Newsletter 82

Slowing down AI, China/Taiwan tensions and good vs bad writing

Apr 16, 2023

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. (1,262! - welcome all new arrivals)

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 irregular (~~weekly~~) missive!

Reading list - the best stuff to read

(The best reads I’ve come across, with excerpts, links, authors and how long it will take to read. Climate change, Ukraine and China are the stories at the top so are semi-permanent)

🌏 Climate change & biodiversity destruction

A really interesting look at rust batteries (or iron-air). Also see pod recommendations.

Woodford then turned to an unlikely source for inspiration: the disposable hand warmers he used on cold days growing up in Pennsylvania. Crushing these little sacks filled with crystals begins a process of rusting iron—it essentially reacts with oxygen to form iron oxide—that’s sped up with chemicals. The reaction releases energy in the form of heat. Woodford and Chiang, Form’s science brains, thought that with the right setup they could use the same reaction to release energy as electrons instead.

Finland has finally turned on its new nuclear power station. While Germany has ended its use of nuclear energy.

Construction of the 1.6 gigawatt (GW) reactor, Finland's first new nuclear plant in more than four decades and Europe's first in 16 years, began in 2005. The plant was originally due to open four years later, but was plagued by technical issues.

We need to draw down carbon—not just stop emitting it

So how can we move forward to develop and scale carbon removal solutions without creating a risk of delaying the needed cuts in climate pollution?
We should start by being skeptical about any portrayal of carbon removal as a silver bullet that avoids the need for deep emissions cuts. We should ensure that any net-zero commitments from governments and corporations assign a substantially greater role for emissions reduction than carbon removal. We should be wary of any modeling that bases high use of future carbon removal on the idea that spending money to tackle climate change in the future is easier than spending money today. And as a simple rule of thumb, we should aim for a world where we reach net zero by cutting more than 90% of current emissions and removing less than 10%.

🇨🇳 China / Taiwan

A note to readers: it’s become increasingly difficult to track all of the developments in relation to this subject. There are almost daily announcements throughout the APAC region (many of them military ones) that in one way or another will have impact (or help lead to) future conflict in the region. This is not a good sign: when you are doing pattern recognition and try to scenario plan, a plethora of new developments is a pattern in itself. I am not optimistic that conflict will be avoided and, if anything, conflict could arise sooner than previously forecast (ie before the latter half of the decade).

I might need a “still in my tabs” just for APAC developments.

A really good long read on how China has tried to steal secrets from US companies. (There’s also an audio version (54 mins)).

The messages in Xu’s iCloud account enabled investigators to make another damning discovery. Xu had helped coordinate a cyberespionage campaign that targeted several aviation technology companies. Those attacks — described in a report by CrowdStrike, a cybersecurity firm — started in 2010, shortly after the state-owned Commercial Aircraft Corporation of China (COMAC) announced that it had chosen a joint venture between G.E. Aviation and Safran to supply a custom-made engine for China’s first domestically manufactured commercial airliner, the C919. The plan behind the campaign, which was directed against Honeywell, Capstone Turbine and Safran, among others, became clear only later when security researchers connected the dots. “When I started putting all these victims together — I was like, OK, these are all component manufacturers for different pieces of the C919,” Adam Kozy, a cybersecurity expert who runs the security firm SinaCyber and was the lead author of the CrowdStrike report, told me. Although COMAC was prepared to procure components needed to build the aircraft from these companies, the Chinese government was evidently also working to steal intellectual property from those suppliers in order to make domestic manufacturing possible in China, according to the report.

🇺🇦 Ukraine / Russia

A note to readers: the leaks of intelligence gave greater detail than we had previously known about the three army corps being created in Ukraine to begin the Spring counteroffensive.

We now know to some degree the composition of the 9 brigades being created, and what they are equipped with. In particular two brigades appear to be “tip of the spear” brigades composed of only Western equipment and lacking artillery support. The goal will be to break through Russian lines, and force them into a lacklustre retreat. The wide availability to fight at night could also be key for a Ukrainian success.

The timeline for the beginning of an offensive - likely mainly in the south in the direction of Berdyansk and/or Tokmak - is expected any time between late April and late May.

Ukraine’s Best Chance - A Successful Offensive Could End the War With Russia

There are different ways to imagine the less-than-ideal outcomes, but consider two. In the first, Russia would retreat to pre-invasion battle lines, retaining Crimea and the land in the Donbas that the two pro-Moscow statelets occupied before the war. In the second, Ukraine would recover all of Zaporizhzhia, and therefore the Azov Sea littoral, and Kherson, including the Russian-occupied segment of the Black Sea coast. Russia would retain Crimea and much of the Donbas—Luhansk Province and the part of Donetsk Province that it controls—but relinquish the land corridor to the peninsula it now has through the southern parts of Donetsk and Zaporizhzhia Provinces.
Even the latter scenario would amount to a significant Ukrainian victory. Putin’s plan to install a quisling government in Kyiv and partition Ukraine by annexing its north, east, and south, including the Black and Azov Sea coastlines, would have failed. So would his bogus attempt last September to annex Donetsk, Luhansk, Kherson, and Zaporizhzhia Provinces, three of which he only partly controlled at the time. Ukraine would retain approximately 93 percent of its pre-invasion land area and about 89 percent of its pre-2014 territory. Although a settlement along these lines would fall well short of Kyiv’s ideal denouement, it would spare Ukraine further death and destruction, allow it to proceed with the mammoth task of postwar reconstruction, and enable millions of refugees and internally displaced people (together, nearly 14 million) to return home.

Binkov has a good analysis on Ukrainian air defence and missile numbers, based on recent leaks of US intelligence documents. (20 mins)

This is a very good detailed analysis of the battle for Hostomel airport right at the start of the war. Contains some footage I had not seen before. (22 mins)

This is an excellent discussion from two of the best Ukrainian analysts, watch the whole thing:

🤓 Augmented reality

Apple plans to launch its new Reality headset in June - its first big new product line since 2008.

The mixed-reality headset will be the star of the show, but the annual conference will be packed with other products too. That includes new Mac laptops and the biggest update to the Apple Watch’s software since the first version was introduced in 2015, I’m told. Altogether, the company is planning to unveil: the Reality headset, the first major new Apple product category in nearly a decade a new xrOS operating system and software development kit new MacBooks iOS 17 iPadOS 17 macOS 14 a major watchOS 10 update

🧠 Slowing down AI

A really good long read on the AI “race”.

Unfortunately, I think the race will continue. It will likely take a major misuse event — a catastrophe — to wake up the public and governments. I personally plan to continue to invest in AI start-ups that focus on alignment and safety or which are developing narrowly useful AI. But I can no longer invest in those that further contribute to this dangerous race. As a small shareholder in Anthropic, which is conducting similar research to DeepMind and OpenAI, I have grappled with these questions. The company has invested substantially in alignment, with 42 per cent of its team working on that area in 2021. But ultimately it is locked in the same race. For that reason, I would support significant regulation by governments and a practical plan to transform these companies into a Cern-like organisation.

Bonus: also checkout John Oliver on AI

Second bonus:

✍️ On writing

I really enjoyed this exploration of good vs bad writing, using Andor (excellent show!) and Obi-Wan Kenobi (not great!) to compare: (20 mins)

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

(A serialised section that started with Greek Tragedy and moved to philosophy. Something to spark ideas. Feel free to go backwards!).

On hiatus!

Documentary

(A good thing to watch - also serialised - so feel free to go back through past editions!)

We’re on Part 2 of Auschwitz The Nazis and the Final Solution.

Podcast(s)

(The best stuff I’ve listened to, or been recommended by subscribers)

This new podcast had a good episode on the future of energy. (59 mins)
A very good and detailed discussion on the semiconductor industry (inc the geopolitics of Taiwan’s role in semiconductors). (57 mins) (Bonus: also read I Saw the Face of God in a TSMC Semiconductor Factory)
There are two good recent episodes of Geopolitics Decanted. First is a discussion about China/Russia (58 mins) and the second about Taiwan and its defences. (63 mins).
There was a fascinating look at Form Energy’s Iron-air battery technology. (25 mins)
A good discussion with Aric Toler, the key journalist (together with the NYT visual investigations team run by a subscriber to this newsletter, Malachy Browne) who uncovered the alleged source of US intelligence leaks. (48 mins)
The New Scientist pod looks at several subject, including figuring out what it will take to feed bigger global human populations, and how to balance that with biodiversity destruction.