The Effect of AI on the Distribution of Developer Salaries

Robert Corwin, CEO
|
Aug 18, 2025
|
Co-Author
Thin wave line.Thick wave line.White wave line.

Important Notice

Diamond-shaped blue icon as bullet point.
Diamond-shaped blue icon as bullet point.

Key Points

Diamond-shaped blue icon as bullet point.
Diamond-shaped blue icon as bullet point.
Diamond-shaped blue icon as bullet point.
Diamond-shaped blue icon as bullet point.
Diamond-shaped blue icon as bullet point.
Diamond-shaped blue icon as bullet point.

In the past week, Sam Altman said programmer salaries are “going up extremely rapidly” while the Wall St. Journal noted many companies are drastically reducing hiring of developers.  How can we reconcile those two observations?

The answer is that the shape of the distribution of salaries is probably changing in the IT industry.  We are seeing huge pay packages at the highest skill levels (consider that Altman is surrounded by the top 1% of talent in the world), plummeting pay at the lowest, and varied effects in the middle. (If anyone with job data is interested in verifying this statistically, pls contact us).

The reasons are simple.  AI is now as good as low-quality developers.  If you’re just writing APIs, database queries, or doing ETL, this is no longer much of a value-add, even at rock-bottom rates.  On the other hand, AI is making experienced developers and architects more productive by a factor of 2 to 5 or more.  Particularly at a big company, such an employee might add millions of dollars of value, so it makes sense to pay them highly.

In the middle, it depends.  Even what was previously considered more complex coding, say, “full-stack engineering and software architecture,” is not immune, as AI increasingly chips away at each part of the stack.  I think those relying only on coding knowledge – no matter how advanced that knowledge seems for now – are going to be pressured.  But those with other, diversifying talents are faring well, at least as viewed from our perch in an AI consulting company.

Examples of this include having deep knowledge about a specific industry, having advanced scientific knowledge about AI, data science or math (the Open AI guys are really better described as exceptional scientists, not coders), having good leadership and communication skills, speaking another language, being a good businessperson, having a good network and professional relationships, graduating from a top-tier school, being involved in charitable causes.  In other words, bringing something else to the table.

I don’t think people should aim to be less technical.  On the contrary, I think future success will require both technical AND business knowledge.  But I think gone are the days of dialing it in remotely cranking out low to medium level coding tasks.  As are the days of the business side ignoring or not believing in advanced tech like AI and data science.

Perhaps I am overstating the case a bit, as coding is currently nowhere near as extreme as other industries which share this type of distribution, such as sports, music, poker, and restauranting.  In them, the market enriches the highest skill level, pays sustenance wages to the middle, and doesn’t have much use for the lowest.  The 1,696 players in the NFL represent about 0.0015% of people playing football in the country.  Or think of the gap between Taylor Swift’s earnings and your favorite local band. But programming seems to be shifting in this direction.  Will we one day have just a handful of “rock-star” coders while the rest of us look on?

Our Use Cases
Get in Touch

Read Next