Software to Blockchain to the Metaverse

I have now watched three different technologies get introduced as the thing that changes everything, taken each one seriously, and then watched the conversation quietly move elsewhere. Every time, I was sure I was at the edge of it. Every time, I wasn’t.

Here is how the three eras actually went, and what stuck.

Software

I started out writing software, which is still where most of my income comes from. The appeal was never complicated. You write some code, it does a thing, and someone uses it. That loop has not changed, and I am suspicious of anyone who claims it has.

What I remember most is debugging late into the night, and the specific dread of shipping something and then waiting to find out what people thought. Those hours taught me more about product than any book did, mostly that product is far more repair than design.

Blockchain

Blockchain got my attention when the interesting problems in software started looking like coordination problems rather than feature problems. Ownership, trust, and who was allowed to move money were all things the existing stack handled badly.

I got in early. I explored decentralized applications, wrote smart contracts, did the NFT thing because of course I did the NFT thing. I made the standard set of mistakes roughly a year ahead of everyone else, which at that age felt like being early rather than being wrong.

The regulatory picture was unclear the entire time. Looking back, that was the real product constraint. Nobody cares what the tech does if you cannot legally operate it.

Spatial

By the time the Metaverse got loud I was skeptical, and curious enough to try it anyway. The honest summary: VR and AR are real, genuinely useful for a narrower set of problems than the hype suggested, and still a long way from being how anyone works.

I spent this stretch building 3D environments. The skills barely overlap with the previous two. Working in three dimensions is a different discipline and I underestimated how much of it is spatial reasoning rather than programming.

What carried over

The tools changed every time. A few things did not.

Ship early and watch what real people do, rather than what you predicted they would do. Get close to users before you build the version for them. Have a plan for when the hype cycle ends, because it always ends, usually faster and more completely than you expect.

I do not spend much energy picking the next wave anymore. Betting on one is fine work. Building something a few hundred people genuinely need is the part I am actually good at, and it has never required me to be early to anything.