I Love Tech. I Hate the Tech Industry.

Oct 6, 2026 · By  Olimi Emmanuel K

Six years. Two degrees. A MicroMasters from the Massachusetts Institute of Technology. Real projects, real code, real nights spent debugging. And in all that time, not one job offer, not one contract, not one paid gig from any of it.

People hear "master's degree" and "New York" and picture someone who made it. They tell me I must be smart. They say it with real admiration. I have learned that admiration does not pay rent.

Here is the confusing part: I love technology. I have loved it since I was a kid pulling radios and old motherboards apart to see what was inside. What I have come to hate is the industry built around it, the one that tells the whole world to come in, learn, and be rewarded, and then quietly keeps the door for a much smaller group.

This is not a polite article. It is what six years of silence looks like when you stop blaming yourself long enough to read the data. Some numbers made me angrier. A few made me change my mind about things I was sure of. I will show you all of it, including the parts that do not flatter my argument, because an argument that cannot survive the facts is not worth publishing.

So here is the question I cannot shake: if this industry is really open to anyone, why does it feel like a door painted on a wall?

Let's start where the lie begins: the promise.

The promise: learn to code and you are set

For twenty years the advice came from everywhere. Parents said it. Teachers said it. Politicians and billboards said it. Learn to code. Computer science was the safe bet, the golden ticket, the one degree that could not fail you.

Kids believed it, and they did exactly what they were told. Between 2010-11 and 2021-22, American colleges went from awarding about 43,000 bachelor's degrees in computer and information sciences to about 108,500. By 2023-24 the figure was roughly 122,000, according to federal data compiled by College Transitions.

Then the music stopped. In fall 2025, undergraduate computer science enrollment at four-year colleges fell 8.1 percent, the steepest one-year drop of any major since at least 2020, reports VnExpress, citing the National Student Clearinghouse Research Center. A Stanford Hoover Institution economist found that computer science's share of enrollment fell for the first time in about two decades.

So what did the golden ticket buy the people who bought it? The Federal Reserve Bank of New York tracks this. Its latest by-major table puts recent computer science graduates (ages 22 to 27) at about 7.0 percent unemployment and computer engineering graduates at about 7.8 percent, among the highest of any major, in the same range as anthropology and fine arts. Its headline number for all recent graduates was 5.6 percent in the second quarter of 2026, with 42 percent underemployed.

Now, fairness, because I promised it. The work itself is not disappearing. The U.S. Bureau of Labor Statistics projects employment of software developers, quality assurance analysts and testers to grow 15 percent from 2024 to 2034, with about 129,200 openings a year (BLS).

The occupation has a strong runway. The major has a brutal on-ramp. Both are true at once, and that gap, between the jobs that exist and the people who can actually get one, is the whole story of this article.

I did not understand that gap when I walked into it. Let me tell you how I got there.

My story: the kid who loved the hardware

I was never the kid who fell in love with code. I was the kid who took radios apart. Old motherboards, gadgets, anything with a circuit in it. The hardware was the magic. The software I learned because I had to.

I studied Information Systems at Makerere University in Uganda and graduated in 2014. I built websites. I ran a firm that was supposed to be a technology company, and it ended up earning more from branding and general supplies than from software. I never saw tech as a cash cow. It was something I loved and had never figured out how to turn into a living.

Then 2020 arrived and AI went from a whisper to a roar. I went back in, hard. Python. Backend. Data. Machine learning. A MicroMasters in Statistics and Data Science from the Massachusetts Institute of Technology in 2022. Then a master's in Data Analytics and Visualization at Yeshiva University in New York, which I finished in May 2026. I built things along the way, including an offline-first AI education platform designed for places where the internet cannot be taken for granted.

Six years. Two degrees, plus the MicroMasters. Not one job offer. Not one contract. Not one paid gig from any of it.

I am applying heavily right now, and some days I feel exactly like I did in 2010: a person who knows nothing, standing outside a building, waiting for someone to open a door. That is a strange feeling after six years of climbing. You can see how far you have come. You just cannot get anyone to count it.

At first I assumed the problem was me. That is what the industry trains you to assume. So I did the one thing I know how to do. I went and got the data.

The ladder that does not exist

Try this thought experiment. Name any serious field, then ask: what happens when you walk in at the bottom?

Medicine has a system waiting. In the 2026 residency Match, 93.5 percent of U.S. MD seniors landed a paid training position, out of 44,344 positions offered nationwide.

The skilled trades have apprenticeships, which are jobs from day one. The U.S. Department of Labor says 91 percent of apprenticeship graduates retain employment, and its newer figures run higher.

I am not claiming these worlds are easy or fair. Medicine has brutal gates too: foreign-born graduates who need visa sponsorship matched at just 54.4 percent in 2026. Acting and farming break people every year. But in each of them you can see where the first rung is, who controls it, and what the rules are.

Tech has no match. No licensing exam. No apprenticeship standard. No agreed meaning for the word "junior." And yet the sales pitch says anyone can walk in, at any level. A one-month certificate and a master's degree enter the same funnel and fight for the same few seats. In medicine, the seat exists. In tech, the seat is a rumor.

The data says the rumor is getting thinner. Venture firm SignalFire, which tracks hiring at the 15 largest tech companies, found new graduates are now just 7 percent of Big Tech hires, down from 15 percent before the pandemic. Stanford economists using payroll data found employment for software developers aged 22 to 25 fell nearly 20 percent from its late-2022 peak while older developers' jobs held steady or grew. The lead author cautions the study cannot prove AI is the cause. It does show who gets squeezed out first: the person with no track record yet.

That is the catch-22 in one sentence: you need experience to get the job, and you need the job to get experience.

So if the bottom rung is gone, who is the door actually open for? Look at who is already inside.

The first gate: who you know, and who screens you

Here is a claim I hear all the time, and one I have repeated myself: that most tech recruiters are Indian, so the very first gate is closed to everyone who is not. I went looking for proof.

I could not find it. The Bureau of Labor Statistics counted about 980,000 human resources workers in 2023, the category that includes recruiters. About 75 percent were white, 14.7 percent Black and 6.0 percent Asian, against 6.9 percent Asian across all U.S. workers. HR is not where the picture is lopsided. I would rather lose an argument than win it with a number that is not true.

One caveat: that category cannot isolate IT staffing agencies, and I found no published data on them. What I have there is experience, not evidence.

Where the data does match my experience is further down the hallway. The same BLS table shows 36.2 percent of software developers are Asian and 6.5 percent are Black, in a workforce that is 6.9 percent Asian and 12.8 percent Black overall. "Asian" bundles many origins, and BLS does not break out Indian workers. But it tells you what the room looks like when the technical interview begins: the people deciding whether you "fit" mostly do not look like me.

There is also a pipeline behind that room. Indian nationals received 71 percent of approved H-1B petitions in fiscal 2024 (283,397 of 399,395, renewals included), and Infosys and Tata Consultancy Services sat among the top ten H-1B employers in 2023. Pipelines become networks. That is not a conspiracy. It is how hiring works. And networks are exactly how this industry decides who gets a look.

When internships drew roughly 2.5 times their usual number of applicants in 2025, a referral stopped being a favor and became a ration ticket.

And networks do not pay out equally. Sociologists David Pedulla and Devah Pager tracked more than 2,000 job seekers for 18 months. Black and white applicants used their networks at similar rates, yet networks produced fewer offers for Black applicants. White applicants knew someone inside the company on 65.2 percent of their network-based applications, versus 56.3 percent for Black applicants. Their contacts reached out to the employer on their behalf 25.4 percent of the time, versus 20.0 percent (summary).

People tell me stories of someone with a two-week certificate and the right introduction beating a candidate with a master's degree and years of work. I cannot verify any single story, and nobody has run that experiment. But this research shows how such a story is possible, and who is most likely to land on the wrong side of it.

Now the harder question. What happens when there is no introduction, and your résumé has to speak for itself?

The shade and the accent

Let me say the uncomfortable part plainly. I am a Black man from Uganda trying to build a technical career in the United States. I cannot prove why any single application failed, because a rejection email never says why. But I can read what researchers have measured, and it is not subtle.

The ladder narrows as you climb. In 2022, the U.S. Equal Employment Opportunity Commission found Black workers were 7.4 percent of the high-tech workforce, against 11.6 percent of all U.S. workers, and only 5.7 percent of high-tech managers. Their share was 6 percent in 2005. Inside the high-tech sector, white employees were 63 percent of all staff but 78 percent of senior officials. Black employees were 8 percent of staff and 3 percent of senior officials.

The agency's conclusion was careful but clear: the size of the gaps, together with other research, suggests discrimination contributes.

Notice what the same report says about Asian workers, who hold 18.1 percent of tech jobs against 6.5 percent of the workforce. Even they slip to 15.3 percent of managers. The top of this industry is not owned by any one immigrant group. It is mostly white, and it gets whiter the higher you look.

 


EEOC, High Tech, Low Inclusion (2024), 2022 American Community Survey data. Index = a group's share of high-tech workers (or managers) divided by its share of all U.S. workers.

Same résumé, different callback. A meta-analysis of 28 field experiments covering 55,842 applications found white applicants received 36 percent more callbacks than equally qualified Black applicants, and that this gap did not change over 25 years. In another study, researchers sent 1,600 fabricated résumés to employers in 16 U.S. metro areas. Black applicants who scrubbed racial signals from their résumés got callbacks 25.5 percent of the time, versus 10 percent when they left them in. Asian applicants went from 11.5 percent to 21 percent. Employers that advertised diversity did no better.

That second finding matters. It means this is not one group gaining at another's expense. It is a system that rewards whatever looks and sounds familiar.

Kang, DeCelles, Tilcsik and Jun, Whitened Résumés, Administrative Science Quarterly (2016); results as reported by Harvard Business School.

Now the machines. University of Washington researchers tested three AI résumé-ranking models on 550-plus real résumés with 120 swapped first names. White-associated names were favored 85 percent of the time. The models never preferred a Black male name over a white male name. It was a lab test of open-source models, not an audit of any employer's software, but companies are adopting these tools fast.

And the accent? You cannot audit an accent with a fake résumé. It shows up in the phone screen and the interview, in rooms nobody measures. I believe it matters. I cannot give you a number, and I will not invent one. That is part of the problem: the test I fear most leaves no record.

None of this proves anyone is rigging the game on purpose. It proves the game is tilted, and tilted games do not need villains.

There is one more number the recruiting ads never mention: your age.

Too young, too old, never just right

I assumed experience would be my friend. In most industries, older means wiser, and wiser means better odds. Tech runs a different calculation.

The EEOC found that 25-to-39-year-olds are 40.8 percent of the high-tech workforce but 33.1 percent of all U.S. workers. The share of tech workers over 40 fell from 55.9 percent in 2014 to 52.1 percent in 2022, slipping below the 53.1 percent in the workforce as a whole. And age claims made up 19.8 percent of EEOC discrimination charges against high-tech employers, versus 14.8 percent in other sectors, a statistically significant gap.

At the giants, the numbers get starker. Companies do not publish the average age of their staff, so the best public comparison is a decade-old PayScale survey of self-reported ages, reported by USA Today in 2017. Treat it as a snapshot, not a census.

PayScale survey data reported by USA Today (2017), via EconDataUS. Self-reported ages, not company disclosures; treat as indicative.

But the squeeze runs both ways. Compensation firm Pave found that Gen Z's share of employees at large public tech companies fell from 15 percent to just under 7 percent between January 2023 and August 2025, pushing the average age of those workforces from 34.3 to 39.4. SignalFire found the average age of technical hires rose by three years since 2021. Meanwhile, a former Meta senior director has sued, alleging company data showed workers 40 and older were 1.5 times as likely, and those 50 and older 2.5 times as likely, to be cut in Meta's February 2025 layoffs. Those are allegations, not findings.

Put it together and you see the trap. Too inexperienced to be hired as a seasoned engineer. Too far from 22 to be the cheap entry-level hire the industry used to love. And if you are changing careers, you are standing in exactly that gap.

So where did all the money go? These companies are richer than any in history.

The richest industry on earth, and the thinnest door

I used to think: these companies are so wealthy, surely they can afford to take people in. Then I checked the filings.

In 2025, Alphabet reported $402.8 billion in revenue and 190,820 employees. Meta reported $201.0 billion and 78,865 employees. Together that is about $603.8 billion and 269,685 people. For its fiscal year ending January 2026, Walmart reported $713.2 billion in revenue and about 2.1 million associates. That is 18 percent more revenue than Alphabet and Meta combined, spread across 7.8 times as many people.

To be fair, tech does hire. Alphabet's headcount roughly tripled between 2015 and 2025. But its revenue per employee still climbed by about three-quarters over those years. Between 2023 and 2025, Alphabet's revenue rose 31 percent while its headcount rose under 5 percent.

Alphabet revenue and year-end headcount from company filings, history via TickerLeague. Meta (2025) and Walmart (fiscal 2026) from their annual reports.

Meanwhile, layoffs.fyi has tracked more than 700,000 tech layoffs since 2022 (the 2023 total alone was 262,735). And the money is going somewhere else. Alphabet says it spent $105.7 billion on capital expenditures in 2025 and expects $175 billion to $185 billion in 2026. Meta spent $72.2 billion. Capital spending means chips, servers and buildings, not first-year engineers.

In a factory, more revenue means more workers. In software, more revenue can mean the same workers and more servers. That is the economics nobody printed in the recruitment brochure.

So the door is narrow, the room inside is clustered, and now a new problem is making that door even harder for honest people to get through.

Now even the real candidates are suspects

A newer problem has joined the old ones: hiring fraud.

In June 2025, the U.S. Justice Department announced charges in a scheme in which North Korean IT workers and their facilitators compromised the identities of more than 80 Americans to land remote jobs at more than 100 U.S. companies, many of them Fortune 500 firms. Prosecutors said that one scheme alone generated over $5 million. Research firm Gartner predicts that by 2028, one in four candidate profiles worldwide could be fraudulent in some way, as AI makes fake faces, voices and credentials cheap.

The victims are the companies, the Americans whose identities were stolen, and, quietly, every honest applicant behind the fakes in the queue. Employers respond the only way they know how: more verification, more suspicion, more reliance on people already vouched for.

I cannot show you data on that last part. But it is not hard to guess where a trust-starved employer turns. It turns to the candidate somebody it knows has already vouched for.

And we have already seen who tends to have the most of those.

Six years of this does something to a person. Let me tell you what.

What six years of silence does to you

The money cost is easy to count. Tuition, equipment, subscriptions, and years when I could have been earning somewhere else. The other cost does not fit on a spreadsheet.

For six years I have had a quiet voice in my head asking whether I am simply not good enough. It is a cruel voice, because the industry feeds it. In most jobs, you master your tools and relax. In tech, there is always another framework, another cloud platform, another AI tool you have not touched. Everyone around you seems to know something you do not. That feeling has a name, the impostor phenomenon. Tech did not invent it. Psychologists Pauline Rose Clance and Suzanne Imes described it in a 1978 paper on high-achieving women. But an industry that never stops changing is rich soil for it.

And for some of us, the industry adds its own fertilizer. The Kapor Center's Tech Leavers study, cited in the EEOC's report, found that nearly a quarter of underrepresented men and women of color in tech had experienced stereotyping, twice the rate of white and Asian workers.

Here is what I am slowly learning. When you send out hundreds of applications and hear nothing, the silence is not information about your worth. It is information about a funnel. The funnel has a collapsed bottom rung, a clustered network, a tilted screen, and a growing suspicion of strangers. You can improve your work every day and still be shut out by all four.

That is not an excuse. I keep building. But I have stopped treating the market's silence as a verdict on me.

Which leaves one question: what do we do with all this?

What I still believe, and what I am asking for

I still love technology. The radio on the bench, the model on the screen. I will keep building, including tools for places where the internet is a luxury. The craft was never the problem. The doorway is.

So here is what I am asking for.

From employers

• Real junior roles, with requirements that match the title, and paid, supervised ways to learn on the job. The apprenticeship model works: the Department of Labor reports 91 percent of apprenticeship graduates keep their jobs.

• Name-blind first screens, and independent audits of any AI tool that ranks candidates.

• Published data on age and race at every level and every hiring stage, not just company-wide totals. The EEOC has already told the industry to identify and address the barriers. Ask what they found.

From schools and bootcamps

• Outcomes before promises. How many graduates had paid, relevant work within a year, and how many already had connections or experience?

For anyone considering this road

• Study the path into paid work as hard as the curriculum. Ask who was hired with your starting point last year.

• Treat relationships as a skill. One real conversation a week with someone inside a company you want. The research says that is where the offers come from.

• Look beyond the famous names. The EEOC found Black workers were 8.1 percent of high-tech workers outside the high-tech sector, versus 6.4 percent inside it. Pay is lower outside ($89,500 median versus $104,500), but the door may be easier to find.

• Describe projects as problem, user, result. Not a list of tools.

For parents

• Before you tell a child to learn to code, ask what happened to last year's graduates.

I love technology. I hate the industry that gatekeeps it. Both are true, and the second does not make the first a mistake.

If this made you angry, good. Anger is information that something is out of proportion. Send it to the next person who says "just learn to code."

Sources

Every figure keeps the period it was reported for. Ratios and percentage changes are my arithmetic from the cited numbers. The personal account is mine; the research describes broad patterns and cannot say why any one application failed.

Government and official data

1. U.S. Equal Employment Opportunity Commission, High Tech, Low Inclusion: Diversity in the High Tech Workforce and Sector 2014-2022 (2024)

2. Bureau of Labor Statistics, Employed persons by detailed occupation, sex, race, and ethnicity, 2023 annual averages

3. Bureau of Labor Statistics, Software Developers, Quality Assurance Analysts, and Testers

4. Federal Reserve Bank of New York, The Labor Market for Recent College Graduates; by-major figures as reported by SSTI

5. National Resident Matching Program, 2026 Main Residency Match results; IMG rates via The Derm Digest

6. U.S. Department of Labor, Registered Apprenticeship outcomes

7. USCIS, FY 2024 H-1B approvals by country of birth, as charted by Datawrapper; top employers per Asia Samachar

Company filings

1. Alphabet, 2025 Form 10-K; Q4 2025 earnings release; headcount history via TickerLeague

2. Meta, 2025 Form 10-K and Q4 2025 results

3. Walmart, fiscal 2026 annual report

Research and reporting

1. SignalFire, State of Tech Talent Report 2025

2. Brynjolfsson, Chandar and Chen, "Canaries in the Coal Mine?" (Stanford Digital Economy Lab, 2025), via HR Executive and Fast Company

3. Quillian, Pager, Hexel and Midtbøen, Meta-analysis of field experiments shows no change in racial discrimination in hiring over time, PNAS (2017)

4. Kang, DeCelles, Tilcsik and Jun, Whitened Résumés, Administrative Science Quarterly (2016), via Harvard Business School and University of Toronto

5. Pedulla and Pager, Race and Networks in the Job Search Process, American Sociological Review (2019), via the American Sociological Association and a research summary

6. University of Washington, AI tools show biases in ranking job applicants' names (2024)

7. PayScale ages via USA Today, reported by EconDataUS (2017); Pave data via LeadDev; Meta lawsuit via People Matters

8. San Francisco Standard on Glassdoor internship applications and SignalFire hire ages

9. Layoffs.fyi totals via NerdWallet and TechCrunch

10. Computer science degrees and enrollment: College Transitions and VnExpress

11. Justice Department North Korean IT worker case, via Staffing Industry Analysts; Gartner forecast via Epstein Becker Green

12. Clance and Imes, The Imposter Phenomenon in High Achieving Women (1978)