IOUNet - an update

IOUNet is an idea that I have been working on since 2014 when I first posted about the potential for an exchange system that doesn't need money. Earlier this year, I made a lot of progress - with help from Anthropic's Claude - and had a new website iou-net.org that provides a wealth of information about the idea. I did a couple of posts about IOUNet here, on the 17th April, and then again on the 6th May, following a presentation of IOUNet that I gave as a plenary lecture at a conference in Fez. The talk is available on YouTube and has been seen 1167 times so far. 

Rather naively, I had assumed that it would be enough to have the website online and links from the Fez conference for the idea to take off. But that take off never really happened. 

That said, I'm still hopeful and the project has continued to move forward. 

Firstly, I had a planetary level simulation of IOUNet that was running continuously on a single MacStudio at home  for nearly 6 months.  The simulation was complete overkill - with 10 billion users, doing around 20 million IOUs a second continuously. A couple of days ago, I decided to end the run, and relaunch with a more sensible version that again simulates 10 billion users, but where the number of IOUs processed per second is only around 350,000 - corresponding to an average of  3 IOUs per user per day. This is a sensible number because even someone sending IOUs for all their waking activity would only manage around 10-20 a day, whereas there will be a lot of people who only use the system rarely.  You can see the new simulation here. 

But, for the record, here is the final screenshot for phase 1 of the planetary simulation. 

In 179 days, the simulation had handled a staggering 292 billion human years of cooperation! In addition, I was running a complete "Jubilee" where everyone in the system is reset to zero every two hours. The simulation did a total of 1943 complete Jubilees.  As I say - complete overkill. But it makes the point that the calculations needed to run IOUNet are trivial - and can easily be handled by a single desktop computer. Sure, for phase 1, my MacStudio was using 24 of its 28 CPU cores and lots of memory - but it didn't even get hot!

For the new phase 2 simulation,  my MacStudio is hardly working at all - CPU cores usage is a few percent, and the CPU is running on less than 7 watts. And, with 350K IOUs per second, I estimate that the entire traffic could be handled by my 1Gb second fibre optic ethernet connection. 

How is this possible? It's because the basic operations needed to process the IOUs are very simple. Each basic IOU only requires a minimum of 40 bytes of data, although this could increase to 128 or even 512 bytes if you wanted to include lots of additional information. At 40 bytes that's about 110 Mb/s — even the richest 512-byte version would need only a little more than a gigabit. That's still trivial. And the only thing the system needs for  the cancelling operation is each user's net position. At jubilee time, the users with large positive positions get users with large negative positions to do something for them. They then send an IOU for the time spent, which cancels out the debt. If you do that once for every user in the system, you can reset the entire system back to zero in very little time. Magic!  If you log in to the phase 2 simulation at midnight or noon (CEST) you can watch this happen live. It takes a few minutes to go from having over 9 billion users with active accounts to zero. Meanwhile, all the positive cooperative human activity has been done and "paid for" without any need for money to change hands. 

The phase 1 and phase 2 simulations were surprisingly realistic. The 10 billion users are dispatched geographically, and you can click to see the breakdown by continent, country, region and even to the level of the department in the case of France.  For example, in the image below, you can see the number of minutes of human cooperation for Haute Garonne, the department where I live, after just a couple of days of simulation.

In the next figure, you can see a breakdown by activity type - showing the percentage of activity that was in the general category of "Home and Garden", and within that, the proportion that corresponds to "Childcare and Family"

Of course, all of this is totally fictional - I just programmed the system to allocate activity according to a set of plausible settings. But the point is that exactly the same information would be available if the system was running real activity. One of the big advantages of using an IOUNet style system is that, unlike economic analysis that relies on Gross Domestic Product (GDP) which only factors in activity that is paid for using conventional money, IOUNet includes all forms of human activity. Vital activities that include caring for the elderly that simply don't count in conventional economics are now given the importance they deserve. 

Another new feature that I want to point out is the latest version of the basic IOUDemo code that you can find by clicking here  iou-net.org/ioudemo.

The new version has been seriously improved. You set the number of users, and the number of IOUs per users and get a series of IOUs between the users, each with a specific label. In the figure below, we have 5 users and 5 IOUs each resulting in a ranking where Dan, Bob and Emily are all net positive, whereas Carol and Alice are net negative

Is that a problem? No, because if you click the Jubilee button, you can see that everyone's net positions can be cancelled out by 4 new actions. For example, Alice did some "Team Sports Coaching" for Dan that Dan paid for by sending an IOU to Alice. With 5 users, 4 jubilee actions are enough to get everyone back to zero. 

By increasing the number of users and the number of IOUs, it is easy to see that the underlying computations are incredibly simple and efficient. Hundreds of users and tens of thousands of IOUs can be generated in a fraction of a second, and even with large groups, the jubilee cancellation is virtually instantaneous. Even 10 billion users can be cancelled out with less than 10 billion IOUs being transferred -that's what the planetary level simulator does every 12 hours.

But there's a new feature that I have recently added to the Demo of which I am particularly proud. There is now a "Town Council" option that you can click on. When you do, each user sends an IOU to the Council in the form of Civic Dues at the beginning of a run. In the example shown below, all users send an IOU for 60 minutes to the Council. You can also select the percentage of funding used by the Council with three positions - 50%, 100% and 150%. In the screenshot below, I have 10 users, and 5 IOUs each, with council spending set at 50%.

The council has a budget of 600 minutes, but only used 300. So, at jubilee time, the settling process involves not just reciprocal balancing between individual users, but also the council 'paying' for 300 minutes of work from someone who was negative. Specifically they voted to pay Tomaso to organise a community event. 

Importantly, no one is obliged to work for the council. When the council wants something done, it advertises the task and the time it is worth — road & pavement repair, elderly visits, library duty — and picks from whoever offers. The dues are the only obligation, and there are two ways to meet them: do community work during the period, or do work for other people so that the IOUs you hold balance the one you gave the council. Someone whose community work has covered their dues arrives at the jubilee with nothing left to do; someone who has done nothing for anyone arrives in debt and works it off then. And someone who builds up a large positive position during the period is exactly the person who, if the council overspends, gets to tell the council what to do.

I think that this demonstrates how it would be perfectly possible to get many things done in the community, either simply by individuals doing things for each other, or by using a town council to get things done on the basis of decisions made collectively. IOUNet provides the tally that makes all this possible. And the simulations show that the computational costs of doing this are absolutely trivial. 

The remarkable point is that all this can be done without the need for any conventional money. And the fact that no money is involved means that there is no need for borrowing and the payment of interest. There is a debt crisis due to the fact that global debt has now reached a staggering $353 trillion at the end of Q1 2026. There are no official figures on the interest costs of that debt,  but on a reasonable estimate — around 3.7% average across governments, companies and households — the interest bill is of the order of $12 trillion a year.

Note that an IOU is a claim, but it carries no interest and is extinguished by the jubilee rather than rolled over, so no stock ever accumulates to charge interest on. In my opinion, the availability of a fully functional IOUNet based system should at least allow people to ask the question - do we really need to do everything with conventional money?

Pretty much the only thing that IOUNet needs now is users who can try it out. If you are interested, you can join up here https://iou-net.org/join. But if you have questions or comments, why not leave them below.