But what about all the tech layoffs?

The layoffs are real and worth reading precisely rather than fearfully. What they are: large companies cutting routine and duplicated roles - AI was the #1 named reason in March 2026 (~25% of announced cuts), concentrated in exactly the work AI now does. What they are not: evidence that companies stopped needing software people - the same companies doing the cutting are simultaneously hiring AI-capable engineers at premiums, and the broader economy (banks, telcos, retailers, government-linked companies in Malaysia) keeps digitising regardless. The pattern is a workforce SWAP, not a workforce shrink: out with the profile AI replaces, in with the profile that directs AI. Headlines report the "out"; payrolls show the "in."

Would it have been better to get into tech 5 years ago?

For the old profile, yes - 2019 was a better year to become a routine-implementation junior than 2026 is, and pretending otherwise would be dishonest. But for the new profile, the honest answer flips: 2026 is arguably the best entry window in a decade, for three reasons. The premium exists BECAUSE it’s early - AI-capable juniors out-earn traditional ones by roughly 40-60% only while supply lags, and premiums like that historically close as markets catch up. The learning path is faster than it’s ever been - AI tutoring compresses what took years into 400-600 focused hours. And the incumbents’ head start is unusually small: everyone, including 15-year veterans, has only had a few years with these tools. You missed the old boat; you’re early for the new one.

How long does it take to get job-ready for tech in 2026?

The honest number we’ve measured across Sigmaschool graduates: 400-600 focused hours from zero to job-ready - roughly 3-4 months full-time or 6-9 months alongside a job. "Job-ready" means the profile employers currently pay premiums for: fundamentals solid enough to judge AI output, fluency directing AI tools, and 2-3 real deployed projects you can defend in an interview. Beware both distortions of this number: the "learn to code in 3 weeks" pitch (produces the vibe-coder profile that fails technical interviews) and the "you need a 4-year degree" gatekeeping (contradicted by our own hiring outcomes - several graduates hold developer roles with no degree, hired on demonstrated ability).

How do I know if tech is right for me before committing?

Run a cheap real test instead of an expensive imagined one. The free trial exists for exactly this decision: one signup, no card, and you get the 6 Projects in 6 Days precourse on Sigmo (build real things immediately - the fastest way to learn whether building energises or drains you), a live Buildroom session with an instructor (see how the actual teaching works), and the full programme e-booklet (see the whole path before paying anything). A week of real building tells you more than a month of reading articles like this one. If the projects frustrate you in the "I want to solve this" way, that’s the signal - that frustration is the job, and some people love it.

All posts

Careers · AI & your future · Getting started

Is now a good time to get into tech? One door closed. One opened wider.

You’re holding two contradictory feeds: endless layoff headlines saying stay away, and endless AI-opportunity posts saying jump in. Both are describing real data about different doors - and the whole decision comes down to knowing which door is which.A coding school answering “should I get into tech” has an obvious interest, so everything below is a number you can check.

Deric Yee Updated 25 August 2026 9 min read

The door that closed - the headlines are right

Start by conceding everything the scary feed says, because it’s true. Traditional entry-level tech - the junior role whose job was routine implementation - is in genuine decline.Entry-level postings fell 15% in the US and 29% globally in a year. Goldman Sachs tracked around 16,000 US jobs a month being cut with AI cited as a factor; by March 2026, AI was the #1 named layoff reason, behind roughly a quarter of announced cuts. Stanford research puts employment for 22–25-year-olds in AI-exposed roles down 13%. If your plan was the 2019 plan - learn syntax, get hired to write CRUD code under supervision, grow from there - that ladder’s bottom rungs have been sawn off, and anyone selling you that plan in 2026 is selling a stale map. We’ve compiled the full dataset in the State of AI Hiring report.

The door that opened - the payrolls are also right

Now the half the scary feed omits. The same employers cutting routine roles are paying premiums they’ve never paid before for a different junior profile: people who direct AI tools fluently, judge their output against real fundamentals, and ship working software. In Malaysia that profile starts at RM 6,000–9,000/month against RM 3,500–6,500 for the traditional profile- a 40–60% premium that exists for the least mysterious reason in economics: demand outrunning supply. And the demand side is broader than “tech companies” - banks, telcos, retailers, logistics firms, and government-linked companies across Malaysia and Southeast Asia are all digitising and all short of people who can build with the new tools. The layoff wave and the hiring premium are one event seen from two sides: a workforce swap. Companies are trading the profile AI replaces for the profile that directs AI. Headlines cover the trade-out; the salary guide documents the trade-in.

Why the window is now, not “someday”

Premiums close - that’s what premiums do

A 40–60% junior premium exists only while supply lags demand. Every early-mover market in history - spreadsheet-fluent analysts in the early 90s, web developers in the late 90s, mobile developers after 2008 - paid its biggest premiums to the first wave and normalised as training pipelines caught up. The AI-capable premium is currently in its wide phase. Entering during the wide phase is the entire meaning of “good timing.”

The incumbents’ head start has never been smaller

Normally a career-switcher at 28 competes against people with a 10-year head start. But on the skills employers now pay premiums for - directing AI agents, evaluating model output, building AI-powered products - everyone started within the last few years, veterans included. The playing field re-levelled in a way it does roughly once a generation. It will tilt again as the era matures.

The learning path itself got faster

The same AI that changed the job compressed the training: a 24/7 tutor that explains your exact confusion, on your exact code, at 2am. The honest number across our graduates is 400–600 focused hours to job-ready - months, not years - and several hold developer roles with no degree, hired on demonstrated ability. The route in is shorter at precisely the moment the destination pays more.

The catch, stated plainly

None of this works if you build the wrong profile - and the wrong profile is the easy one. The open door does not admit vibe coders: people who can prompt an AI into producing an app but can’t read, debug, or defend what it produced fail at exactly the interview stage where employers test judgement - and employers in 2026 test hard for it, because they’ve been burned. The premium profile is fundamentals plus AI fluency, never fluency alone; the full argument is in vibe coding vs fundamentals, and the stage-by-stage route to the right profile is in how to become an AI-native developer. So: is now a good time to get into tech? The old door is closed - grieve it briefly if you must. The new door is wide, measurably better-paying, and admitting people who prepare properly. And finding out whether it’s your door costs nothing: the free trial - one signup, no card - gives you six real projects in six days, a live instructor session, and the full programme e-booklet. A week of real building answers this article’s question better than the article can.

FAQ

Is now a good time to get into tech?
Yes - for one specific profile, and the data is unusually clear about which. The door that closed: traditional entry-level roles doing routine implementation work, where postings fell 15% in the US and 29% globally in a year, and Stanford research shows employment for 22-25-year-olds in AI-exposed roles down 13%. The door that opened wider: AI-capable builders - people who can direct AI tools, judge their output, and ship real software - who command RM 6,000-9,000/month as juniors in Malaysia against RM 3,500-6,500 for traditional profiles, because demand for that profile outruns supply. So "should I get into tech" is the wrong resolution; the right question is "which tech profile am I building?" - and the open-door profile is learnable in months.
But what about all the tech layoffs?
The layoffs are real and worth reading precisely rather than fearfully. What they are: large companies cutting routine and duplicated roles - AI was the #1 named reason in March 2026 (~25% of announced cuts), concentrated in exactly the work AI now does. What they are not: evidence that companies stopped needing software people - the same companies doing the cutting are simultaneously hiring AI-capable engineers at premiums, and the broader economy (banks, telcos, retailers, government-linked companies in Malaysia) keeps digitising regardless. The pattern is a workforce SWAP, not a workforce shrink: out with the profile AI replaces, in with the profile that directs AI. Headlines report the "out"; payrolls show the "in."
Would it have been better to get into tech 5 years ago?
For the old profile, yes - 2019 was a better year to become a routine-implementation junior than 2026 is, and pretending otherwise would be dishonest. But for the new profile, the honest answer flips: 2026 is arguably the best entry window in a decade, for three reasons. The premium exists BECAUSE it’s early - AI-capable juniors out-earn traditional ones by roughly 40-60% only while supply lags, and premiums like that historically close as markets catch up. The learning path is faster than it’s ever been - AI tutoring compresses what took years into 400-600 focused hours. And the incumbents’ head start is unusually small: everyone, including 15-year veterans, has only had a few years with these tools. You missed the old boat; you’re early for the new one.
How long does it take to get job-ready for tech in 2026?
The honest number we’ve measured across Sigmaschool graduates: 400-600 focused hours from zero to job-ready - roughly 3-4 months full-time or 6-9 months alongside a job. "Job-ready" means the profile employers currently pay premiums for: fundamentals solid enough to judge AI output, fluency directing AI tools, and 2-3 real deployed projects you can defend in an interview. Beware both distortions of this number: the "learn to code in 3 weeks" pitch (produces the vibe-coder profile that fails technical interviews) and the "you need a 4-year degree" gatekeeping (contradicted by our own hiring outcomes - several graduates hold developer roles with no degree, hired on demonstrated ability).
How do I know if tech is right for me before committing?
Run a cheap real test instead of an expensive imagined one. The free trial exists for exactly this decision: one signup, no card, and you get the 6 Projects in 6 Days precourse on Sigmo (build real things immediately - the fastest way to learn whether building energises or drains you), a live Buildroom session with an instructor (see how the actual teaching works), and the full programme e-booklet (see the whole path before paying anything). A week of real building tells you more than a month of reading articles like this one. If the projects frustrate you in the "I want to solve this" way, that’s the signal - that frustration is the job, and some people love it.

Back to all posts The open-door profile, step by step →

The window is open. Test it free.
A week of real building beats a month of articles.

Six real projects in six days, a live instructor session, and the full programme e-booklet - one signup, no card. Find out if the open door is yours before spending a ringgit.

Start the free trial See the hiring data