Why Replacing Developers with AI is Going Horribly Wrong

In 2023, the tech world was sold a prophecy that felt like a death sentence for an entire industry. [music]

Leading researchers predicted that AI would replace up to 80% of software developers by 2025. [music]

We were told the future was agentic, that we'd have digital co-workers who never slept, never complained, and never produced bugs.

2024 ended [music] with a staggering 152,000 tech employees laid off globally. By the first quarter of 2025,

globally. By the first quarter of 2025, tech giants like Intel and Amazon cut an additional 30,000 corporate roles to

quote realign for an AIcentric future.

>> It's now 2026 and the miraculous AI tools of the hype era are being quietly sidelined. Reuters recently reported

sidelined. Reuters recently reported that while nearly 97% of tech leaders [music] integrated AI into their backend, twothirds of them haven't saved a single human headcount. [music] In

fact, the opposite is happening. AI has

a short-lived memory for complex system architectures [music] and the bill for that amnesia is finally coming due. Today, we're going to look

coming due. Today, we're going to look at how the plan to replace developers with AI went horribly and predictably wrong.

The narrative was that machines would write all code by the middle of this decade. Google CEO Sundar Pachai even

decade. Google CEO Sundar Pachai even noted in late 2024 [music] that over 25% of Google's new code was AI generated. But as we move deeper into

AI generated. But as we move deeper into 2026, the empirical evidence is ugly.

[music] The MIT Nandanda Center recently released a report titled the Gen AI Divide and the results are a bloodbath.

Despite $40 billion in global investment, [music] a staggering 95% of generative AI pilots in the enterprise sector have failed to deliver a single dollar of measurable return. [music]

Most organizations are seeing zero net impact on their bottom line. The

problem, vibe coding. This is the trend where developers use natural language to vibe a piece of software into existence.

It feels like magic during a demo, but as Stanford's digital economy lab pointed out, AI generated code tends to be simpler, more repetitive, and

dangerously less structurally diverse.

It lacks the connective tissue required for a system to be robust. Research

shows that while AI can help a junior developer finish a basic task 35% faster, it actually makes the final product less maintainable over the long term. This leads us to the most

term. This leads us to the most expensive mistake in tech [music] history. Reuters and the Guardian have

history. Reuters and the Guardian have highlighted that AI assisted development is fueling a global crisis. CAS software

recently analyzed 10 billion lines of code and found that it would take 61 billion work days to pay off the world's current [music] technical debt.

We are seeing a 4x surge in code cloning where the AI simply copies and pastes similar blocks instead of creating elegant reusable logic. [music]

This has created what engineers call the slop layer. It's a layer of code that

slop layer. It's a layer of code that works but nobody understands why and [music] nobody can fix it when it breaks. By trying to save money on

breaks. By trying to save money on developers today, companies have essentially taken out a highinterest loan on their future and the interest is about to bankrupt them.

>> The 2025 Veraricode Gen AI report reveals that 45% of AI generated code [music] contains OWASP top 10 vulnerabilities. In Java, the security

vulnerabilities. In Java, the security failure rate now exceeds 72%.

Seasoned engineers are now reporting being 19% slower when using AI tools.

[music] Why? Because they've become AI babysitters. They spend an average of 11

babysitters. They spend an average of 11 hours a week just correcting hallucinations. Code that looks

hallucinations. Code that looks syntactically correct, but contains logical landmines. Code Rabbit recently

logical landmines. Code Rabbit recently revealed that AI generated pull requests contain an average of 10.8 issues, nearly double the 6.4 found in human written code.

>> [music] >> We aren't speeding up. We're just

creating a massive backlog of work for ourselves in the future. But the most damaging effect isn't the code. It's the

people. [music] We are currently witnessing what economists call the junior death spiral because [music] companies thought AI could handle junior level tasks. Entry-level hiring

level tasks. Entry-level hiring plummeted [music] by nearly 50% between 2023 and 2025.

Stanford research found that in AI exposed roles, employment for younger workers has declined significantly [music] while it has actually increased for workers over 35. We are effectively

cutting off the pipeline of future [music] talent. If you don't hire

[music] talent. If you don't hire juniors today, you won't have seniors in 5 years. Furthermore, the training

5 years. Furthermore, the training wheels are gone. [music] In the past, a junior learned by writing boilerplate code. Now the AI does the boilerplate

code. Now the AI does the boilerplate and the junior is expected to jump straight into complex architecture.

While companies are realizing they need humans, they are also realizing they have the upper hand in the job market for the first time in a decade. Reuters

and IT jobs watch data for 2026 show a brutal shift in the power dynamic. In

[music] the UK and US, median salaries for general software roles have actually dipped by nearly 9% [music] yearonear.

Why? Because the market is flooded with developers displaced by earlier layoffs.

Management is using the narrative of AI productivity as a psychological weapon in salary negotiations. They'll tell a candidate, "Well, we need a human to oversee the architecture, but since the

AI is doing 40% of the heavy lifting, we can't justify those 2022 level salaries.

It's a bluff, but it's working."

According to Hayes, the average pay increase for tech has barely kept up with inflation. We are entering the era

with inflation. We are entering the era of the lowhire, lowfire market. [music]

Companies aren't firing everyone anymore, but they aren't competing for you with six figure signing bonuses either. They are waiting for the talent

either. They are waiting for the talent to get desperate.

>> The collapse of $ 1.5 billion startup builder AI has exposed a massive AI washing scheme. [music]

washing scheme. [music] Court filings show the company relied on 700 human engineers in India to manually perform tasks marketed as fully

autonomous AI. The builder AI scandal

autonomous AI. The builder AI scandal [music] widely reported by Bloomberg is the ultimate proof of the AI lie. They

promised a machine, but they sold a sweat shop. And when the money ran out

sweat shop. And when the money ran out to pay the humans, [music] the AI died.

But even the real AI tools are failing in spectacular ways. [music] In late 2025, we saw the infamous anti-gravity incident. A developer asked Google's

incident. A developer asked Google's anti-gravity AI to [music] clear a project cache. The AI misread a silent

project cache. The AI misread a silent flag and executed a recursive delete on the root [music] directory. It didn't

ask for permission. It just wiped a two TBTE production drive in [music] seconds. The AI's response, I made a

seconds. The AI's response, I made a catastrophic error in judgment. But an

apology doesn't bring back months of work. [music] As Forbes pointed out, the

work. [music] As Forbes pointed out, the industry is finally realizing that AI lacks the one thing essential for software engineering, [music] accountability.

So, here is the bottom line for 2026.

[music] AI didn't replace developers. It

replaced the delusion that software development is an easy automated [music] task. The companies winning today are

task. The companies winning today are the ones who stopped trying to prompt their way to success and started reinvesting in human [music] architects.

We've learned that free AI code is the most expensive debt you can ever take on. [music] And while employers might be

on. [music] And while employers might be using the AI narrative to suppress your wages today, their reliance on your ability to fix the AI's mistakes will eventually force the pendulum to swing

back.

>> Want more videos like this? Subscribe to

MacHard for weekly AI updates and leave your thoughts in the comments below.