Misleading news should not worry IT engineers
The article’s headline is attention-grabbing, but the interpretation needs considerable caution. The most interesting finding is actually the apparent contradiction: AI/ML professionals are among the most worried about AI-driven job cuts, while engineers are reportedly among the least worried.
1. The 66% figure does not mean 66% will lose their jobs
The wording matters.
“66% expect job cuts within 3–6 months”
means expectation/perception, not an observed probability of individual job loss.
There are at least three very different propositions:
- 66% believe their team may see headcount reduction.
- 66% believe their own job is at risk.
- 66% will actually lose their jobs.
Only the first appears to be claimed. Headlines can easily make readers interpret it as the third.
2. Why would AI/ML professionals be more worried?
This is actually economically logical.
AI/ML professionals have unusually high AI exposure. They understand what current models and agents can automate because they work with them every day.
Consider an AI team:
Earlier
20 ML engineers → build models → maintain pipelines → evaluate models → deploy models
Increasingly
8–12 engineers + foundation models + AI coding agents + automated evaluation + cloud infrastructure
AI does not necessarily eliminate the entire occupation. It can reduce the number of people required to produce a given amount of output.
This is the distinction highlighted by : AI often automates tasks, rather than eliminating occupations wholesale.
3. “Engineers are least worried” may actually be the most interesting finding
Engineering is an enormous category.
A software engineer who previously spent:
- 30% of time coding
- 20% debugging
- 20% testing
- 15% documentation
- 15% requirements/design
may now use AI to compress several of those activities.
But that doesn't automatically make the engineer redundant.
The engineer who can specify, architect, review, validate and integrate AI-generated output may become considerably more productive.
The danger therefore isn't necessarily:
AI replaces engineer
but:
AI-enabled engineer replaces several conventional engineers.
That is a much more plausible labour-market mechanism.
4. The real threat is probably the junior-to-mid-level career ladder
This is where I would be more concerned than the headline suggests.
Traditional technology careers often worked like:
Junior → Senior → Lead → Architect/Manager
Junior employees performed relatively structured tasks while gradually acquiring expertise.
AI can increasingly perform some of those structured tasks.
That creates a potentially serious problem:
If AI eliminates much of the junior work, where do future senior engineers acquire experience?
This could produce a “career bottleneck” rather than immediate mass unemployment.
5. AI may reduce employment while increasing AI investment
There is no contradiction here.
A company can simultaneously say:
“We are investing ₹500 crore in AI”
and
“We need fewer employees.”
In fact, the second can be part of the business case for the first.
Suppose:
Without AI
100 employees × ₹15 lakh average cost
= ₹15 crore/year
AI investment = ₹3 crore
If AI allows the company to accomplish the same output with 70 employees:
70 × ₹15 lakh = ₹10.5 crore
Potential labour saving = ₹4.5 crore/year.
So the company can increase AI spending while reducing headcount.
6. But there is a major counterargument
AI adoption can also create new demand.
If productivity rises dramatically, the price of producing software/services can fall.
Lower prices can stimulate demand.
For example:
AI makes software development 40% cheaper → more companies can afford software → software demand expands → some of the displaced labour gets absorbed.
This is the classic productivity → price → demand → employment mechanism.
Whether this offsets displacement depends heavily on the elasticity of demand and how rapidly AI capabilities improve.
7. India has an additional vulnerability
India's IT/services model has historically benefited from large pools of relatively inexpensive skilled labour.
That creates a particular risk.
If a US company previously outsourced:
100 developers → India
but AI-enabled teams can accomplish the same work with:
40 developers
India doesn't necessarily lose the entire business.
But it may lose 60% of the labour requirement.
That could pressure:
- IT services
- BPO/KPO
- software development
- testing
- technical support
- documentation
- routine analytics
- low-complexity consulting
8. The winners may not be “AI specialists” alone
The strongest position may increasingly be:
Domain expertise + AI + decision-making
rather than merely:
AI expertise
For example:
AI + healthcare
AI + finance
AI + semiconductor engineering
AI + cybersecurity
AI + legal
AI + manufacturing
A person who understands the business problem, domain constraints, regulatory environment and AI capabilities can be harder to replace than someone whose role is primarily executing a technical task.
9. The biggest misconception in the article
The question shouldn't simply be:
“Will AI replace jobs?”
A better question is:
“How many people will be required to produce the same economic output after AI?”
That is the crucial variable.
If:
100 people → ₹100 crore output
becomes
60 people → ₹150 crore output
then society may become more productive while employment in that particular function declines.
The resulting economic problem is distribution, not necessarily technological failure.
10. My overall assessment
I would classify the article's claim as:
Directionally important, but insufficiently nuanced.
The 66% statistic is useful as a sentiment/expectation indicator, but shouldn't be interpreted as a forecast that two-thirds of AI/ML professionals will lose their jobs.
The more credible medium-term scenario is:
AI doesn't eliminate most occupations.
It reduces the labour required for many tasks within those occupations.
And that creates a particularly interesting paradox:
The people closest to AI may be the most frightened by it—not because they misunderstand AI, but because they understand it better.
For India, we should watch IT-services revenue per employee, employee additions, fresher hiring, utilisation rates, billing rates, and revenue generated per engineer much more closely than AI-layoff headlines. Those indicators would tell us whether AI is merely changing work—or structurally reducing India's labour advantage.
Comments
Post a Comment