AI Hiring Gets More Specific, Role-Based Searches Rise

AI hiring is getting more specific, and I think that is an important signal. We are moving from broad interest in AI toward clearer demand for the people who can apply it in specific roles, workflows, and business problems. One of the most interesting takeaways from Upwork’s July hiring data is that three of the top 10 AI searches were role based, while AI UGC became the fastest growing AI related search. For leaders, the implication is pretty clear: the question is becoming less about whether to invest in AI, and more about what work needs to change, what skills are required, and who can make that happen. Worth a look if you are thinking about how AI is reshaping hiring and workforce strategy.

I’ve noticed a huge surge in AI-specific roles across all sectors. "Head of AI Strategy" is the trendy title of the moment. This reminds me of the late 90s, when companies created specific roles for the Internet because no one knew exactly where it fit in. Over time, those responsibilities were absorbed into IT, marketing, operations, and general management. I think the same will happen here, and rightly so. For now, however, it makes sense to bring someone into the organization who can explain the opportunity, set policies, and rein in reckless experimentation. The problem is that a job title doesn't make someone an expert. Neither does following the news, taking a course, or putting together a brilliant presentation. I use the word "expert" loosely, as I’m not sure anyone can truly be one in a field that evolves so rapidly. But there are three characteristics you can look for. A creator has built something. Not necessarily the next OpenAI; it could be an internal tool or a workflow. What matters is that they’ve experienced the gap between a demo and something that actually works in day-to-day operations. A user works with AI seriously and on a daily basis. They know which model suits which task. They are aware that an impressive answer can be wrong, and they have a method for spotting it. A thinker looks beyond the technology and focuses on people. An organization isn't software; it’s habits, incentives, fears, and politics. Implementing an AI product is trivial compared to changing the way people work. This brings me to an idea I keep coming back to. There is a "back room," where AI is created, and a "front room," where people are expected to use it. Most technology decisions are made by someone who understands only one of these two sides: developers who know the system but not the user, or executives who know the customer but not the tool's limitations. The ideal person is someone capable of moving between both rooms without the need for a translator. Agree or Disagree, I'd genuinely like to hear it. #Hiring #FutureOfWork #AI #Artificial Intelligence

I presented recently on sensemaking in insight generation. Data, then insight, then implications. Slow, brain-intensive work. A young colleague took the framework, fed it into AI, and showed me how well it ran. It was good. Faster than I would have been. I hit pause. Not on their process. On the thinking. Sensemaking is iterative and inductive. You go round the loop, get it wrong, notice what you missed, go again. Outsource it and you get the efficiency gain. You also get a learning loss that shows up years later. That exchange sent me into the research, and the research became a piece for Forbes. More than a third of entry-level jobs now screen for #AI fluency. We hire young people for it. We reward them for eighteen months for producing alone at speed. And we remove the exposure that used to build judgment, because it was bundled inside the slow work we automated. Five to eight years on, we assess them on judgment and the ability to drive something through a real organization. Then we call the shortfall a readiness gap. They met every criterion we set. We just set two of them, in different rooms, and never put the lists side by side. Link in comments…

Boolean search is not dead. It's just underused. Too many recruiters are still leaning on AI tools and stopping there. Those getting the best results are treating AI as step one, not the whole search. AI tools are genuinely fast at surfacing candidates. For most roles, they'll get you a solid list in minutes. But AI surfaces who it knows how to find. Niche skill sets, unusual titles, and non-traditional backgrounds often slip through. Boolean search doesn't rely on a model's pattern matching. You build the logic yourself, so you control exactly who shows up. When an AI search comes back thin, Boolean is often the reason you find candidates at all. It's the difference between "there's no one out there" and "I hadn't looked in the right place yet." The recruiters still fluent in Boolean have an extra gear the rest don't. It's not their main tool anymore, but it's the one they reach for when the easy tool runs dry. You see... Boolean didn't get replaced. It got demoted to backup, and backups matter most when the main plan fails. Drop a Boolean string in the comments that's still saved on your desktop. #boolean #AIRecruiting #Kleo30

Been looking at our own survey data again this week. 31.6% of companies say they use AI somewhere in hiring right now. I thought that number would be bigger, not gonna lie. Everybody talks about AI like it already took over hiring. Reality is many companies haven't touched it yet, or they tried it and backed off. And even the ones using it can't agree on whether it's actually helping. Curious where you land on this one. #AIinHiring #HRLeaders #TalentAcquisition #BackgroundChecks #HiringTrends

The tech industry is pushing us into an unprecedented paradox, and we’re heading toward a collective cognitive breakdown if we don't address it. Call it our defining cultural dichotomy: An operational obsession with AI, paired with a visceral rejection of its output. On one hand, we are eagerly handing over the keys to every facet of work. We want algorithms to streamline everything, finance, HR, accounting, campaign planning, and ad optimization. Look at modern ad platforms. Meta champions dynamic creative generation from product catalogs without a human touch. Google Performance Max asks marketers to "turn it on and leave it," letting AI draft ad copy, set bids, and handle targeting automatically. We crave the efficiency. We demand the scale. But here is where the breakdown happens. The second AI output enters the public domain, the ecosystem violently rejects it. On social channels, and increasingly on LinkedIn, content with telltale signs of synthetic generation suffers a massive drop in reach and impressions. Platform algorithms are actively re-tuned to penalize "AI slop" and reward authentic human signals. Consider the surreal reality we’ve built: 1) Behind the scenes: We automate processes and surrender human control for speed. 2) In the public square: We mock, reject, or algorithmically throttle content that feels automated. Nowhere is this mutual destruction loop more absurd than recruitment. Candidates use AI to tailor CVs to beat ATS algorithms. Employers deploy counter-AI detectors to flag those CVs. We have AI judging AI. Systems evaluating systems, while human connection gets erased. As Jamie Bartlett explores in his book "How to Talk to AI,: the central issue isn't what the technology can do, but human intent and governance. We don't need to ban AI, nor should we blindly surrender our voice to it. The path forward requires a clear boundary: 1) Invisible Utility: Let AI handle the heavy lifting behind the curtain, data parsing, structural optimization, predictive bidding, and workflow mechanics. 2) Visible Expression: Keep opinion, creative signature, and relationship-building uncompromisingly human. If we keep using AI to shortcut human connection while using AI to filter out synthetic noise, we just end up with algorithms talking to algorithms. Are you using AI as an intellectual multiplier, or are you letting it replace your human signal? #AI #ArtificialIntelligence #MarketingStrategy #FutureOfWork #DigitalTransformation #Leadership #AdTech #Automation

71 searches later, AI changed what I thought I knew about the skills employers actually want today Not what people on social media say you should learn or which AI course is trending. I wanted to know what the actual data says. At first, I tried researching it the way I normally would: searching for reports, opening tabs, comparing sources. But I was still the one doing all the research. So instead, I gave my AI employee, Viktor, the job of becoming my Research Lead. He ran ~71 searches across 3 research tracks, reviewed ~60 sources, and examined ~95 claims. Only 8 were cleared. After going through all of that research, these were the 7 skills that kept coming up: For everyone:
- Functional AI literacy
- Judgment & verification of AI output
- Workflow & process redesign
- AI governance & responsible use For more technical roles:
- Agentic AI: orchestration & context engineering
- Evaluation & error analysis
- Data fluency + Python But the most interesting part wasn't the list. It was what he told me not to share. He flagged ~24 claims as unusable or needing caution because they were outdated, misleading, poorly supported, or, in a few cases, referenced reports that didn't even exist. He even analyzed my own LinkedIn data and told me the sample wasn't strong enough to confidently use. That's when it clicked: I hadn't just asked AI to research something. I had delegated the research function. AI can now research, compare sources, challenge conflicting evidence, recommend the strongest angle, and build a fact-checking package for my team. I still own the thinking, judgment, and what ultimately gets shared. But my team and I don't have to spend hours collecting and validating information before we can get to that point. That's the shift I'm finding most interesting with AI: not just asking him more questions, but giving him more responsibility for the work that happens before the decision. Hire viktor.com for your team. $100 in credits included, no card. Full link in first comment. In partnership with Viktor. #Viktor #AIemployee #FutureOfWork 

AI fluency is now a hiring requirement. But is your team able to define what good looks like? Tomorrow at 2:00 p.m. ET, we’re bringing together three experts to discuss what AI fluency looks like in the real world. Join Bonnie Dilber of Zapier, Melany Austin of GoDaddy, and Adebayo Ojuri of CoderPad for a conversation about how hiring teams can move beyond buzzwords and build a clearer, more consistent approach to evaluating AI skills. They’ll dig into: • What separates genuine AI fluency from simply knowing how to use AI tools • What strong AI fluency actually looks and sounds like during an interview • How leading organizations are incorporating AI into hiring and competency models • How to design assessments that produce real signal as AI continues to evolve If your organization is still figuring out where the AI proficiency bar should be, this conversation is for you. Join us tomorrow, August 26 at 2:00 p.m. ET. Register free: https://lnkd.in/gdnjCvtG Can’t make the live broadcast? Register anyway, and we’ll send you the recorded playback shortly after the session ends.

47% of candidates now list AI skills on their resume... and 64% of them say no one has ever tried to verify those skills. 😱 So on October 7 at 3pm ET / 12noon PT we're sitting down with Terrence Fox from CoderPad to get the details on what can be done: what AI fluency really means; why it's now table stakes for GTM, finance, ops and marketing roles (not just engineers); and how to test for AI skills in the hiring process you're already running. If you're still taking "AI-fluent" as a conversation about the last time someone looked something up with Claude, you're hiring on vibes. Let's talk about how to do better. Registration link here ⤵️ https://lnkd.in/ey85FmWp

Introducing SFI – Search Friction Index: a new series where we break down the friction behind specific hires, so founders and recruiters can understand what will actually make a search difficult. Remember our UA Lead search for Rask AI? https://lnkd.in/dwPHWV_A Let’s look at it through the Search Friction lens. #Hiring #AI #UserAcquisition #Recruitment