It's a gamble in a global game with next to no barriers to entry and it's hot now.
This is the blessing and curse of freely available open source applications and code libraries, along with cheap, infinitely scalable cloud computing resources, and app marketplaces that reach hundreds of millions of people.
Everyone has an idea to create something, or to automate a process, and the tools to build things with are right there under your nose.
Success is not guaranteed by any means, quite the opposite in fact.
But the lure of cobbling something together, giving it a Strayng Prodkct Neim and selling out after a few years for millions — or even billions if you get enough users — is immensely powerful.
If the dream doesn't come true the first time, maybe it'll work the second. Or third or fourth. All it takes is that special young person's superpower, not sleeping and subsisting on energy drinks.
That and resting assured in the knowledge that very few people understand the workings of what's being developed, like blockchain.
As long as what you're doing seems hip enough, investor money should roll in.
Interestingly enough, many startup people I've spoken to can't code much and can't be bothered to learn the intricacies of different programming languages like how you optimise apps for performance rather than just throwing more computing power at them when user numbers grow.
As long as the copy-and-paste snippet from a website works, great.
Who can blame them? Coding is only fun in small doses. Long projects become a soul-destroying chore that is difficult and repetitive, requiring the ability to understand badly written documentation and unclear, ambiguous programming language features.
Getting code more or less right (it'll never be a 100 per cent correct) also takes heaps of time; time that nobody has in 2018.
Attempting to alleviate the skills shortage with artificial intelligence and machine learning will make things worse, not better. AI and ML are not ready to replace programmers, but the threat's there and coders know that their jobs will go as soon as the machines reach the good-enough for purpose stage.
Young techies know that programming in general is destined to become a dead end career, so why not get that idea coded up while there's still time and a market for it?
(Alternatively, super brilliant coders should move into AI/ML because they'll be the rare ones who understand how the technology works and basically take over the running of the world, but let's park that one for now).
There are no easy fixes in other words, but given where we're headed, businesses need to love their existing skilled code workers even more and look after them well because finding new ones will be hard.