I’ve had the pleasure this winter of doing some summer reading, a sentence I don’t often get to say. I was lucky enough to head down to Australia to visit family, where for Christmas, my nephew had received Noah Harari’s book Nexus: A Brief History of Information Networks from the Stone Age to AI.

The topic piqued my interest, so I read most of the book before heading back to the UK. Thankfully, it was definitely worth the time and actually proved to be a fascinating lens through which to view our work at Open Climate Fix.

I won’t try to summarise the book, as it’s extensive in its analysis, but I have selected a couple of key themes:

Harari pays much attention to how information is managed in centralised versus decentralised networks.

Centralised networks have one source of information, controlled by a central power. Error correction is performed by the same group issuing the information. In highly centralised networks, information is assumed to be the highest possible quality, as any errors have already been corrected at source.

An extreme example of a highly centralised information network is totalitarian governments, such as Stalinist Russia. There the regime positions itself the source of all truth through a state media system. All information emanates from the centre and the regime is assumed to be infallible - no other opinions or information is therefore taken into account.

Contrast this with decentralised information networks, where information is generated from many sources within the network, and is then passed around in multiple directions. Error correction can be performed by multiple parties.

In political terms, consider a country with a free press: There is a diverse set of opinions on how the country or the economy is faring and how it should be run. Any journalist or commentator can have an opinion whether it is in agreement with the central government or not. Error correcting is done continually through the public assessing truths they feel the most compelling or valid.

The thesis is this: Centralised or dictatorial information networks are simple and decentralised ones are complicated.

However, error correction is weaker in centralised networks and stronger in decentralised networks as there are more checks on the information. In fact Harari defines democracies as distributed information networks with strong self-correcting mechanisms rather than equating democracies to political systems who solely run elections. By contrast totalitarian regimes have highly centralised information networks to help them maintain control.

The electricity grid is extremely complex. Sometimes called the most complex machine ever created by human kind, in some cases grids stretch across multiple countries and impact hundreds of millions of electricity consumers. To operate the electricity grid, supply and demand must be kept perfectly in balance at every point in time, and voltage, frequency and power quality must be continually maintained.

Before the 1990s, each electricity grid was managed by a single body - the system operator. The system operator would be responsible for instructing generation on or off to match electricity demand while ensuring power quality was maintained at the lowest cost. In order to perform this task the system operator would amass a huge amount of information on the present and future state of the grid, including the coming minutes, hours and weeks. Using this data, the system operator would generate predictions, such as when a generator should turn on. This was a highly centralised network with few other parties collecting and generating information. Further, almost all the error correction was the responsibility of the system operator - a considerable burden.

Now around the 1990s, a wave of deregulation, decentralisation and in many cases, privatisation of electricity grids, swept through the major western countries. Electricity markets were established and electricity was positioned as a commodity to be bought and sold at the best price. This was motivated by popular policy ideologies of the time that viewed private enterprise control as a way to drive down costs through market competition.

Similar to how Harari defines democracy, we can view the market based electricity grid as a decentralised, self-correcting information network. Market participants - generators, battery operators, electricity retailers and traders - all assimilate information on the market and decide prices they believe electricity should be bought and sold at. This information is implicitly exchanged via the market traded price. Error correction is performed by all market participants, as a generator who sells their power at the wrong price will either have to put effort into improving their price estimation or go out of business. System operators, like governments in democracies, still have to make the most important decisions about how to manage the grid, but the data in the decentralised energy information network helps correct errors, driving better decisions on how to operate the grid.

The modern grid continues to become even more complex. The number of generators is moving from dozens to millions, due to both the electrification of transport and the integration of renewable energy. The amount of information required to run the grid is becoming unbelievably greater than even ten years ago. Hence, having energy information networks working effectively is more important now than ever, and much more important than the architects of the market in the 1990s could have envisaged.

One of the key types of information in the electricity information network are forecasts of how much electricity will be used by consumers and generated by weather dependent generation: Solar, wind and hydro-electricity.

In a healthy decentralised energy information network, we have not just one forecast, but multiple. Error correction takes place via generators, electricity retailers, traders and system operators, all assimilating which forecasts are the best in any given situation and taking actions based on the most accurate. The best forecasts will flourish and the worst will perish.

Consider the Great Britain (GB) electricity grid. At Open Climate Fix, we believe we are producing the most accurate forecast of short-term (hourly) solar generation prediction. We are providing that forecast to the National Energy System Operator (NESO), and they are using it to save 100s of thousands of tonnes of carbon and tens of millions of pounds per year.

But that alone isn’t going to maximise the benefit to the grid. In a decentralised energy information network, we want as many people to use ours, as well as other, high quality forecasts. In this way the grid will operate more efficiently. So we also provide our leading UK National Forecast, Quartz Solar, to other key players in the energy system, from solar power generators, smart home optimisers, battery operators and energy traders.

This is a model we plan to replicate in other grids, providing high quality energy information to as many users as possible. By supporting a diverse, decentralised energy information network, we can also support the efficient running of the grid, as a free press supports a democracy.