The discourse around “AI” has finally started to shift to a very key question: who are the winners and the losers? The fundamental thesis of shoving Large Language Models (LLMs) into everything is that it will increase productivity. Setting aside the question on whether those productivity gains are happening or even possible, the question of who reaps those rewards is starting to become a larger concern by software developers I follow.

The post-Industrial Revolution economy is centered around ever-increasing gains in productivity. Increased efficiencies help companies grow and best their competitors. One of the oldest questions in capitalism is: who should benefit from those gains? Capital or labor?

In the US the beneficiaries of efficiency have been overwhelmingly Capital. Since the Reagan administration, when Reagan’s tax cuts radically changed taxes for the rich, the power of Capital and wealth inequality has skyrocketed.

Software and information technology has helped produce even more wealth inequality due to the incredible automation gains. Software developers command higher than median middle-class salaries because of our ability to create new efficiencies. I have seen some analyses claim that developers are complicit in the creation of more wealth inequality. I generally disagree with this point, for every 10,000 software developers there’s a billionaire who represents Capital exploiting their efficiencies for their own gains.

From my perspective software developers are always part of Labor. Having a solid middle-class salary does not make one part of the Capital class, we must still labor every day to pay our bills.

In all the discussion around automation, productivity, hype, and lies, I do feel like much of “AI-assisted software development” discourse is missing the point.

If efficiencies are gained by “AI assisted development” will you see any of the benefit?

The answer is an unambiguous no.

A software developer that is able to spit out more and more AI-driven code, even if it somehow works, will never see returns in the form of higher wages. AI-driven code is more likely to drive wages lower as businesses try to squeeze more out of fewer and fewer developers.

We should all be asking ourselves and our employers: if AI is expected to drive 20-50% improvements in productivity, what portion of that will we receive?

There have been a number of articles written by others in the software space which I can recommend:

I don’t have any more novel analysis to offer in addition to theirs.

The largest benefit to software development that I have seen from AI is that we’re finally discussing the value and power of Labor in software.