Aim for nuclear peace first

PANews reports: 

In April 2026, a hair dryer was used to heat a meteorological sensor at Paris Charles de Gaulle Airport, causing abnormal temperature spikes. The attacker exploited rule vulnerabilities in Polymarket's prediction market, winning $34,000 in prizes. Polymarket relies on a single data source and does not revise data after settlement, making physical intervention possible. After the incident was exposed, Polymarket quietly changed the data source without publicly acknowledging the vulnerability. This is known as a 'physical oracle attack,' with low cost but high returns. A hair dryer "blew" $34,000 away from Polymarket, PANews, 23 April 2026

Campbell's Law states that as soon as you use a specific number to judge people's success, people will start "gaming the system" to make that number look good, even if it ruins the actual quality of the work. The hairdryer is a relatively trivial example; a more serious one is 'teaching to the test'. It's a problem that afflicts current policymakers and could, potentially, derail a Social Policy Bond regime, which depends entirely on the definition and targeting of quantitative variables. A solution would be to target goals that are, or are inextricably lined to, what we actually want to achieve. This could be made to work well for bonds that target broad goals at a highly aggregated level and in the long term. Take literacy: if we want to see improvements in the literacy of, say, 7-year old children in a country over the next decade, we could sample children of that age over a wide area, waiting until the end of the decade to take the sample. Nobody would know in advance the identity of those children, so only a genuine improvement in the national literacy level of 7-year olds would signify that the goal had been achieved. 

A current preoccupation of mine is the urgent need for conflict reduction. Coming up with a range of indicators that we could target for world peace, or even regional (such as the Middle East) peace could be done: I envisage an array of indicators, including: numbers of people killed directly by conflict; people seriously maimed; people made homeless; expenditure on weaponry; numbers in armed forces; numbers of schoolbooks inciting hatred; and many others. All these indicators could be debated, then rigorously defined for targeting purposes. This would take time, but would be worth doing, given the payoffs that we would target, such as peace in the world (or region) sustained for at least thirty years. But simpler, and only slightly less ambitious, would be to target nuclear peace: detection of a nuclear explosion is a relatively simple matter: from Wikipedia:

There are many different ways to detect a nuclear detonation, these include seismic, hydroacoustic, and infrasound detection, air sampling, and satellites. They have their own weaknesses and strengths, as well as different utilities. Each has been used separately, but at present the best results occur when data is used in tandem, since the energy caused by an explosion will transfer over to different mediums. Source

I've written about Nuclear Peace Bonds here. I'd hope that, once nuclear peace were targeted, we could then begin the more arduous work on defining exactly what we'd want World Peace Bonds or, say, Middle East Peace Bonds to target.