AI is eating the software that is eating the world

About 12 years ago, Marc Andreessen penned his famous “why software is eating the world” essay in WSJ. His famously prescient claim had a strong why now to it:

“Six decades into the computer revolution, four decades since the invention of the microprocessor, and two decades into the rise of the modern Internet, all of the technology required to transform industries through software finally works and can be widely delivered at global scale.”

At the time he made the claim, which today is a considered almost a cliche, there were about 2B internet connected users in the world.

Here are some other additional ‘state of union’ facts to bear in mind from then

• Git hub hosted about 2Mn repos and A16z had yet not invested in the bootstrapped co
• Docker, Zapier and multiple other dev tools had yet not been released
• Devops had just started becoming the major theme in software zeitgeist | Debois had just coined the term 2 years ago.
• MLops hadn’t even been originated (2015 was when Sculley and team published “Hidden Technical Debt in Machine Learning Systems”)

Today

• There are about 5B internet connected users in the world and about 14B internet connected agents (humans and devices included)
• Github has 90M+ developers, 3m+ active organisations and 200M+ active repos were created in 2022 alone!
• MLops is pegged at $50B as an industry
• Developers are enabled by 1000s of tools which address SDLC, many of them open core/source

Today on back of LLMs and AGI, there is a new disruption building with important implications for entrepreneurs, consumers and investors. More importantly, this disruption is not new. It is only coming of age now after 5 hype cycles, with first one starting as early as the first modern computing era around WW-II.

It is time to expand on Marc Andreessen’s maxim that software is eating the world:

AI is eating the software that is eating the world

This drastically benefits consumers but creates challenges for entrepreneurs and investors

As an entrepreneur and investor it is critical to understand dynamics of value creation, capture and extraction. It is easy to chase value unlocks, but far harder to discern which pockets of human endeavour can both capture and extract value, allowing cos to build durable value for themselves.

Enter the new disruption

There are three forces at play which are disrupting software as we know it:

1. We are under-going a fundamental shift in software engineering and development, where the new programming language for soft-tech is slowly but steadily becoming “English”.
2. LLMs have hit an inflection point where neural network stacks are the new protocols for all soft-tech and we are on path to human-interpretable AGI unlocking software creation at breakneck pace
3. Big open source movement and composability unlocks of 1+2 above, most of which is coming of age with advances in devops and MLops and a large community of matured developers

On back of these, world of software is going to get disrupted in ways which many can’t imagine.

I believe, that value capture in software is going to now mostly move away from middle of the stack, to down the protocol layers (LLM and neural net stacks) and upwards to end-consumer facing products (whatever I save in my SaaS bill i can push out to consumer value)

We are now nearing the end of an era in software where businesses solving the problem statement of “you don’t need to write code for non-core task XYZ, here is a SaaS”, shall see value erosion.

To now be relevant as a SaaS co, depth+breadth of workflow is going to be more and more critical. Point problem solutions will find it harder to establish why they capture value.

One could argue that there are always going to be non dev first cos which would continue needing SaaS. And i don’t deny that. But many of that SaaS would come open core/sourced now because all the dev first cos which now have unlocked dev productivity —> devs there will be contributing more and more to the open source.

This is the true promise of composability and open source, and I'm excited to see it play out.

AI unlocks of dev productivity < > Open source development is THE virtuous cycle!