And to give the teams at Anthropic and OpenAI their flowers, this has not been a failure of their tools. The tools work, more or less. This is all about something stranger and more important than a failure of the LLMs: the necessary work of software has already been done. We havenā€™t seen a lot of new types of software because we already have what we need. You canā€™t flood a vessel thatā€™s already full, and the amateurs, God bless them, have been pouring this new power into a jug that weā€™ve been frantically filling throughout the 21st century.

[...]

And, yes, professional human software engineers will be a thing tomorrow and next week and in five years and in ten. What I mean is narrower but far-reaching and profound: the era of new software categories is over. There is no more true need for new types of software than there is need for new lines on the periodic table.

[...]

for the end of one S-curve to become the beginning of another requires that a new reality thatā€™s waiting underneath the old one. My claim is that software, having reached the ceiling of natural human need in anything like the recognizable human world, has no new reality move into. Thereā€™s no more fundamental layer of computation to migrate to. The next curve, if it comes, wonā€™t be built on top of more code. Instead we need a different kind of new development, one which will come in the world of atoms, not the world of bits, which means weā€™re stuck in the same problem.

[...]

There are limits. What Iā€™m arguing is that the adjacent possible in software is remarkably constrained and largely similar to what weā€™ve already gotten.

[...]

Software was never going to deliver the future. Software was the thing we did instead of the future, because it has orderly rules, because itā€™s cheap to make, because inside its narrow confines the returns come fast and the failures are easy to identify and contain.

https://freddiedeboer.substack.com/p/the-project-of-software-is-complete

Hmm... I guess we shall see where all this goes. The worthwhile things are probably always going to be hard.