By 2030, 90% of developers would be

replaced by AI. Those were the experts

estimates a few years ago, but today the

reality is very different.

>> Gartner just said that 50% of companies

who laid off workers because of AI will

rehire for the exact same roles by 2027.

>> So far in 2026, software engineer hiring

is skyrocketing. It's true that it could

wreck havoc, but I think it's much more

likely to reshape jobs rather than to

erase them altogether.

>> But this is not necessarily because

companies are not using AI, but because

AI makes so many mistakes when writing

code, the expert developers have to

correct it n out of 10 times.

>> Tech giants like Microsoft and Google

are outsourcing more and more coding to

AI in a productivity push. But some new

research shows the tools might not be as

helpful as some expect. This not only

affects developers productivity but also

companies finances since the increase in

AI generated junk code raises the

workload.

>> Seasoned engineers were actually 19%

slower when using AI tools instead of

speeding these engineers up. The AI

often gave them suggestions that looked

helpful but actually required

time-consuming corrections. All of this

has made companies reconsider their

decision to delegate code creation

entirely to AI because instead of

reducing costs as previously thought, it

is now increasing them. So why are

companies quietly rehiring software

engineers?

Over the past few years, the tech

industry has repeatedly told us

something that seemed inevitable.

Companies like Google, Amazon, Meta, and

many others have tirelessly repeated one

phrase. Artificial intelligence is going

to replace workers, especially

developers. Since early 2024, companies

began reducing teams, freezing hiring,

and betting on automatic code generation

tools that promise to build complete

software in seconds. It is estimated

that since 2024, companies have laid off

around 124,000 software developers.

>> Amazon, CHEG, Microsoft, Meta,

Salesforce, the list continues. Tech

companies just are cutting tens of

thousands of jobs.

>> The idea seems simple. Fewer developers,

more automation, and higher profit

margins for companies. However, what

started as an apparently unstoppable

revolution quickly began to show its

limitations. News such as Amazon

experiencing four critical errors in

just 90 days due to AI assisted code

changes made the world realize that AI

was not as promising as believed.

Reality began to show that writing code

is not the same as developing software

and that artificial intelligence

although powerful still heavily depends

on human supervision. Gradually,

companies began to notice that reducing

teams was not generating the expected

results, but instead creating new

problems, more complex, and costly to

solve.

>> The study suggests that the return on AI

coding, it may be more uneven, less

immediate than investors have priced in.

>> While AI proved to be extremely

efficient at generating code quickly,

the truth is that it also began

producing large volumes of code that

required constant correction. Recent

studies have shown that AI generated

code contain up to 1.7 times more errors

than code written by humans. This has

forced companies to spend more time

reviewing and fixing these issues.

Additionally, massive code generation

has caused companies to have up to 38%

more code to maintain, increasing system

complexity. The problem is that software

development is no longer what it used to

be. Because now the challenge is no

longer writing code, but maintaining it,

optimizing it, and fixing it. A relevant

fact is that 61% of companies that

adopted AI tools for programming

increased their hiring of senior

developers in 2026, mainly to review AI

generated code. Now, some of those

mid-level Google engineers are shaking

free. This is your dream. So, you just

need to hold your hand up in this job

environment if you're a former Google

engineer and you're going to get offers.

>> Another major problem is that artificial

intelligence generates code without

understanding business context.

According to Gartner, more than 50% of

errors in AI generated code are related

to a lack of business context

understanding, not syntax or programming

errors. AI can write functions,

structures, and algorithms based on

learned patterns, but it does not

understand business objectives,

technical constraints or strategic

decisions. This means that many times

the generated code works in isolation

but fails when integrated into real

systems. This has been notable as a

study by IBM found that four out of 10

development teams reported compatibility

issues when integrating AI generated

code into existing infrastructures. As a

result, human developers must intervene

to adapt that code, fix errors, and

optimize performance, increasing the

need for experienced developers.

>> The tools, they're certainly being used.

They're here. The payoff may just be

more uneven than the hype suggests. and

perhaps it plateaus at a certain point.

>> But perhaps the biggest problem with AI

is that unlike a human developer, AI

cannot correct itself. When a developer

makes a mistake, they can realize it in

time. But when AI makes a mistake, it

does not detect it unless you point it

out. Researchers at Princeton University

discovered that AI models failed to

self-correct in more than 60% of cases,

even when asked to review their own

code. This has led companies like

Google, Amazon, and Meta to begin

reconsidering their initial strategy.

>> Companies are trying these tools.

They're not always working, and so

they're asking eventually, "Where's the

value?"

>> And experienced developers are aware of

AI's limitations. Recent studies have

indicated that up to 96% of developers

do not fully trust AI generated code.

This has turned code supervision into

one of their central tasks, leaving

aside important tasks such as

innovation. As a consequence, the real

productivity of the few remaining

developers began to decline. A GitHub

study found that 49% of teams reported a

decrease in real productivity and

ultimately this translates into higher

costs for companies.

>> Big tech now pouring more capital than

ever into AI. Alphabet, Microsoft, Meta,

Amazon expected to spend a combined $600

billion in capex. That's this year

alone.

>> All these limitations of AI generated

code have created a bottleneck in

development teams. As a result,

companies have begun rehiring employees

who were previously laid off. It is

estimated that up to four out of 10 new

hires are software developers who are

former employees of companies that have

laid them off after replacing them with

AI. This phenomenon known as boomerang

hiring is rapidly growing in the tech

sector as these professionals can

integrate faster, understand internal

systems, and detect complex errors that

AI fails to identify. I want you back.

It's a catchy tune and it's also an

increasingly popular hiring trend. It's

what's called boomerang hire. 35% of new

hires being made up of past employees.

While it is true that AI is replacing

some software developer roles, these are

mostly junior positions since AI can

replace the work of a novice developer,

but not that of an experienced one.

Basic programming tasks traditionally

assigned to beginner developers can now

be performed by AI tools. This has

caused many companies to reduce junior

hiring and increase demand for

experienced developers. More than 54% of

companies have indicated they plan to

hire more senior developers while

reducing junior positions, reflecting a

structural shift in the tech industry.

>> Now, this latest suggests that while it

can help that group level up, it may

actually be increasing reliance on

senior talent because someone still

needs to debug, refine, and ship the

final product. The tech industry is

beginning to recognize that artificial

intelligence does not replace developers

on its own. And this is reflected in

recent hiring trends. For example,

around 20% of the software engineers

hired by Google in 2025 were former

employees who were rehired. But this

return is happening quietly without

major announcements yet with a

significant impact on the labor market.

The developers being rehired are those

with experience capable of supervising

artificial intelligence and improving

generated code. This demonstrates that

artificial intelligence is not at the

level of an experienced developer.

Sometimes it is simply used as an excuse

to boost stock valuations and lay off

workers.

At Economy Media, your opinion matters

to us. Subscribe and let us know what

you think in the comments below.