Ben Pastor

LinkedIn just launched something big! Hiring Assistant is the first AI agent built into LinkedIn Recruiter and Jobs, and it’s built for enterprise scale. It’s not just smart, it’s trained by real recruiter actions and shaped by feedback from some of our largest charter customers. It’s already helping teams surface overlooked talent, reduce busywork, and hire faster. #LinkedInHiringAssistant is here to help recruiters do more of what matters, creating world class experiences for our candidates and hiring managers. Curious what it looks like in action? Check it out: https://lnkd.in/gHKbxnZm

84

Layoffs. It’s the talk of the town. The recent layoffs at companies like Expedia and Amazon have the market panicking, and understandably so. It’s hard not to question what the job landscape looks like next. If you’ve been impacted by a layoff (wherever you are), don’t panic. Easy to say, I know, especially when the advice is coming from LinkedIn. But grounding yourself in action helps. Here are 3 practical steps you can take right now to move towards your next role 👇 1️⃣ Update your CV Focus on outcomes, not duties. Highlight your key achievements & measurable impact. 2️⃣ Refresh your LinkedIn 95% of recruiters and hiring managers will look you up. Make sure your profile clearly shows who you are, what you do and the value you bring. 3️⃣ Set up job alerts Subscribe to notifications on job sites relevant to your country and industry so opportunities come to you. These are starting steps. Simple, but powerful. There is a place for everyone. I truly believe that. And yes, you WILL find a job 💛

50

CNBC recently reported that there may be millions of “ghost jobs,” based on the 2.2M gap between job openings and hires. The gap is real (it's based on BLS data), but the interpretation is up for debate. They cite analysis from "My Perfect Resume", and they draw the conclusion that ANY gap represents a "ghost job". Some other considerations: 1. They measure openings and hires at different times. If a company posts a job in September, makes the hire in November, and the new hire starts in December, we have 3 openings but 1 hire. In other words, even if 100% of jobs that were posted were filled, we would see a discrepancy unless every role is filled (with butt in seat) in the same month it was posted. 2. The longer it takes to fill a role, the greater this gap will appear to be. Roles are taking longer to fill - I know that doesn't seem intuitive in this market, but the reality is that those who are looking for work may not align well with the roles companies are trying to hire in terms of skills, location, etc. Layer on top of that, that companies are hiring fewer people and being more selective. They would rather hold out 4 months for the right person than take a risk on a stretch hire in 2 months. This means openings are being counted for many months before a hire is recorded. 3. Openings are self-reported by companies, and most economists say it's the least reliable metric in the JOLTS report. There IS a definition, but it's kind of fuzzy and there's no audit. There's not necessarily a posting that would represent a "fake job" to an applicant. Let's just consider jobs posted on LinkedIn: we see 3.8M jobs posted in the US. Not all employers post here, but about ~75% of recruiters use LinkedIn so this might suggest a total of ~5M job postings. If 30% of all job postings were truly fake, we’d expect to see far fewer than ~5M hires per month happening in the labor market. None of this changes the fact that: - people in certain fields are really struggling in this job market - that there's real misalignment between what jobs openings and the skills, interests, and locations of the workforce - that many people are underemployed - that many companies are unrealistic in the bar they're holding - that many companies have crappy hiring processes. - that many companies But none of that means that employers are intentionally posting fake jobs en masse. I think that nuance is important. I think it impacts how people approach their job search. If you believe every company is out to get you and 30% of jobs are fake, you may engage in an adversarial way or try to "game the system" (with mass appliers, fake resumes, etc.). If you believe companies are hiring in good faith but being extremely picky, you're more likely to be thoughtful about what jobs you apply for, take steps to stand out, and show up as the best possible candidate. And one of those strategies leads to much better results for applicants than the other.

43

30 Comments

Great insights from Gem‘s new Recruiting Benchmark Report👇🏻 ⓵ Hiring Processes Are Becoming More Selective - Interviews per hire increased by 33%. The report notes this is primarily driven by technical roles, but from my experience, Go-To-Market (GTM) roles are just as selective. Our passthrough rate at the TA interview stage is only 20–30%. These roles are critical for any SaaS organisation, so finding the best possible talent is non-negotiable. ⓶ Sourcing Remains One of the Most Effective Hiring Channels - Sourced candidates are ~8x more likely to be hired. This makes sense — sourcing lets us target the exact profiles needed for the team, rather than waiting for them to emerge in the inbound pipeline. Most TA teams also maintain warm talent pools over time, ready to activate the moment a role opens. Stay tuned for more insights from the report later this week! #TalentAcquisition #Recruiting #HiringTrends #GTMHiring #SaaS #Sourcing #RecruitmentInsights #HRAnalytics #GemBenchmark

16

1 Comment

A little peek behind the recruiting curtain as we kick off 2026: In the past 24 hours, I posted 3 new roles and received ~500 applications — all of which are hand-screened (no AI filters here...yes, we do exist! 🙋‍♀️) A few quick stats: • Each role requires candidates to reside in NY or TX, outlined at the very top of the description • Over half of the applicants were outside those states • Out of 500 applications, just 3 people messaged me directly on LinkedIn • 1 of those messages checked all the boxes (right location + strong experience) That candidate? 👉 Interview scheduled & completed 12 hours later 👉 In front of the hiring manager within 24 hours of reaching out to me The takeaway: ✅ Be specific and intentional when you apply ✅ Take a moment to review the role requirements and see how your experience aligns, you don't want to waste your time on roles you can't be considered for ✅ And yes — spending a few extra minutes to message the recruiter can put you miles ahead of the pack Not a guarantee, but it absolutely helps 💬🚀 Happy job hunting!

674

137 Comments

Great to see an uptick in Talent Acquisition roles across the tech industry — especially in tech recruiting. It’s no secret the past couple of years have been challenging for recruiters, but the market is shifting. Companies are building again, and it’s encouraging to see TA being brought back into focus as a key part of growth. For those still in the thick of it: stay positive — momentum is building, and your next opportunity is coming. #talentacquisition #recruiting #techhiring #recruiters #careergrowth

151

6 Comments

Thank you to Cassie Shouger, Head of TA for Roblox, who was our TLIX: Talent Leader Information Exchange Peer Mentorship Program guest speaker in September. Cassie dazzled us with demos of how Roblox is using AI for her recruiting team today. She also shared AI tools they’re evaluating for their team’s future use and monitoring for potential candidate use – including a new tool called Cluely, a hands-free AI teleprompter that listens to online interviews and provides suggested answers without any candidate typing required.  Equally helpful was her overview of the journey Roblox has been on to thoughtfully incorporate AI into their hiring process. Cassie shared that her team started with mapping the entire recruiting journey, then developing a point-of-view on which processes should be managed by automation or people by weighing the impact and lift. Nicely, due to Roblox’s internal tech capabilities, they also had the opportunity to consider build vs. buy scenarios. Cassie shared how her team partnered with legal along the way and noted that they’re actively navigating various global and regional AI regulations. Roblox is still on their AI adoption journey but so far, they’re enjoying significant recruiter and interviewer efficiencies, as well as increased candidate scheduling flexibility. Cassie confirmed that 100% of their hiring recommendations and decisions are made by people, not AI. We closed with a demo showing how Roblox is also using AI to assist employees with those laborious performance reviews. It was fascinating and likely a new trend we’ll see proliferating. TLIX: Talent Leader Information Exchange IX Communities #TAleadership #TAstrategy #RecruitingAI #FutureRecruiting

38

4 Comments

🚨 New H-1B Fee Alert: What Hiring Leaders in AI, ML & Data Need to Know A sharp shift just landed: the U.S. government now requires a $100,000 one-time fee for new H-1B petitions — up from the ~$2,000–$5,000 range. This change is already rippling through tech hiring, and it’s especially concerning for high-demand domains like AI, ML, and data. ⚠️ Major Challenges Ahead 1. Skyrocketing Sponsor Costs Every new technical hire requiring H-1B now carries a six-figure cost. That shifts the calculus for hiring decisions, especially for more junior or borderline roles. 2. Stricter Selectivity on Who Gets Sponsored Expect more scrutiny on “must-have” vs. “nice-to-have” roles. Jobs that aren’t clearly mission-critical or high-value may be deprioritized if the cost can’t be justified. 3. Uncertainty, Confusion & Risk Companies are still seeking clarity around how the fee is applied (which roles, exceptions, timing, renewals, etc.). That uncertainty slows hiring and adds risk to planning. 4. Disadvantages for Startups & Smaller Firms Deep-pocketed tech giants can more easily absorb a $100K barrier. Smaller companies, leaner R&D shops, or developing AI teams may struggle to compete for H-1B-based talent. 5. Shift to Alternative Talent Strategies Many will look harder at U.S. — born talent, remote-first models, international hiring without U.S. relocation, or alternative visa categories. The pressure is on to rethink traditional hiring pipelines. 💡 What Smart Hiring Leaders Can Do Now - Revisit Eligibility Criteria Be crystal clear about which roles must be sponsored. Align sponsorship only to positions tied closely to strategic AI/ML or data objectives. - Budget & Forecast Early Embed the new H-1B cost in your hiring models. If sponsorship is part of your talent strategy, you’ll need to justify it (or decline it) up front. - Explore Alternatives Look into other visa types (O-1, L-1, etc.), contractor / remote options, or staffing firms specializing in US talent. Don’t put all your eggs in the H-1B basket. - Watch for Exemptions / Policy Clarifications Monitor potential “National Interest Exceptions” or regulatory fine print. Legal counsel and immigration experts will become more critical than ever. - Double Down on Domestic Pipelines & Upskilling To reduce dependence on foreign sponsorship, invest more in U.S.-based talent — training, apprenticeships, transitions from adjacent roles. The bottom line? This isn’t just a budget change — it’s a structural shift in how competitive U.S. tech firms will source global AI/data talent going forward. 🗣️ Hiring leaders: are you already seeing changes in pipeline, candidate behavior, or internal decisions based on this new H-1B cost?

9

12 Comments

AWS is building Project Rainier, one of the largest AI superclusters ever constructed, to support Anthropic’s next-generation foundation model training. This infrastructure spans hundreds of thousands of Trainium2 accelerators across multiple U.S. regions, delivering unprecedented scale and performance. Project Rainier is designed to meet the growing demands of AI model training with: o 2.9TB/s memory bandwidth per chip o Under 10 microsecond network latency o Custom-built interconnects and cooling systems This isn’t just about raw computing—it’s about enabling innovation at a scale that was previously unimaginable. With Rainier, Anthropic can train a single model across the entire cluster, unlocking new frontiers in AI capability and efficiency. Read on:

17

1 Comment