Somewhere in the U.S., 3.5 million jobs are sitting open at a $103K starting salary - and employers can't find anyone to fill them. Not because the people don't exist. Because nobody's looking there.

Today we're covering that hiring gap, why AI is quietly adding $15-25 a month to your electricity bill whether you've ever used it or not, and the career moves that are actually working in a market that's frozen at the top and wide open underneath.

Let's get into it.

Cybersecurity has a hiring problem that runs in the opposite direction from everything else right now. There are an estimated 3.1 to 3.5 million unfilled cybersecurity roles globally, with the Bureau of Labor Statistics projecting 35% job growth for information security analysts through 2031.

Median starting pay sits at $103,700, with mid-level roles averaging $107,000–$130,000, and more than half of U.S. employers say they would raise starting compensation for candidates with in-demand skills. AI security, cloud security, and threat intelligence command the biggest premiums. The field has had more openings than candidates for years — and that gap is not closing.

Workers who changed jobs in January 2026 saw annualized pay growth of just 6.4%, compared to 4.5% for those who stayed — a gap of 1.9 percentage points, the smallest since ADP started tracking this data in 2017. At the peak of the Great Resignation in April 2022, that gap was 16 points. Fortune put it plainly: switching roles "nearly has the same payoff as staying loyal to an employer."

The cause is structural — job openings have nearly halved from their 2022 highs, and employers no longer need to outbid each other to hire externally. The job-hopping playbook that worked in 2021 is now producing diminishing returns for most people.

U.S. hiring jumped in February 2026, delivering the strongest month since November of last year, according to the Indeed Hiring Lab. Construction led the gains again, followed by education and health services. Pay for new hires rose to $19 per hour in January, lifted by accelerated hiring in construction and financial services.

The headline number keeps getting treated as a single story. It's actually two markets running in parallel — broad contraction in tech and white-collar, genuine expansion in trades, infrastructure, and skilled services — and where you sit determines almost everything about your next move.

Unlock a focused set of AI strategies built to streamline your work and maximize impact. This guide delivers the practical tactics and tools marketers need to start seeing results right away:

Stay ahead of the curve with these top strategies AI helped develop for marketers, built for real-world results.

Download the Free Report

AI is raising your electricity bill, whether you've ever used it or not

U.S. data centers consumed approximately 176 TWh of electricity in 2025, equal to 4.4% of all national power. The IEA projects that figure to reach 325–580 TWh by 2028 as AI workloads scale.

Goldman Sachs estimates that data center demand alone will push core inflation up 0.1% in both 2026 and 2027. The average U.S. household electricity bill is expected to rise $15–25 per month as utilities pass on grid upgrade costs.

Northern Virginia (home to the world's largest data center cluster) has Dominion Energy requesting emergency rate increases and fast-tracking new natural gas plants to keep up. The costs of AI aren't landing only on the companies building it. They're showing up in everyone's monthly expenses, including people who have never opened a chatbot.

Fujitsu partnered with Japanese hospital operator Genshukai to deploy AI across hospital management workflows: scheduling, documentation, and patient flow coordination. The measured results at a single facility were 400+ management hours saved and a 10% revenue uplift, equivalent to $1.4 million annually.

The AI handled administrative routing that previously required manual coordination across departments, and the deployment is now scaling to additional sites. Healthcare AI coverage tends to focus on diagnostics and drug discovery.

The administrative overhead story is less glamorous, but it affects every hospital system in the world, and the ROI numbers here are specific enough to take seriously.

Researchers at the University of Canterbury published findings this month showing that a machine-learning wildfire system consistently outperforms the standard Fire Behaviour Index across multiple locations, improving prediction accuracy by 10–30%. The model analyzes weather station data to detect pre-ignition patterns before fires start, and the research team included an economic valuation of what earlier warning is actually worth.

A separately funded NASA wildfire digital twin is now integrating satellite feeds, UAV data, and ground sensors to model fire spread in real time, with pilot deployments underway. U.S. wildfire losses have exceeded $150 billion in recent years.

A 10–30% improvement in prediction is not a footnote.

The traditional model (climb one ladder at one company toward one title) is giving way to something that looks more like a portfolio: a full-time role combined with advisory work, project consulting, writing, or teaching. Robert Half's 2026 workplace trends research found this pattern accelerating across all age groups, not just younger professionals.

The people executing this most effectively aren't abandoning stability. They're using it as the foundation and building adjacent income alongside it.

The side portfolio builds skills the day job doesn't, generates outside income, and creates leverage in negotiations. A single income attached to a single employer is a single point of failure.

More people are starting to treat it that way.

A new category of independent consultant has emerged: professionals charging $150–$300 per hour, or $5,000–$15,000 per project, to help small and mid-sized businesses implement AI workflows they can't figure out on their own. These aren't engineers - they're former operations managers, marketing directors, and consultants who learned the tools and packaged them as a service.

The client math is clean: a $10K engagement that saves 20 hours per week at $50/hour pays back in under two months. Supply is still far behind demand, because most people with the skills haven't yet built a practice around them.

That gap is the opportunity.

The hyperscale data centers being built across the U.S. require someone to construct and maintain them, and that someone is increasingly a licensed electrician or HVAC technician clearing $120,000–$150,000 with overtime. Northern Virginia has become one of the most competitive skilled trades markets in the country specifically because of data center density.

The buildout is projected to continue through at least 2028, driven by AI infrastructure investment that isn't sensitive to broader market conditions. The AI economy is generating tech jobs, yes.

It's also generating a specific category of high-paying trades jobs that most people aren't looking at, and these require credentials, not four-year degrees.

Paste your resume or a summary of your background into Claude or ChatGPT with this prompt: "Based on my background, which cybersecurity roles might I qualify for without starting from scratch? Map my domain knowledge and transferable skills to specific job titles, and tell me the single certification that would make me most competitive for each." Then run a LinkedIn search for those specific titles filtered to entry or mid-level and read the actual job requirements.

Healthcare, finance, legal, and operations backgrounds translate more directly than most people expect. Domain expertise is exactly what many cybersecurity teams say they can't find.

The point isn't to pivot tomorrow. It's to understand what the path looks like from where you're already standing.

📖 INTERESTING READS

The largest-ever review of medicinal cannabis found it doesn't treat anxiety, depression, or PTSD. Despite millions of people using cannabis specifically for these conditions, the most comprehensive evidence review to date found no reliable therapeutic effect for any of them. Worth reading before the next conversation where someone suggests it as a mental health solution. ScienceDaily (March 2026)

Ozempic's most surprising side effect might be a significant drop in depression and anxiety

A large-scale study found that GLP-1 medications produced measurable reductions in depression, anxiety, and psychiatric hospital visits - a mental health benefit nobody was looking for when the drugs were designed. The implications for how doctors think about prescribing these are significant. ScienceDaily (March 2026)

The pay premium for switching jobs is basically gone. Here's what Axios says to do instead. Short, data-backed, and directly relevant to anyone who's been waiting for the right moment to make a move. The calculus has changed and this piece doesn't sugarcoat it. Axios (Feb 18, 2026)