On the State of IT in 2025

I have been in IT since 2003. Long enough to have seen the field transform from a niche populated by people who genuinely could not imagine doing anything else, into an industry that attracted everyone seeking high compensation and flexible hours — and is now in the process of correcting for that. What follows is my attempt to describe honestly what is happening, why it happened, and where it goes from here.

1. The Flood of Inexperienced Specialists

Over the past several years, enormous numbers of newly-minted developers, analysts, testers, and product managers entered the market. Many have limited genuine experience and questionable output quality. They arrived with high salary expectations, a preference for 2–3 hour effective work days, requirements for premium office amenities, and strong resistance to in-person interviews (which make it harder to cheat on technical assessments).

The phrase I hear often: "I'm already 1.5 years in, I'm practically a mid-level." In any other skilled trade, 1.5 years would make you a junior who is still making foundational mistakes. In IT, a cohort of people decided it was enough to command senior compensation.

What happens next: Companies are already tightening screening. In-person interviews are increasing. For remote candidates in cities where a company has no office, organizations are beginning to arrange in-person assessments with senior staff who live nearby — cheaper and more reliable than five to ten rounds of video calls where it is impossible to verify that the person answering the questions is the person who will show up for work.

2. Job Postings That Demand Supermen

The flip side of the oversupply problem: job postings that list requirements no single human could satisfy, for roles that functionally need a capable junior. Candidates respond by copying the job description language directly into their CV. Neither side is communicating honestly, and the hiring process devolves into a keyword-matching exercise that benefits no one.

This is not new. Twenty years ago, students faced the same paradox: entry-level roles requiring three or more years of experience. The solutions then were: teach at computer courses, fix neighbors' computers, run cable at a small company, or join an organization at the bottom and grow internally. Those paths still exist and still work. The difference now is that a generation of candidates has been told they can skip them.

What happens next: Companies will rediscover that growing talent internally is more effective than trying to hire ready-made senior staff from the market. The best entry point into IT remains technical support.

3. Resume Fraud (Not New, But Worse)

Fabricated credentials and fake references have always existed. What has changed is scale and social acceptability. I know of cases where a colleague's spouse completed their technical assignments during probation. I know of a business executive who claimed personal credit for projects an entire engineering team delivered.

Organizations are slow to acknowledge this because doing so would require admitting they paid inflated salaries to underperformers for years — which reflects poorly on the managers who hired and retained them. It is easier to quietly let someone go and say nothing.

What happens next: Reference checks will intensify. Three to five references instead of one or two; cross-verification of consistency across contacts; multi-year social media background checks. This reduces the problem; it does not eliminate it.

4. Experienced Professionals Caught in the Middle

Here is the genuinely painful part. Developers with 15 to 20 years of experience are struggling. They spent years building deep expertise. They hold themselves to exacting standards — zero tolerance for their own mistakes, constant self-evaluation. They watched the market inflate around them, with juniors earning comparable or higher salaries. And now they face age discrimination: companies are reluctant to hire anyone over 40, let alone 45.

The result is a cohort of highly competent professionals who feel professionally stranded. They know that leaving their current role means a long and demoralizing job search. They stay, and they feel the slow accumulation of that awareness every day.

If you are a spouse watching someone you love go through this — the weight loss, the difficulty sleeping, the flat affect that looks like depression — it is worth knowing that this is a specific professional crisis with a specific cause. It is not about material dissatisfaction. It is about someone who has invested decades in mastering a craft watching the market signal that mastery does not matter. That passes. The expertise does not disappear, and eventually the organizations that need it recognize what they have.

What happens next: The market will correct. Companies that optimize purely for cost discover, expensively, that quality problems compound. The professionals who remained rigorous during the hype cycle become the most valuable people in the room.

5. HR Under Pressure from Both Directions

Recruiting staff are in an impossible position. They are criticized for being too selective while simultaneously dealing with hiring managers who cannot articulate what they actually need. The intake of candidates includes people who genuinely reinvented themselves (a former construction worker who learned Python over two years and can outperform many CS graduates) alongside people who have fraudulent credentials, coached interview answers, and no underlying skill.

Differentiating between these cases is extremely difficult without deep technical knowledge that HR generalists do not have. Technical interview panels help but do not scale.

What happens next: AI-assisted screening will provide partial relief. It will not solve the problem but will allow HR teams to handle higher volume with better consistency.

6. Leadership That Does Not Understand What IT Is

Outside major metropolitan areas, many executives still regard all IT staff as "the computer guys who fix the printers." They resist remote work on principle, find agile methodology suspicious, and are baffled by the cultural norms of technical teams. They make IT decisions with the confidence of someone who does not know what they do not know.

This is not a small problem. An organization whose leadership cannot distinguish between strategic and commodity IT work, or between a talented engineer and a mediocre one, cannot make good decisions about either. It will overpay for commodity work, underinvest in strategic capability, and be surprised when the results are poor.

What happens next: Basic IT literacy needs to be treated as a leadership competency, not an optional extra. Operating in 2025 without understanding information technology is analogous to not knowing how to use a bank account.

7. Speed at the Expense of Everything Else

Development velocity has increased. Code quality has decreased. The norm in many organizations is: ship fast, hand off to support, move on. Technical debt accumulates. The support team receives the consequences of decisions made by developers who have long since moved to other projects.

The romanticized idea of "moving fast" has a downstream cost that does not appear in the sprint velocity metrics. Legacy system volume grows exponentially. Every shortcut taken today is maintenance work for someone tomorrow — and that someone is usually a less experienced developer who does not know why the original decision was made.

"1.5 years" or even "3 years" as a metric of seniority is meaningless. Genuine seniority — the kind that produces architectural decisions that age well — requires 7 to 10 years, enough time to accumulate a full cycle of mistakes, consequences, corrections, and hard-won judgment. The 10,000-hour principle applies here without exception.

What happens next: Legacy system volumes will multiply. The cleanup work will fall to the experienced professionals discussed above. This is, ultimately, part of how the market re-establishes the value of depth.

8. The Clubhouse Effect

IT's current trajectory closely mirrors what happened to Clubhouse, the social audio platform. In its early days, Clubhouse was exclusive, interesting, and populated by people who genuinely cared about the medium. When access opened to everyone, the people who made it valuable left, the atmosphere evaporated, and the platform became irrelevant within months.

IT made a similar transition: from a specialized field populated by people who were passionate about it, to an industry that marketed itself as accessible to anyone willing to spend three months at a bootcamp. Many of those people are talented and will succeed. Many will not. The economic principle at work is simple: when the supply of people capable of doing your job increases dramatically, the price of that labor decreases. Taxi drivers, delivery couriers, call center operators all perform difficult, essential work. They earn modestly because the labor pool is large.

Unlike Clubhouse, IT will not disappear. But the market will expel participants who are not genuinely committed. The people who remain — the ones who love the work and cannot imagine doing something else — will spend years cleaning up code written by people who moved on when the salaries stopped being remarkable.

How This Happened: Three Turning Points

The 2008 Financial Crisis. Before 2008, large IT budgets funded multi-year waterfall projects that produced software nobody ever saw. The crisis forced business owners to demand accountability. Agile methodology emerged naturally from this pressure — not as a philosophy, but as a survival response. Value had to be delivered quickly and visibly.

The Innovation Center Era. Companies competed for talent by creating environments optimized for creative professionals: game rooms, smoothie bars, no dress codes, mental health support, flexible hours. The intent was to attract genuinely excellent people by removing friction. It worked — and then it attracted people who came for the amenities rather than the work. The distinction is hard to detect in an interview process designed to be welcoming.

The Pandemic and the Salary Wars. Remote work demand exploded. Some companies responded by offering three to five times market salary for any candidate who could pass a phone screen, reasoning that it was cheaper to overpay during probation and cull aggressively than to interview carefully. The candidates who did not survive those probation periods entered the broader market expecting equivalent compensation. Bootcamps that had once placed graduates into appropriate junior roles now placed them into inflated markets where inappropriate salary expectations became normalized.

What Should Change

The most effective intervention I can suggest is one that has been out of fashion for twenty years: mandatory time in technical support before moving into development roles.

Not as punishment. As education. A developer who has spent three months answering support tickets understands what broken software costs real people. They know what "it works on my machine" means to a user who cannot submit an expense report before a deadline. They have heard, directly, the frustration of someone who cannot do their job because of a decision made in a sprint planning meeting.

This experience changes how people write code. It changes what they prioritize. It filters out people who are not genuinely interested in the work — because support is unglamorous, and people who came to IT for status and salary will not stay through it. The ones who do are usually worth keeping.

For anyone considering entering the field: if you genuinely love it, come in. The path is longer and harder than the bootcamp marketing suggests, but the work is real and the career is durable. If you are here because IT salaries looked attractive and the hours seemed flexible — those conditions no longer reliably apply, and the window is closing further.

On AI and the Longer View

AI coding tools are changing what junior work looks like, faster than most organizations have processed. Within a few years, the language of programming will effectively be English for many tasks. What will not be replaceable is the engineer who understands enough fundamentals to recognize when the AI is generating plausible-looking garbage — who can read the output, identify the subtle wrong assumption, and know why it matters.

That engineer needs years of accumulated experience, not months. They need to have written code that broke production and lived with the consequences. They need to have read other people's terrible code and understood, from experience, what made it terrible.

The hype cycle selects against these people in the short term. The long term belongs to them.