Imagine a world where your washing machine can recognize it needs more detergent and orders it for you. Now imagine a world where your self-driving car needs to fill up, or charge if it is electric, and does that on its own time. Then go a step further and imagine a self-driving truck crossing borders, not only filling itself up on the way but dropping off goods and buying new ones, based on an artificial intelligence (AI) that selects the best profit margin available and paying any customs fees – all without human input. That’s the difference between the current Internet of Things and the future Internet of Payments. It has profound implications for the future of banking, trade, and finance in general. The finance industry may be about to experience the same disruption that the internet inflicted on print media.

The crucial thing about the Internet of Payments is that it involves transactions undertaken by machines – whether they be actual tangible things like a self-driving car or a virtual company that exists only in lines of code. While humans own the machines and see the payments come out of their pockets or the profits that accrue to them, their direct human authorization is not needed for every one of them.

This is a major change. It will mean the possibility of many more payments taking place. As long as the human is happy with the arrangement and can cover them, the ease of the transactions will likely mean more of them happening. This is because of a phenomenon that economists call “transaction costs” – the higher the transaction cost, the less chance of it occurring, and vice versa. If you can “set it and forget it,” you’re more likely to make the transaction.

This should be a happy thing for banks. Moreover, the advent of such mobile technology offers banks the ability to reduce costs and concentrate more on customer service. Banks will be able to communicate more with the customer through their phones or other central device, such as a home “hub” like Amazon’s Alexa device. Indeed, developing the API (application program interface) technology to enable bank systems to talk to other systems to enable all of this should be top of any bank’s to-do list.

However, most current payments require human authorization through at least one third party. Often there are several parties to the transaction – customer, vendor, each party’s bank, and probably payment processors and networks as well. Each of the third parties charges something for the service. Those too are transaction costs. They can reach as high as 3 percent of the total payment. Fixed fee elements also make small payments problematic.

Importantly, the Internet of Payments also holds out the possibility of these costs lowering, and thereby presents a threat to banks’ income streams almost as problematic as the disappearance of print advertising has been for newspapers.

This is because Internet of Payments transactions will probably require a much more scalable payments system than the one currently dominating the payments market. Customers will want to have their machines engage in small-scale payments as much as large ones. If the technology exists they might want to have their self-driving cars pay others small amounts to get out of their way, for instance.

Such a system already exists in the form of cryptocurrencies like bitcoin. The nature of these currencies allows for extremely small payments to be made at little or no cost.

Indeed, lowering transaction costs was at the heart of the original case for bitcoin. The white paper written by the mysterious Satoshi Nakamoto starts with a discussion of transaction costs and how the need for third party involvement raises them. Nakamoto’s solution to this problem was to have financial transactions mediated not by a trusted third party but by a distributed public ledger, also known as distributed ledger technology, called the blockchain.

The ledger would be available to anyone participating in the project and its distributed nature, being held on huge numbers of computers around the world, would ensure that discrepancies would be avoided.

Thus the bitcoin system disintermediates not just payment processors but also banks. It reduces transaction costs substantially and allows for micropayments. Ironically, the possibility of micropayments could provide a lifeline for cash-starved news organizations who would be able to charge acceptably small sums for people reading an article (the writer was involved in discussions in the early years of blogging about how to monetize the content of popular blogs – a discussion that always foundered on the rock of payments transaction costs).

Furthermore, cryptocurrencies by their very nature provide an income stream to those who provide the computing power to run the blockchain on their machines, who are called “miners” by analogy with resource extractors. Already startups like San-Francisco firm 21, Inc., are developing dedicated mining modules for Internet of Things devices. This means that some devices could pay for themselves even without an AI.

However, bitcoin and other cryptocurrencies are not for everyone. Ironically for systems based around trust, they suffer from reputational difficulties given their early adoption by criminal enterprises – a problem that was at the heart of the recent decision by the U.S. Securities and Exchange Commission to reject an application for an exchange traded fund based on bitcoin. There is a persistent problem with exchanging assets held in bitcoin for the equivalent in national currencies and the risk that exchanges might prove vulnerable to hacking. And there are issues of governance of the blockchain, which resembles more an old-fashioned commons than a privately managed corporation.

This means that banks and other finance industry players have a chance to find the best of both worlds – combining the low transaction costs of cryptocurrencies with their own operational advantages. Small wonder that most major banks in the developed world are experimenting with distributed ledgers to see if it can increase trust and reduce their own costs. Some are using public blockchains distributed beyond their own environment, others private blockchains they can control, and yet others are developing hybrid systems or other distributed ledger systems based on similar concepts of consensus.

For instance, Westpac is working with shared ledger company Ripple to experiment in cross-border payments, while BNP Paribas is testing distributed ledger technology to enable faster payments. Among non-banks, NASDAQ is using distributed ledgers to help power its Private Market Platform aimed at enabling pre-IPO trading among private companies.

Meanwhile, startups aplenty are providing other use cases. Projects include providing banking services for the unbanked, micropayments for social networks, and escrow services for the gaming industry. Many of these startups will presumably be bought up by banks as their technology starts to show promise, but the possibility remains that one of them will be as disruptive as TransferWise has been to European banking.

Moreover, distributed ledger technology provides a variety of innovative non-financial uses that banks could easily monetize if they become the undisputed masters of the technology.

Blockchains are already being used to provide land titling services in countries like Honduras and Georgia. They can similarly be used to provide proof of ownership for assets as diverse as a diamond, a song, or medical marijuana. Things that are easy to steal, like a diamond, can be traced back to their original owner, while things that are difficult to transfer legally, like a song, can now be sold. People who require proof that they are entitled to own something that other people are not, like medical marijuana, can obtain that proof. Indeed, the technology could even provide a solution to the vexed question of ownership of items derived in times past from endangered species like ivory chess sets, allowing them to be traded legally, even across borders, while not encouraging any new trade.

It should be apparent that there is a Holy Grail in all this development. As Internet of Payments thought leader Roger Bass of Traxiant, Inc., puts it, the potential “killer app” for the Internet of Payments is interoperability. Just as the original internet only reached critical mass when standard protocols were developed to allow the interoperability of systems like email and the world wide web, standard protocols for API, the anonymization process known as tokenization, and similar technologies could be what  turns payments from a headache for developers into a “plug and play” proposition. The rewards to be reaped by the creator of such standards could be immense, or they could present the technology as a gift to the world.

Indeed, the world stands to benefit greatly from the Internet of Payments. As mentioned above, one potential use of the technology is as a facilitator for trade. In the post-Brexit world where trade deals could come to be defined by mutual recognition agreements for regulation, as I argued in a previous issue (Cayman Financial Review 4Q/2016), automated trade will become much easier. If the presumption is no border inspections between countries with such agreements, then proof of carriage can be provided by automated means, probably using some form of distributed ledger, and cross-border transaction costs can be significantly reduced.

There is a policy lesson here for governments. To reap these benefits and grow economies and household wealth, they should provide as innovation-friendly an environment in this arena as possible. The U.K. government has already decided to follow that route with its “regulatory sandbox” for financial technology innovation. Other Anglosphere governments would be well placed to follow suit, given their similar approaches to such things as dispute resolution using the common law. The U.S. could profitably move away from its hyper-regulation of financial activity by doing so.

The emerging Internet of Payments therefore represents both a threat and an opportunity for financial institutions. If banks and payment networks allow nimble technology startups to steal the water they should be rowing in, they could find themselves sinking fast. On the other hand, if they embrace the technology and, crucially, are allowed to do so by regulators, they could find themselves with vastly expanded markets in which to operate, not to mention an increased volume of profitable transactions. Banks will have only themselves to blame if they wonder why self-driving cars are not pulling up to their drive-thru windows.