Newsletter 52

Troll farms, AUKUS subs, learning fast

Hi, I’m Gavin. This is my experimental newsletter that explores thinking - how we might think better and learn together as we do so.

I explore several key topics through the lens of several core themes: systems thinking, scenario planning, trends, and cross-disciplinary innovation. These often relate to key issues: climate change, pandemics, astronomy, physics, health, history, philosophy, culture, rocketry, conflict, the impact of technology on society and more (lol!). With a larger question behind it all: how do we progress and how do we progress better?

I hope you like where we go. (827- nope - 834 of us now! - welcome all new arrivals!) I would love to get to 1,000+ subscribers by the 52nd edition so if you like the newsletter please do tweet about it!)

Gavin

PS: if you like the newsletter please share it! (And I always appreciate tweets about it too!) 🙏

Share Gav's Newsletter

If you support this newsletter and really like it you can also become a paid subscriber - you get nothing different except the satisfaction of knowing that you’re supporting this current weekly missive!


Reading list - the best stuff to read

🌏 Climate change & biodiversity loss

  1. A reminder for this section:

  2. Hopefully these trees are safe. (3 mins by Rosanna Xia)

The 21,777-acre lighting-sparked wildfire — dubbed the KNP Complex fire after the Colony and Paradise fires merged into one — grew by more than 3,900 acres overnight, but officials said Sunday that hundreds of firefighters have valiantly kept key areas of the forest under control. The park is located east of Fresno.

In an upbeat report Sunday, fire officials said they were feeling fairly confident about protecting the Giant Forest, home to thousands of towering sequoias. Numerous well-established walking trails meander through this iconic part of the park, so firefighters have been able to move around and work from multiple locations.

  1. The writing is one the wall for the current scale of beef and dairy. And it’s not just their impact on emissions, it’s biodiversity loss too. (4 mins by Damian Carrington)

Support for the “outsized” meat and dairy industry in rich countries must be reduced, while subsidies for polluting chemical fertilisers and pesticides must fall in lower-income countries, the analysis said.

The report, published before a UN food systems summit on 23 September, said repurposing the subsidies to beneficial activities could “be a game changer” and help to end poverty, eradicate hunger, improve nutrition, reduce global heating and restore nature. Good uses of public money could include supporting healthy food, such as vegetables and fruit, improving the environment and supporting small farmers.

“The true costs of our food system have been hidden for too long,” said Morgan Gillespy, the programme director at the Food and Land Use Coalition. The damage caused to nature by subsidy regimes was $4tn to $6tn, according to a recent review, she said.

“Changes in subsidy regimes are likely to be politically controversial and could spark protests among farmers and other groups,” Gillespy said. “But just because it is hard, doesn’t mean it shouldn’t happen. The facts are now clear.”

  1. Meanwhile the EU and US have pledged to reduce methane emissions. (2 mins by Fiona Harvey)

The pact between the US and the EU sets a target of cutting at least 30% from global methane emissions, based on 2020 levels, by 2030. If adopted around the world, this would reduce global heating by 0.2C by the 2040s, compared with likely temperature rises by then. The world is now about 1.2C hotter now than in pre-industrial times.

  1. We need to be moving faster, and some models are not really understanding interdependency, or ambiguity. (4 mins by Kate Mackenzie)

Pitman, director of a multi-university center on climate extremes in Australia who has also contributed to previous IPCC reports, points to financial stress tests and macroeconomic modeling as one example of where this kind of thinking goes wrong.

The instruments are meant to estimate the effects of higher levels of warming, but “if it tells you you are resilient at 4°C, that doesn’t mean you’ll be okay. It means your analysis is crap,” he said. It’s like asking “what would happen if you jumped off a 50-meter cliff and then finding you’d land at the bottom and you’d be fine.”

Talking about that uncertainty and the limits of what modeling can currently show has long been a double-edged sword. Climate deniers have pounced on it as a way to discredit climate science.

In fact, the opposite is true. “For me, the remaining uncertainties should be used as an argument for acting as fast as we can,” said Seneviratne.

🦠 COVID-19

Zeynep had a very good thread on whether we should be rolling out booster shots:

🇨🇳 China - AUKUS

This was big news and fits into this newsletter’s frequent look at the development of China’s military and how the rest of the world is reacting (in particular the US). France is not happy, and it could easily have delivered nuclear powered attack submarines to Australia - but Australia decided to break the partnership and begin nuclear submarine development with the US and the UK for 8 new subs. Of course, this is also about money.

  1. Just days after the AUKUS partnership was announced, the UK announced the beginning of its programme to replace its current Astute class subs. No doubt this model may be for sale to the Australians. As may be the US Virginia class.

  2. The Drive also has a good overview of the announced deal.

  3. Why would Australia want nuclear powered subs instead of diesel?

  1. Binkov has a good overview too (don’t let the puppet fool you, always good analysis here). He speculates that Australia may receive UK expertise to build an Astute class or similar submarine in Australia (likely in Perth). (10 mins)

🧠 Brains - neuron computation

Well this is interesting. Nature isn’t too bad at figuring computation out. (5 mins by Jason Dorrier)

In a fascinating paper published recently in the journal Neuron, a team of researchers from the Hebrew University of Jerusalem tried to get us a little closer to an answer. While they expected the results would show biological neurons are more complex—they were surprised at just how much more complex they actually are.

In the study, the team found it took a five- to eight-layer neural network, or nearly 1,000 artificial neurons, to mimic the behavior of a single biological neuron from the brain’s cortex.

Though the researchers caution the results are an upper bound for complexity—as opposed to an exact measurement of it—they also believe their findings might help scientists further zero in on what exactly makes biological neurons so complex. And that knowledge, perhaps, can help engineers design even more capable neural networks and AI.

🧠 Innovation - how to learn fast

This is a really interesting and thought provoking thread on how to learn via Cognitive Task Analysis.

🏛 Society - Facebook troll farms

Great reporting as ever from Karen Hao - this time on Facebook troll farms. (8 mins by Karen Hao) (also see podcast section for recent WSJ revelations)

The report looks specifically at troll farms based in Kosovo and Macedonia, which are run by people who don’t necessarily understand American politics. Yet because of the way Facebook’s newsfeed reward systems are designed, they can still have a significant impact on political discourse.

Sophie Zhang, a former data scientist at Facebook, revealed that it enables global political manipulation and has done little to stop it.

In the report, Allen identifies three reasons why these pages are able to gain such large audiences. First, Facebook doesn’t penalize pages for posting completely unoriginal content. If something has previously gone viral, it will likely go viral again when posted a second time. This makes it really easy for anyone to build a massive following among Black Americans, for example. Bad actors can simply copy viral content from Black Americans’ pages, or even Reddit and Twitter, and paste it onto their own page—or sometimes dozens of pages.

Second, Facebook pushes engaging content on pages to people who don’t follow them. When users’ friends comment on or reshare posts on one of these pages, those users will see it in their newsfeeds too. The more a page’s content is commented on or shared, the more it travels beyond its followers. This means troll farms, whose strategy centers on reposting the most engaging content, have an outsize ability to reach new audiences.

Third, Facebook’s ranking system pushes more engaging content higher up in users’ newsfeeds. For the most part, the people who run troll farms have financial rather than political motives; they post whatever receives the most engagement, with little regard to the actual content. But because misinformation, clickbait, and politically divisive content is more likely to receive high engagement (as Facebook’s own internal analyses acknowledge), troll farms gravitate to posting more of it over time, the report says.

🍰 Biology - fat

Basically everything you think you know about fat is likely wrong. This is an excellent read. (18 mins by Gary Taubes)

This energy-in-energy-out conception of weight regulation, we argue, is fatally, tragically flawed: Obesity is not an energy balance disorder, but a hormonal or constitutional disorder, a dysregulation of fat storage and metabolism, a disorder of fuel-partitioning. Because these hormonal responses are dominated by the insulin signaling system, which in turn responds primarily (although not entirely) to the carbohydrate content of the diet, this thinking is now known as the carbohydrate-insulin model.

Its implications are simple and profound: People don’t get fat because they eat too much, consuming more calories than they expend, but because the carbohydrates in their diets — both the quantity of carbohydrates and their quality — establish a hormonal milieu that fosters the accumulation of excess fat.

🏠 Society - housing

This piece caused quite a bit of debate on Twitter. In principle the diagnosis is correct. One criticism is that if focusses too much on the market delivering housing - or that it fails to mention housing delivery as a function of the public good - just like delivering other types of public infrastructure (cycle paths, roads, railways). (Disclosure: I went to college with and have had pints with the lead author lol).

It riffs on the same themes as some of my own Twitter threads. (20 mins by Sam Bowman et al)

Where you live affects nearly everything about your life – where you work, how you spend time off, who your friends and neighbours are, how many kids you can have and when, and even how often you get sick. Most people’s most valuable asset is, by far, their own home. And housing is so important for the overall economy because it determines the location and supply of the most important ‘resource’ of all: people.

Bits of my thread here from early September:

🚀 Space - Inspiration4

We looked at this a few weeks back. The entire mission was a complete success and the crew returned to Earth yesterday. Watch their return here. Also check out the Netflix documentary.

Philosophy Corner (a journey through thinking about thinking every week)

There’s a lack of good content on YouTube for this stuff so I’m gonna switch to podcasts. The History of Ideas is a good place to start. First up - Hobbes. (60 mins)

Thomas Hobbes’s Leviathan (1651) reimagined how we could do politics. It redefined many of the ideas that continue to shape modern politics: representation, sovereignty, the state. But in Leviathan these ideas have a strange and puzzling power. David explores what Hobbes was trying to achieve and how a vision of politics that came out of the English civil war, can still illuminate the world we live in.

Documentary

The third part of the Adam Curtis documentary. (59 mins)

Podcast(s)

  1. The WSJ has been doing an excellent investigation into Facebook based on a throve of leaked documents. It’s worth listening to each of the available episodes in turn. Facebook’s corporate behaviour gets worse with each episode. Part One, Two, Three, Four (about 30 mins each)

  2. Ezra Klein had two good episodes.

    The first is the chat with Tyler Cowen (80 mins). I don’t fully buy into Cowen's libertarian streak but he has always challenged my preconceptions. You may also recall his 2019 essay with Stripe’s Patrick Collison on the subject of progress. Cowen’s love of classical music is great too. Worth your time - and it’s largely about issues and things this newsletter is persistently interested in.

    The second is about economics with Adam Tooze. 8/10. (86 mins)

Still in my tabs

Looking for Interstellar Monuments

One hiker’s peak of desperation

Ancient Camel Sculptures Are Older Than the Pyramids and Stonehenge, Scientists Discover

Artists can use artificial intelligence as a creative tool to transmit themselves from the afterlife.

How China Avoided Soviet-Style Collapse (from the aforementioned Adam Tooze)

Can America's Fastest Supercomputer Defeat Covid for Good?

Vitamins or whole foods; high-fat or low-fat; sugar or sweetener. Will we ever get a clear idea about what we should eat?

A Massive Subterranean ‘Tree’ Is Moving Magma to Earth’s Surface

New type of dark energy could solve Universe expansion mystery

For the week that’s in it: this is good

Loading more posts…