Computer In Drydock

One of the motivations for building one’s own computer is that if something goes wrong with it one should be able to repair it. Alas, I am putting this to the test right now. After booting perfectly for 13 months, the workstation that I built in May 2025 started having trouble booting. At first this wasn’t a big deal — I would just power it up several times and it would boot. Watching the POST (power-on self test) LED indicators it was clear this problem was hardware related — the boot failure was occurring before the computer loaded in the operating system code. Just before we left for Spain, the computer changed from booting after several times to only booting once in a blue moon.

After reading up on POST failures and testing all the components I could think I concluded by elimination the computer has a motherboard problem. The good news is that the motherboard is the least expensive electronic component in the system (the extraordinary increase in the price of memory is part of that) and that it is still under warranty; the bad news is that the computer needs to be taken mostly apart to remove the motherboard so it can be shipped off. I spent yesterday afternoon dissecting my computer and today I went to the shipping store downtown to send the motherboard off to the manufacturer (total shipping cost $24). We will see if this leads to a working computer in a couple of weeks. These are very bad times for the PC building hobby, all part of the enclosure movement occurring all across information technology.

Lessons from the East India Company

In my last post I wrote about William Dalrymple’s Return of a King: The Battle for Afghanistan, 1839-1842. This is the second book I’ve read by Dalrymple, the first being The Anarchy: The Relentless Rise of the East India Company, published in 2019. Dalrymple recently wrote an opinion piece for the New York Times entitled “What It Took to Dismantle the Most Powerful Company in the World” discussing the parallels between the East India Company and today’s Big Tech.

The East India Company did a really good job of subverting the British state. But in 1772 the East India Company ran out of money, and started asking for bailouts. The company was too big to fail, and its stock was owned by 40% of the House of Commons. In 1773 trying to claw some money back the British state came up with the Tea Act, shipping the tea to America to undercut the colonial merchants. This provoked the American Revolution. Dalrymple cites John Dickinson:

This “almost bankrupt company,” having been occupied in wreaking “the most unparalleled barbarities, extortions and monopolies” in Bengal, had now “cast their eyes on America, as a new theater, whereon to exercise their talents of rapine, oppression and cruelty,” wrote the American John Dickinson. Within three years, America declared independence.

In 1788 Parliament impeached Warren Hastings, former governor-general of Bengal, but he was eventually acquitted. It took another 70 years before the British tamed the East India Company, only after the Indian Rebellion of 1857 with 800,000+ lives lost. At that point the state dissolved the company’s navy and absorbed its army into the Crown.

When Dalrymple finished The Anarchy in 2019 he noted thankfully that today’s corporations did not have private armies. Things aren’t quite as clearcut in 2026. Starlink operates at the whims of Elon Musk, technology that has played a significant role in the fighting in Ukraine. Palantir holds a $1.3 billion contract with the Pentagon for its AI targeting program. Google and Amazon hold a $1.2 billion contract to supply AI and cloud services to the Israeli government.

But the British state did in the end tame the East India Company. In our own times, the Chinese state has cracked down on its big corporations, e.g. Alibaba and Tencent. Still, as Dalrymple says:

History shows that when a company starts acquiring the attributes of a state — the revenues, the courts, the treaties, the capacity to make war and peace — its transformation can damage the parent country at least as much as the territories it exploits. Its lobbying corrupts the legislature. It captures the treasury, because the corporation’s solvency becomes a public interest. It converts foreign policy into an extension of a balance sheet. And it shows that the threshold, once crossed, is astonishingly hard to recross. The window in which any of this can be undone closes fast.

An Interlude On Public Transport

Apart from the excursion up to Vall de Boí via rental car, basically all of travels throughout Catalunya were via public transport. How did this go?

The train (Renfe). Train service in Spain in general and Catalunya specifically can be divided into high-speed rail (the AVE) and regional rail lines. Spain has invested a lot into building out the AVE — Spain has more mileage of high-speed rail lines than any other country besides China — but this has been at the cost of lack of investment in the regional rail network. We rode the AVE twice, once to go from Barcelona to Girona, and the other time to go from Girona to Lleida. These trips were fast as advertised but navigating the stations, particularly Barcelona-Sants, was tricky. We rode the regional rail three times: first to get from Barcelona airport to downtown, then to go from L’Hospitalet d’Infant to Tarragona, and then from Tarragona to Barcelona. The latter two trips were miserably overcrowded and I had to stand all the way both legs. The trains however were on time. There was an episode of Easy Catalan a year ago where the hosts interviewed people on the street in Vilafranca del Penedès on whether they took the train anywhere. The answer was almost without exception no, the trains in Catalunya were just too slow and unreliable.

The metro in Barcelona. We rode it extensively the last three days of our trip. It is a large system and we found that changing between lines took a lot of effort in going up and down to the different underground levels via escalator or elevator. Our base for our stay in Barcelona was a hotel on the metro line running directly to the airport so we had to do a lot of line changing. We also learned that summertime is when they do maintenance on the system; specifically on our first trip from the hotel to downtown we discovered that the L1 line which had the easiest interconnection to downtown was closed for five stations, replaced by a special bus. We then had to route around this discontinuity in the metro lines the rest of our stay. One thing that made our travels easier in Barcelona was buying a three-day Hola Barcelona pass, unlimited travel for that time period on the city buses and the metro. Another fun diversion was taking the funicular up Montjuïc that is directly connected to the metro system.

Buses. They are harder to research than trains or the metro but the buses can be quite enjoyable. Returning to our hotel in Barcelona the last day we found a bus that ran from Paral·lel straight to our hotel — it was much more efficient than the metro plus it skirted the lower slopes of Montjuïc, passing by some magnificent architecture on the way. And my favorite transit experience the entire stay in Catalunya was taking the bus from Lleida south to Mora d’Ebre. There are only two buses a day on that route but the trip passed through some really interesting countryside. Also the bus station in Lleida is right next to the train station which is quite convenient.

Taxis. We used these fairly often for short legs. In general taxis are easily found but we had a stressful twenty minutes waiting for a summoned taxi to take us from our hotel in Girona to the train station to catch the AVE.

All of which has taught me to pay more attention to bus transport while traveling. A good guide for researching buses worldwide is busmaps.com. As an example, this shows all the agencies running buses through Lleida.

Zettlring After Nine Months

Back at the beginning of last October I reported on on how I was experimenting with a notetaking method called Zettelkasten, and had just starting using a software package called Zettlr to do so. I am happy with how this project has gone, and am using the system to store short thoughts, extended notes, and drafts of pieces of writing. As of now I have about 132 entries in my Zettlr workspace. This count may not live up to my hopes from then of one entry a day, but it is continuing to grow in diverse directions.

Each entry in Zettlr is a single markdown file — markdown is a simple structured text format — and a workspace in Zettlr consists of a nested directory of markdown files. I am not at all worried about when I end up having thousands of files instead of hundreds. The Zettlr software seems to scale well, and in addition directories full of markdown files have become a common way to organize information in our brave new world of AI-augmented knowledge management systems.

On Green AI

Contrary to the normal progression of things, the computer that I built a year ago in May has increased in value by about $1150. This is because of the extraordinary demand being placed on supplies of memory and storage components of all sorts by the buildout of AI data centers — most spectacularly in my case, the 64 GB of DDR5 RAM which I purchased in May of 2025 for $165 is now priced at $915. With my workstation, I am continuing to play around with local AI models. The quality of these local models has increased dramatically over the past year, which is leading many developers to explore how they can be used in preference to the huge models from providers in the cloud such as Anthropic, OpenAI, and Google.

One could make the argument that running AI models locally is more sustainable than invoking the resources of huge server farms every time one calls upon AI. But are there other threads that might constitute “green AI”? Sharon Stein has a good piece about this topic on resilience.org. She identifies three visions for green AI. The first of these technical greening. This means running AI with fewer resources, for instance by using smaller models that demand less energy and materials. The second of these is ecological intervention. This refers to developing AI applications that help us understand and ameliorate environmental change, for instance by optimizing agricultural resource use. The third vision is relational reorientation. This focuses on questioning the assumptions around nature, intelligence, and relationships that are built into contemporary AI. A lot of this vision comes from Indigenous perspectives: one interesting research program in this vein is called Abundant Intelligences (also see an open access paper titled Abundant intelligences: placing AI within Indigenous knowledge frameworks).

On a more amusing note, here is a literally hand-cranked solution to the cost and energy use of contemporary AI systems.

Disarming AI

Today I read through the entirety of Pope Leo XIV’s newly issued encyclical Magnifica Humanitas, subtitled On Safeguarding The Human Person In The Time Of Artificial Intelligence. It’s a very rich document which anyone who is concerned with the ethical and moral issues around AI should add to their reading list. Pope Leo’s namesake predecessor Pope Leo XIII in 1891 wrote the encyclical Rerum Novarum about the condition of the working class. In its first two chapters Magnifica Humanitas traces the history of Catholic social teachings from the time of Rerum Novarum to the present. These themes of human dignity and working towards the common good structure the discussion of AI beginning in chapter three.

Whether through direct study or not Pope Leo XIV has absorbed a lot from the discipline of science and technology studies. For instance there is no such thing as morally neutral AI:

In reality, every technical tool embodies choices and priorities through what it measures, ignores and optimizes, and how it classifies people and situations. If a system is designed or used in a way that treats some lives as less worthy, or excludes them without the possibility of appeal, then it is not merely a tool “to be used well,” since it has already introduced criteria that contradict the inalienable dignity of the human person. (Section 104).

Discussions of digital colonialism are readily found among academic critics of information technology but it is much weightier to hear this topic coming from the leader of the Catholic church:

Even today, colonialism assumes new forms. It no longer dominates only bodies, but appropriates data, transforming personal lives into exploitable information. Entire regions, especially those marked by structural fragility and limited geopolitical relevance, are currently subjected to a new mindset of extraction: that of health data, epidemiological profiles, genetic maps and demographic information…Here lies one of the most urgent moral challenges of our time: to ensure that shared knowledge becomes a true common good rather than an instrument of dominance. This requires restoring to individuals not only the data that describes them, but also the ability to decide how it is used, by whom and for whose benefit. Otherwise, the digital age will not be post-colonial, but colonial in another form. (Section 178).

Pope Leo XIV is fond of the expression “to disarm” and this phrase is at the heart of his critique of AI:

To disarm means discrediting the assumption that technical power automatically confers the right to govern. To disarm does not mean rejecting technology, but preventing it from dominating humanity. It means freeing technology from monopolistic control and opening it to discussion and debate, therefore making it human-friendly and restoring it to the plurality of human cultures and ways of life. Our task today is not only ethical or technical. It is ecological in the deepest sense, for it concerns a new dimension of our common home. AI is already an environment in which we are immersed, as well as a force with which we must engage. For this reason, merely regulating it is insufficient; it must be disarmed, welcoming and accessible. (Section 110).

It is easy to find blanket rejections of AI in many quarters but I am heartened that there is a lot more nuance in this encyclical. Technical innovation is part of the divine act of creation but going along with that comes a responsibility towards affected communities and cultivation of the common good.

Awaiting A Moon Launch

The world is in even much more of a train wreck than when I first reported on this at the end of January, but there might be a launch to the moon tomorrow! At the moment we have 16 hours before launch of the Artemis 2 space rocket that will send four astronauts on a loop around the moon. It’s quite possible the launch will be scrubbed tomorrow due to a mechanical difficulty or weather (there’s a 20% chance of the latter), but the launch window for this month extends a few more day so they’d be able to try again the following day.

I will be following this actively tomorrow afternoon. Trump is scheduled to make some pronouncement about the Iran war in the evening — nobody knows what direction he’ll go on this — and is usurping all broadcasting for that; I will not be watching that. Ugh.

Wikipedia At Home

Both Pica and I had the Encyclopedia Britannica at home when we grew up, always at hand to dig into some topic of interest. Nowadays Wikipedia plays that role, the jewel of the Internet, a master reference only one bit of typing away. But what if…something happen to the Internet? A tech enclosure movement or collapse of the root level domain name system or rampant cyberwarfare?

It turns out one can download all of Wikipedia pretty easily. Since 2007 there has been a project called Kiwix that has created a system for taking extremely large knowledge stores and compressing them into single files that are easily viewed with special software offline.

This software is not difficult to run and is available for all major platforms. Their catalog lists 3458 different works that have been compiled into this file format in many different languages. I’ve had a go and have downloaded the complete English Wikipedia, Wikispecies, and resources from iFixit. The Wikipedia file is 115 GB in size and took several hours to download.

It’s a comfort having Wikipedia on my own hard drive. You never know.

The Ghost Airport of Aragón

There was an good piece yesterday about the epistemic collapse of the Iran war: nobody has any idea of what the facts are on the ground, and the information bubbles we all (separately) occupy are not communicating across to each other. Some interesting stories are emerging though, sometimes on the periphery of it all.

One of these stories concerns the airport of Teruel – Caudé in Aragón. This is not an airport for passengers. Rather, it’s a facility for aircraft maintenance and storage. When the war broke out, it began to receive big jetliners from the fleets of the Gulf States airlines to take refuge from the war. Planes from Qatar Airways, whose home base is Doha, are especially represented as part of the new arrivals, with about 20 of their widebody jets having come in by last Saturday.

For Fun Or For Beauty?

Thanks to a link on MetaFilter, recently I watched a masterful long-form video by jazz bassist and YouTuber Adam Neely entitled “Suno, AI Music, and the Bad Future”. Suno is the technology company that has become the biggest powerhouse in AI music. A usual workflow with Suno is to give it some text prompts describing the characteristics of the music you want and perhaps some lyrics and then Suno will take that and generate a fully instrumented song for you.

Needless to say, Neely is not a fan.

Commercial generative AI is bad in ways which are different from other disruptive music technologies of the past, like MIDI sequencing and samplers, because there is a sociopolitical agenda behind its adoption. This agenda will be bad for musicians, it will be bad for music lovers, and it will leave us feeling more alienated and alone…Like any culture war, it’s a distraction from the real war, the class war. You see, there is a class of techno-capitalists who are currently using generative AI as a means of wealth extraction. They do this by circumventing intellectual property laws…

Fifty-nine minutes into the video, Neely links AI music to technofuturist movements past and present. He discusses the Italian futurists, in particular Filippo Marinetti who in 1909 pens The Futurist Manifesto. Nine years later he goes on to write the original Fascist Manifesto. Today’s technocapitalists such as Marc Andreessen view Marinetti as a patron saint. Theirs is the same worldview that produces abominations such as Reflect Orbital’s plan.

Suno describes their core values as being “Music, Impatience, Aesthetics, and Fun”. Neely would replace these values with Service, Patience, Craft, and Beauty.

Neely believes generative AI will cause music to split into two different art forms, just as theatre and cinema are different but related art forms. These will be live and recorded music, and he has a lot of hope for the future of live music.