about whether AI will help recruiters is settled. The argument now is whether your team picks up the tools in the next 90 days or watches another quarter of candidates slip to the competitors who already did. Start somewhere small: take your last 10 interviews and run them through Metaview's interview notes. Look at the time you save. Then look at the conversation quality. The decision tends to make itself.

If you're a recruiting leader, the question is not "should we adopt AI." It's "are we building the team that wins the next 24 months or the team that gets replaced by the team that does." The talent strategy of the next two years rides on this. You can read more on how to think about it in our take on talent density, and on what excellent looks like when partnership and AI work together in our Wall of Love. I'll close the same way I closed in 2023: I personally can't wait to see what the future of Recruitment holds. I'm just more confident now than I was then about who wins it.