TheEndOfSoftware

Software isn’t ending; it’s just beginning. However, in a future where anyone can quickly build software, the focus of builders will shift towards being wrong for the least amount of time possible.

A decade from now—in a hyper-competitive and automated landscape—quickly carving out a unique niche with enough of a foothold to land and expand will be more challenging than ever.

Software is a means to an end; it’s one way to solve one key problem. The problem with past software is that it was built (& funded) by the wrong people with the wrong incentives, leading to a 90+% failure rate. Software has historically been an inefficient bottoms-up process where “visionaries” fund and build based upon intuition and hope they can iterate their way towards product-market fit.

In the future, software will be top-down developed by the same people who need it, cutting out much of this middle bloat along the way and making room for exceptional, personalized software.

In the end, software doesn’t need to work for everyone, it just needs to work for you.

Self-assemblage

In a world where anyone can build, everyone will build (often without realizing it). The digital world will be better at self-assembling based upon acute pain, needs, and greed.

People will be able to craft their own experiences and solve their own problems, bypassing the scar tissue of why most human-powered experiences (SaaS, healthcare, housing, regulation, etc.) are inefficient.

Ideas were once the exhaust of experience, but this will shift towards a broader, machine-driven understanding of needs and trends as orchestrators gather and solve a collective of global needs.

Broad empowerment will shift to creative, abstract thinkers, but acute empowerment will also shift to the individual, enabling them to solve any need for the right price.

Energy and compute will become the currencies required to fulfill many needs once individualized paywalls and isolated software fall by the wayside.

Functional pragmatism will beat branding

Less than 30 years after broad adoption of the internet, startups already need to be a sharp point to pierce the world due to competitive pressure and mindshare fatigue.

As automation increases, builders will necessarily stop wastefully building net-new platforms and apps as siloed off, branded experiences. Instead, software will follow an extension model where each software package provides a functional, pragmatic worker that executes its work within a larger orchestration engine to satisfy people’s & businesses’ needs.

Orchestrators will fluidly assemble and compose all these components to that end, replacing the manual API integrations that are the bane of many engineers' existence today. Instead of excuses, end users will have solutions.

Single-use software

As orchestrators are built, software will also become single-use in many scenarios. People will be able to request infinite variants of experiences and needs which permutate and extend some underlying software, only to have those software tweaks thrown away afterwards.

Paths will also be forged, broken down, and reforged as software understands how to tune itself towards compartmentalized end goals.

In the end

As has always been true, even of art: the only thing that matters is building what people want—especially yourself. What will ultimately change is who is doing the building.

If Amazon enables everyone to get any physical product they want, who will do the same for digital experiences?

Software won’t be written; it'll be summoned, shaped, and set free. Tap start to begin.