I’m not even remotely worried about AI eliminating software development jobs. In fact, I’m pretty sure there will soon be a boom in both software development jobs and the amount of software available to everyone.
People have always worried about automation causing massive unemployment. Each time a breakthrough happens, folks are sure that “it will be different this time.” Only it never is different.
But the worriers persist.
It’s paradoxical
You can tell them all about the Jevons paradox — the observation that as something becomes more efficient, demand for that more efficient thing increases rather than decreases. In the mid-19th century, William Jevons noticed that the use of coal became more efficient. Humans figured out how to get more heat and energy out of less and less coal. The common belief was that, because less coal was needed for the same amount of energy or heat, there would be less demand for coal as a result. Everyone was concerned that coal miners would lose their jobs. But Jevons noticed that demand for coal actually went up, as the more efficient processes led to more widespread uses for coal.
The same thing happened half a century earlier with the introduction of the automated loom. Despite fears that the power loom would destroy jobs for weavers, it made the production of clothing and other textile products cheaper, increasing demand for such products and increased employment in the textile industry.
This phenomenon can be seen over and over again. Spinning jennies, automobiles, computers, robotic manufacturing, tractors, sewing machines, and countless other inventions all caused widespread fears of job loss, but the fears were never really realized. When a company can suddenly produce 10 times more with the people they have, they have always wanted to produce 10 times more, not cut their workforce by 90%. Yet here we are, with everyone sure that AI is going to put us all out of work.
It’s not going to happen — especially in the software development realm. You know what is going to happen? The same thing that always happens. That which is automated and made more efficient will find new and different ways to express itself. Existing software will suddenly be vastly more useful as the backlog of features can be implemented. New software ideas that were previously too complex for humans to write and manage will be created.
Marc Andreessen was never so right as when he said that “software is eating the world.” Sure, software was eating a lot when humans wrote every line of code. But now that code can be written 10 or 100 times faster, software’s appetite will go from hungry to ravenous. The work that can be done has expanded rapidly. And that work will be done because there is too much money in building what we have always wanted but that humans alone could not deliver.
A positive-sum game
The world is never a zero-sum game, but humans seem hard-wired to view the world that way. Only now, with AI, we have what Daniel Jefferies delightfully calls “Fear Mongering as a Service,” running rife through our industry. Yet while all the Chicken Littles decry the job market falling out of the sky, job postings continue to actually increase, and it is becoming harder to fill those jobs.
Now that doesn’t mean the market isn’t shifting. The demand is strong for experienced engineers and weaker for entry-level jobs, a situation that is creating a bit of a paradox all by itself. The skills that worked for many years may not be as valuable going forward. Writing good code and getting an AI agent to write good code are two different but related skills.
Source – https://www.infoworld.com/article/4167463/no-ai-wont-destroy-software-development-jobs.html