Prior to the pandemic, I had commenced work on a project that would consider Artificial Intelligence (AI) safety practices in reference to the development of nuclear power. At that time I was struck by the comparatively limited work done on AI safety, considering the potential downside risks.

Below a selection of some news pieces, articles and podcasts that have caught my attention. I might try doing notes like these more regularly if there is interest. The full name is ‘Very Intense Tropical Cyclone Freddy’. From FT: What is remarkable is that the cyclone circled around and hit Madagascar and Moazambique twice . It came back for a second shot. And add in that Malawi is experiencing its worst cholera outbreak on record.

I was fortunate to have the opportunity to chat with Andrew Keen as part of his ‘Keen on’ podcast series. The prompt for our discussion was my recent note: From that, we have a wide ranging conversation about a growing array of examples of institutional failure and state fragility.

Simone de Beauvoir, The Ethics of Ambiguity (1947): - Robert Oppenheimer, Hearing Before Personnel Security Board (1954): - Bill Joy, ‘Why the Future Doesn’t Need Us’ (2000): - Victoria Krakovna, ‘Risks from general artificial intelligence without an intelligence explosion’ (2015): - Ezra Klein, ‘This Changes Everything’ (2023): - Henry Kissinger, Eric Schmidt and Daniel Huttenlocher, ‘ChatGPT Heralds an Intellectual Revolution’ (2023):

I am not someone who notices anniversaries, and yet, each year as 11 March approaches, my mind returns to Tohoku, and the earthquake, tsunami and nuclear accident that struck the region in 2011. I was in Tokyo at the time, and while I was fortunate enough not to be immediately impacted, the experience and subsequently researching the disasters has had a lasting impact.

It is hard not to get a sense that there has been some kind of step change in recent months with AI-related technologies, and the consequences that might follow from the manner in which these are being tested on societies. The point of this note, however, is not to present another judgment or prognosis on chatGPT. Rather, it is to reflect on what we might be losing or forgoing as we rush towards an approximated world.

The notes in which I collect pieces from different sources and reassemble I normally entitle ‘fragments’. Not this time. These are not fragments, but sounds and echoes. Put aside the banal hot takes about chatGPT and listen: there is plenty else to hear. Crack crack. Things are breaking. Not unimportant things. Basic infrastructures and core services. States. And not in one or two places.

Last week marked two years of this Substack, so it seemed like a good moment for a brief update and review of the project. The original aim was for it to be a ‘thought sketchbook’, and that is roughly how it has developed.

Paul Valéry, ‘The Crisis of the Mind’ (1919): - Osip Mandelstam, ‘The Nineteenth Century’ (1922): - Walter Benjamin, ‘Experience and poverty’ (1933): - Stefan Zweig, Diary (Autumn 1939): - René Char, Hypnos (1946):

Australia’s Lowy Institute has just released its 2023 Asia Power Index. Like all such exercises, it is presented to have that feel of it being objective and scientific. It is full of numbers and measures, percentages and tables. In the end, it is effectively just a measured set of judgements, and it can be helpful in that regard.

Noah Smith tells us: Martin Wolf is comfy in his ‘90s time-warp: And don’t worry, the gliberals have not yet been replaced by GPT, although you have to feel that day is getting close. Timothy Snyder announces: And Timothy Garton Ash looks towards a great European future: One does not need to venture far to find similar content. Remember what Francis Ford Coppola said: The Marvel model extends to thinking about politics.