Most ICOs are dubious – but it’s still the future of software economics

Yes, I know. Many ICOs, which allow for blockchain-enabled sales of tokens to be used in a variety of projects, are the subject of much controversy, and even contempt in some circles – especially by those who would be disrupted by them such as Jamie Dimon.

Still, the idea that a community can build its own platform via a blockchain-enabled fundraise, by selling tokens to its own users in advance has some profound implications: Who needs a venture capitalist or a corporate sponsor – or even, perhaps, a company – if you can self-fund and self-organize around an idea via an ICO?

Of course, we’re currently in a bubble and the ICO concept is attracting hundreds, or perhaps thousands, of “ideas” that just don’t work. Many projects are force fitting some definition of a “utility token” onto ill-conceived business models (just as the “add Machine Learning to industry X” business meme).

Even as the ICO market has raised more than $3 billion, another stat to watch is that currentlyonly 1 in 10 tokens are actually in use following their sale.  

Though different than a tulip bulb bubble of the 1600s, the frenzy ICOs create is actually leading to an avalanche of advances in protocols, networks, cryptography and other technical building blocks which will benefit us for years to come. Blockchain itself is a transformative technology, which proponents argue will be come as fundamental metric system. Despite current limitations, technology will continue to evolve to allow increased throughput (e.g., not requiring replication to each node), and to support different distributed compute models (e.g., blockchain oracles).

Similar to the technology explosion, there are financial innovations enabled by tokens we are only beginning to understand, for example the rearranging of business organizations as we know them.

Today, ideas need companies to function. Apart from being the delivery of the idea, companies do a lot on inefficient things for the humans involved in them. Companies require lots of overhead (investors and financing, sales, marketing, HR, recruiting, real estate, bean bags, etc.) None of this is fundamentally necessary for many ideas out there.

Simply put: Companies are an inefficient way to cohere multiple brains into a compute  singleton.

Two things are happening, which will obviate the need for many types of software companies. First, the ICO enables a more efficient mechanism to raise capital and token governance can align the interests of participants better than the standard joint stock company. Second, the software stack has matured to the point that it doesn’t require a physical organization for life support. This is the animating force of smart contracts, taken to the extreme with the Distributed Autonomous Organization, an investment organization with no staff, and no compnay.

As network services move on the plumbing is disappearing, and accelerating with technologies. So what happens next? Perhapshe support teams aren’t going to be needed, as issues resolve themselves by automated contracts.

Even systems architects are going to disappear. The new architecture is just business logic (serverless functions) and data logic (map/reduce) without the current 99% code overhead. (Our previous node/express server had 200 files and six “functions”; our new Lambda “serverless” that replaces it has six files: one for each service.)

Of course, it’s not like software is going to be commissioned and integrated through telepathy.

There is a need for new platforms that enable distributed work to be done in ways that are increasingly frictionless. And token economics are pushing the edges of this trend.

The startup numer.ai is a great example of this (at least in principle). It’s a hedge fund attempting to crowdsource human expertise — and is making this possible by designing a conceptual abstraction that coheres the participants without requiring each “node” (i.e., brain) to understand the entire stack (or even communicate with other nodes). I don’t have any expertise in financial modeling, but intuitively it seems that enabling multiple participants to conceive of different approaches to modeling the same data set is perfectly scalable. It’s an extreme version of the wisdom of the crowd.

As a thought experiment, the trend will continue toward singularity since fundamentally there is no need for human programmers for many current use cases. Currently human programmers are mostly gluing together human processes (sales, financial services, recruiting, construction, product development, real-estate, etc.)

What if we didn’t need that?

What if businesses, the “buy side,” could specifically encode their requirements for developers on the “supply side,” who could then build components. Not only could a platform or marketplace potentially manage all of the drudgery of scale and testing, it could also obviate the need for entire corporate structures.

We are already seeing the trend of engineering, design and other talent being commissioned through Gigster or 99Designs. Vetted talent can be accessed without hiring or creating an entire corporation.

These nascent outsourced freelancer technology models still require a platform. They are centralized, they have owners to enforce the rules of the platform. Token economics allow for further automation, and distributing of the risk and utility of the crowdsourcing occurring.

How could a crowd assemble a sophisticated application credibly? Well on close inspection much of what businesses need to run are specific and highly commoditized components.

And even building most new features doesn’t require much more than clear specifications that perhaps soon could be built via AI.

This idea of this collapse of functions around software is daunting: The software sector contributes more than a trillion dollars to U.S. GDP, and supports over 10 million jobs. And there are entire software types that are built largely for software developers. The opportunity and disruption is on a scale rarely seen.

Crypto tokens, immature and crazy seeming as they are today, could provide much of the business model and funding to accelerate the marketplaces and ties to talented developers and the businesses that need them.