“Learn the Basics”: What the Data Really Says About IT Jobs in the Age of AI

On forums, both admins and developers repeat the same advice: don’t just learn how to work with model APIs; learn the basics. That advice makes sense. However, the claim that IT specialists are “returning en masse” to networking, Linux, and hardware is not supported by the data. This weekend’s commentary distinguishes between what can be measured about the job market in the wake of AI’s rise and what is merely the prevailing sentiment in discussion threads.

What is verified

The most reliable data pertains to beginners, not the entire field:

• Young employees in fields vulnerable to AI. According to an August 2026 update to a Stanford study, which analyzes wage data, the employment rate for people aged 22 to 25 in occupations most exposed to AI is roughly 19% below the level it would have reached had it followed the same trend as in less-exposed sectors. No such difference is observed among more experienced workers.
• Number of job openings for entry-level workers. Various analyses indicate a decline in job openings for entry-level workers ranging from 16% to as much as 60–70% starting in 2022. The range is wide because the methods, countries, and definitions of “junior” vary. However, the trend is the same.
• The Entire Field. The U.S. Bureau of Labor Statistics further expects the number of software developers to grow by approximately 15% over the decade, but a decline for programmers in the narrower sense (by approximately 7%). According to available data, job postings for software positions are still below 2020 levels, but they have increased since the beginning of 2025, particularly for senior and AI roles.
• Skills. The demand for AI skills in job postings is growing, and the salary premiums are significant. At the same time, the ability to evaluate and test AI-generated code is increasingly valued.

What is controversial

• Cause. It is unclear to what extent AI is responsible for the decline. A Yale University study from September 2026 found no clear evidence of a general disruption of the labor market by artificial intelligence, and its indicators remain within historical ranges. Other analyses point to the impact of higher interest rates, corporate restructuring, and the offshoring of jobs.
• The Future. The claim that AI will be able to handle routine coding “better in a few months” is a prediction, not a fact. Estimates vary, and the evidence regarding an actual acceleration in developers’ work is mixed.

What I couldn't find

Evidence of a mass return by IT specialists to networking fundamentals, Linux, and hardware infrastructure development. I haven’t found any labor market analysis showing significant growth in these positions driven by AI. Positive signs of a “return to basics” come mainly from recommendations and discussions, not from measured data. Forum discussions are a useful indicator of sentiment, but not proof of a trend. People who feel uncertain tend to speak up more loudly.

What does this mean?

Despite the lack of data on the “big comeback,” the recommendation to learn the basics is well-founded for another reason. While AI makes writing code cheaper, what remains important is what is difficult to generate without understanding:

• Understanding the system. Network, operating system, databases, performance, and reliability. Without this knowledge, you won’t be able to assess whether the proposed solution makes sense.
• Production debugging. Finding bugs in real-world systems requires knowledge of what’s happening beneath the surface.
• Review and testing. According to available analyses, employers are increasingly looking for people who can verify AI-generated code, not just write it.
• Security and operations. Those who understand networks and servers can also secure what the AI generates.

This does not mean that AI should be avoided. The data suggests the opposite: people who combine the basics with effective use of AI tools are in a better position in the market than those who know only one or the other.

Tips for Different Groups

• Beginners: Use AI to learn, but regularly try to do the basics on your own. Build a project that you understand from the network to the database. Consider related roles (operations, testing, support, data) as a starting point.
• More experienced developers: Invest in system design, security, and operations. Learn to critically evaluate AI output.
• Employers: If you stop hiring beginners, you’ll face a shortage of senior developers in a few years. Think about how to train them.

What to Take Away From This

The headline “The IT World Is Taking a New Course” is catchy, but the data tells a different story: it points to a worsening outlook for beginners and a growing value placed on experience and the ability to use AI. The advice to “learn the basics” is sound even without the claim that people are returning to them en masse. The basics are what will help you assess what AI has to offer.