i've been using claude code for a while now and it's addictive. specifically opus. there's something about having a tool that actually understands what you're trying to build. ideas that would take weeks now take days. sometimes hours.
it's overwhelming in a good way. my backlog of "someday projects" is actually getting done. the friction between thinking and building has never been this low.
the bigger picture
this is changing computer engineering fundamentally. but here's the thing, this won't stay confined to software. every knowledge work sector is next. accounting, law, medicine, design. the timeline might be slower but the direction is clear.
is the world ready for this? probably not. our economic systems, education, job markets, none of it was designed for a world where a significant portion of cognitive labor can be automated.
for new developers
if you're starting out in tech right now, learn claude code. not as a crutch but as a multiplier. understand what it's doing, learn from its suggestions, use it to accelerate your learning. the developers who master these tools will have an insane advantage.
what comes next
i keep thinking about hans rosling's work on global economic patterns. he talked about potential systemic shocks that could reshape economies. looking at how fast ai is moving, this technology might be that trigger.
the productivity gains are real. but productivity gains without redistribution of value create instability. we're building tools that can do more with fewer people. that's great for efficiency, complicated for society.
i don't have answers. but i'm watching this unfold with a mix of excitement and concern. the next few years will be interesting.
maybe that's an understatement.