Michael Story ⚓

Here’s to the hard workers, the early starters, the taxpayers, the people building this country 🫡🇬🇧

A Whitehall civil servant who wants to spend £5m often needs direct sign off from the No10 and is subject to a vast number of checks. A judge who orders 10 or 50x that spending is explicitly banned from considering financial implications

open.substack.com/pub/tomwestgar… Good piece about the barriers civil servants face to doing expensive stuff. Very important to compare this list of vetos to the essentially zero checks on judges spending vastly greater amounts of public money, often influenced by niche political fashions

Amazing blog here. Ofgem accurately forecast the Russia/Ukraine war's effect on energy prices all the way back in 2010! (Speaks to an institutional problem w/ forecasting as a discipline; good but gloomy forecasts too easy to ignore)

We’re good at noticing when politicians make bad decisions for their own *financial* benefit (bribes, corruption, conflicts of business interest). But a lot of truly awful stuff happens when leaders pursue self-actualisation and meaning at the expense of their duty to the country

This is locally a kindness but at scale it conceals knowledge about the rewards on offer for making changes to your life. Examples include: getting married, earning more money, moving to a growing economy, becoming a parent

I think the surprise at how far ahead Oxbridge grads are is an example of an interesting social media pattern where, contra the received wisdom that everyone is constantly exaggerating and bragging, many people with great outcomes tend to play them down a bit to avoid disharmony

Why was the DPP trying to create a blasphemy law in the first place? Officials trying to make a name for themselves? Runaway politicised spirals inside a supposedly impartial org? Influence of staff networks? Terror at potential loss of social control?

All policy is underpinned by an estimate of the future - usually, it’s just implicit. At the Swift Centre, we make those predictions explicit so they can be analysed, interrogated and used. Last month, we announced our "Bridge the Gap" AI Policy competition. Our goal is to provide a low-friction way for people to practice high-stakes policy writing, using rigorous forecasts from the Swift Centre’s professional team as their evidence base. Today, we are publishing the five forecasts that underpin the competition: swiftcentre.org/publicforecast… 🎯 Agentic Capabilities: Will an AI agent autonomously discover a Zero-Day vulnerability? 📉 Workforce Impact: Will UK Finance or IT sectors see a 10% headcount drop? 📉 Market Stability: The "AI Bubble" - Will NVIDIA stock collapse by >50% while the wider semiconductor market falls by >25%? ⚔️ Autonomous Weapons: Will a G20 military officially adopt full "human-out-of-the-loop" lethal autonomous weapons? ☣️ Biosecurity: Will an LLM significantly support a biosecurity breach? Right now, our participants are acting as Policy Advisors to a Secretary of State. Their task is to use these forecasts and rationales to determine if a risk is acceptable and to recommend concrete, actionable interventions in response. The question isn't just what will happen, but what a leader should actually do about it. If you haven’t registered to take part yet, you have until March 20th to join. 👇 Registration link in the comments.

How do you run an international airline in a communist dictatorship whose most productive citizens are constantly trying to escape to better lives? What stops the crew running away as soon as their plane lands? East Germany’s Interflug tried this: - pilots and cabin crew recruited from vetted party-loyal families only - state security minders joined long haul flights to stop crew bolting on arrival - in some airports they had special airside accommodation so crew physically couldn’t enter a capitalist county before the flight home - routes were divided by escape risk instead of geography: junior employees most likely to abscond were allowed only on domestic flights, those with a few more years experience allowed to fly to other communist countries, and only after years of stability and a family back home to worry about were pilots and cabin crews allowed to touch down somewhere with a stock market. Coercion is extremely complex and expensive! Lots more interesting stuff in this video youtu.be/miPRO4BA3vI

Good example of AI impact on electoral politics beyond disinformation/deepfakes (surprisingly still a popular topic in some circles). Even well-funded political parties couldn’t have afforded to make something like this in previous years, now everyone can do it

SoTA's Frontiers Night: Forecasting On Tuesday 9th December, we'll be showcasing tech advancing the state of the art in forecasting the future—of human behaviour, geopolitics, the markets, and technology’s trajectory. We’re fortunate to have the following individuals present and discuss their work at @faculty_ai in London: - Toby Shevlane (@tshevl), Cofounder & CEO of Mantic (@_Mantic_AI) - Ben Warner, Cofounder of Electric Twin - Michael Story (@MWStory), Director of the Swift Centre (@swift_centre) Registration link in the comments below.

It's interesting how many of our most far-reaching (and costly) policies are essentially mistakes or afterthoughts which were implemented with little debate or scrutiny

The lesson is that rules can be completely insane and people enforcing them will just carry on while it’s very difficult for political actors to see the wood for the trees

I don’t know if it actually happened but there was supposedly a raid on Gatwick departure lounges to check whether people were illegally taking a cheap flight

I was learning about the *minimum* pricing rules we had on plane tickets from the sixties to early seventies where discounting fares down to the market rate was illegal- unless you were on an “affinity group” tour eg catholics going to Lourdes.