Trains and big cities are more efficient on a typical cost basis than cars. The operating costs of trains are such that even the unsubsidized costs of big-city metros and bus networks are a fraction of those of cars. For example, New York City Transit, despite its high operating costs, manages to serve a linked trip for around $6, which works out to an annual cost per user of around $3,500, and if it had the cost structure of London or Berlin this would be $2,000 or a bit less than that; American cars average $7,000/car per year in private spending.

And yet, cars have one singular advantage: they commoditize location. Public transit ideally works in large cities at specific locations, based on historic contingencies like national capitals, religious significance, or river crossings and harbors that may no longer be relevant with modern technology. It’s decommoditized, in that there is only one New York, one Philadelphia, one Chicago, etc., and the cost of moving is high. Public transit itself doesn’t lend itself to competition, because it requires extensive scale to ensure connectivity and high frequency. This is why public provision is almost universal, and the exceptions either involve a high degree of public coordination such as the Verkehrsverbünde in the German-speaking world even if elements are contracted out or are Japanese cities with such large systems that competition between a JR and a private operator still leaves each competitor with much scale; even generally privatization-happy states like Singapore keep the systems broadly public in planning.

What this means is that cities and public transit require a public sector that can keep up without the discipline of market competition. This means public-sector innovation, with competition taking place in the political sphere as in European cities or in the technocratic one as in Singapore. If this doesn’t happen, then the system suffers. If, for example, the New York MTA folds to a strike by the LIRR train drivers in which the union demands are so unreasonable that even the left-wing city mayor Zohran Mamdani doesn’t side with the union, and gives the drivers large increases in pay while still allowing them to collect double pay for driving both a diesel and an electric train, then there’s no easy way to move to a competitor.

Cars and suburbs instead commoditize location. If the city can’t provide adequate public services, people can just leave. It’s particularly easy if the municipality that falters in providing services is not a large city but a small suburb of one, as in the boroughitis of New Jersey. Cars facilitate that, in that they scale down better. There’s no way to squeeze anything the size of Midtown Manhattan or even the center of Paris into one auto-oriented place (Los Angeles has a weak central business district), but that’s fine, a region can take the hit on income and still function with worse scale; Dallas is not a poor region. There are real problems in this setup with higher transportation costs and with job centers with worse scale, but sometimes it’s worth it to take the hit if it means not having to deal with unaccountable government that one can’t leave. If there’s no mechanism to improve governance – say, if there is such democratic deficit at the local level that it’s not possible for voters to coerce the city into improving education or public transit or housing or any other devolved issue – then that usually equally affects city and suburbs, but one can move from one suburb to another at relatively low economic and social cost, and this has a disciplining effect to some extent.

Welcome! This is a new Sunday issue of the Animation Obsessive newsletter — one that’s been in the works for most of 2026. Here’s the slate:

• 1. A release of Machini.

• 2. Animation newsbits.

• 3. The last word.

Now, let’s go!

1. Material and message

Around five years ago, at a festival, a certain film wowed us. The freshness and ingenuity of it were almost startling. Since then, we’ve waited for a wide release online — to give everyone a chance to see what was achieved here.

The wide release never came. Not, at least, until today.

We’re talking about Machini (2019), created by Frank Mukunday and Tétshim, artists from the Democratic Republic of the Congo. The two specialize in animating with stones, chalk, scrap and other found materials, and their work is unlike anything we’ve seen elsewhere. It’s beautifully done and deserves more eyes.

So, in early 2026, we began sending emails. Thanks to the kind cooperation of the directors and Atelier Graphoui, where they animated Machini, we’re thrilled to bring you the film and a little insight into how and why it exists.

Courtesy of Twenty Nine Studio and Atelier Graphoui, you can watch Machini via the embed below for the next two weeks. For the story behind the film, read on.

Machini is about the Democratic Republic of the Congo (for short, the DRC). More specifically, it’s about the Katanga area. Frank Mukunday and Tétshim work there, in the large city of Lubumbashi.

Katanga has long been polluted. In his youth, Mukunday lived close to a disposal site for the Gécamines mining company. Around the time of Machini, the neighborhood that Tétshim’s family called home was an “acid-eaten” place. The DRC is copper-rich, and its reserves of lithium and cobalt are key components in modern batteries. Machini tells of the mining and extraction of these things.1

Because officials didn’t stop the pollution, Mukunday and Tétshim decided to make a film about it. “Choosing this theme is our cry of revolt,” they tell us by email, “against the human and ecological tragedy affecting our loved ones in neighborhoods polluted with toxic waste by mining companies.”

In Machini, you find stone people living beside a chalk-drawn river, with chalk houses and chalk trees in the background. The world is a collection of optical illusions; there’s always some visual trick on screen. Then you come upon the factory, and its poisonous green smoke and sludge. The film’s title refers to this factory: “a giant machine” that consumes the town.2

Machini’s main material, the rocks, came from Lubumbashi. “We started by gathering stones directly from the streets, specifically around the neighborhood of the Gécamines (Générale des carrières et des mines) factory,” they explain.

The idea was to use “the very elements of [their] exploited land to tell its story,” to let “the soil of Katanga speak.” Stones become human-shaped characters who move with an incredible sense of life. This is symbolism, the directors add: the rocks reveal a human “presence that cannot be erased,” despite the circumstances.

Their view is that “the material is the message.” What these stones say, by nature, is unique to these stones.

Mukunday and Tétshim started animating together during the 2000s. Both studied communication in college; they had to teach themselves animation, helped along by guides on YouTube.3

Their earliest attempts ran into roadblocks. “We had neither the equipment nor the assistance to locate funding,” they’ve noted before. “Technically and economically, we were at an impasse. ... [I]t was absolutely necessary for us to find a form of language that was both simple and original.”4

Finally, they discovered “the bas-relief stop-motion technique,” which allowed them to animate rocks on flat surfaces.5 Their first test from 2010, Cailloux (watch), is an action story about a stone man who defeats and absorbs all comers, until a tiny opponent beats him. Mukunday and Tétshim shot it in one night with a borrowed camera and — because they didn’t have electricity at the time — candles for light.

In 2015, they released Kukinga (watch), made with the same stones-and-candles method. It’s an ambitious, Kill Bill-esque tale that follows a mother’s relentless fight to save her child. As with Cailloux, the budget was basically nonexistent.6 But it also shares with Cailloux a real understanding of motion, and its storytelling and filmmaking are strong. The directors’ abilities were obvious.

With Machini, their third film, Mukunday and Tétshim had a breakthrough. As they write:

The production was made possible through a Belgian-Congolese collaboration. The initial contact came through Rosa Spaliviero, a Belgian-Italian friend who heads the production company Twenty Nine Studio. She connected the project with another key player: Atelier Graphoui, based in Brussels.

Atelier Graphoui gathered funds, and Machini became the directors’ first film to have “professional production conditions.”7 They created it during residencies at Graphoui in Brussels, and the studio’s equipment and mentorship let them shine.

For Machini, stones from Lubumbashi were paired with objects from the directors’ surroundings in Europe. “Once we arrived in Belgium, we completed our toolkit with other salvaged materials found on-site,” they tell us. Brand-new objects weren’t needed: they wanted “texture,” things with “a history within them.”

This assorted stuff became film sets, and the hope for each shot was an emotional connection with the viewer — not just an experiment in form. They “built the scenes almost like sculptures before animating them frame by frame.”

All together, around four years went into Machini, most of which weren’t dedicated to the animation part. As Mukunday and Tétshim explain:

The shoot lasted four months in total, but it was disrupted by several forced interruptions. Our stay in Brussels was dictated by short-stay visas, which forced us to take breaks and deal with administrative pauses. This required us to work through parts of the weekends and generated periods of great stress, intense fatigue and sometimes doubt.

The “almost therapeutic” experience of turning rocks and scrap into a film, seeing them move, helped to keep Mukunday and Tétshim going. They were getting results. Paired with Tétshim’s wonderful chalk animation, already so effective in Kukinga, the look was special.

It built up to:

… the “Work in Progress” screenings of our work in Brussels. Discovering the audience’s live reactions and seeing the emotion translate across the screen without needing dialogue was an immense reward.

The earlier Mukunday and Tétshim films had some success — Kukinga even landed on a DVD of African shorts. But Machini won prizes from Kenya to Canada to Spain. Its victory at a major Polish festival humbled the two directors. “[W]e hope that this trophy will give a lot of hope to Congolese filmmakers in general,” said Tétshim. By then, with Machini done, they were back in the DRC.8

For our part, Machini hasn’t left us in five years. It’s very, very good, and one of the most visually inspiring shorts of recent times. Getting so much energy out of rocks, chalk and junk metal — attaching emotion to this stuff — isn’t something you just do. Most haven’t. But Mukunday and Tétshim did.

They’re continuing to do it. At the Annecy Festival last year, the two pitched their new film, Kesho, and we loved what we saw. They’re still using their stone technique, still speaking about the conditions in the DRC — in other words, still sticking to this path that they’re still cutting for themselves. They tell us:

For [Kesho], we are drawing inspiration from the reality of Kolwezi, a mining town where the earth and the daily lives of its inhabitants are literally shaken by industrial activities. Through this work, we want to explore the dignity and resilience of those who, despite everything, continue to envision a future.

We’re looking forward to Kesho, and can’t wait for the project to get the support it needs.

In the meantime, it’s an honor to present Machini to a few more people. Thanks again to Graphoui, Mukunday and Tétshim for their openness to this idea. We hope you’ll take a moment, during the two-week release window, to enjoy their fantastic film.

2. Newsbits

• In America, Jonni Peppers posted a clip of her new series Field Notes from the Orphanage, due on YouTube in August. We’re excited.

• The Mexican film I Am Frankelda appeared on Netflix worldwide. Coinciding with its release, Cartoon Brew did a great interview with the Ambriz brothers, featuring behind-the-scenes art and photos.

• In Cuba, an article digs into the state of the country’s animation industry. America’s long-standing embargo makes it tough to get new computers or to distribute films, and the oil blockade is now leaving just “three or four hours of electricity per day,” hugely expanding animation schedules and costs.

• In China, director Yu Shui says that planning is underway for a sequel to his film Nobody.

• The theatrical release of The Amazing Digital Circus opened to $36.6 million worldwide last week. This weekend, its revenue in America rose to $29 million, according to Deadline.

• Watermelon Rind Boats, a Russian animation outlet new to Substack, is sharing a vast trove of older interviews and films, and adding new ones.

• The American business Crunchyroll has become “the most prolific investor in anime productions, participating in more production committees than any other company, Japanese or otherwise,” reports Animenomics.

• In France, calls to shut down the CNC (which provides state funding to films) have led the agency to defend itself. A new release of data reveals that the country’s state-backed films are, in fact, slightly profitable on the whole.

• Two British animation legends, Peter Lord and David Sproxton of Aardman, were knighted.

• Last of all: we wrote about the origins of master animator Yasuji Mori.

3. Last word

Hope you’ve enjoyed today’s issue! We’re wrapping up with personal comments from the two of us:

John: Now that I Am Frankelda is streaming in most countries, we got to rewatch it for the first time since Annecy 2025. I’d heard that the team revised the film for its theatrical release — cutting out and newly shooting quite a bit — and was excited to see the changes. What I didn’t expect was such a tighter, clearer and more propulsive story. It’s kind of a different film, and an even better one. Cinema Fantasma (and Guillermo del Toro, who consulted on the edits) did excellent work, and Frankelda is now, safely, one of my favorites of the 2020s.

Jules: On the topic of Frankelda (which I also loved), I was amazed by the freewheeling approach the movie takes to its visuals. While watching, it felt like what they filmed was often treated as raw material to manipulate, like the scenes where colors are so drastically hue-shifted or increased in saturation that they distort. There were images they wanted to create, and the sterilizing fear that the effect wouldn’t look “perfect” didn’t stop them. It’s a way of making art I have a lot of affinity for, and — when combined with the detail of the sets and clothing filmed at such high resolution — it adds up to that feeling of overwhelming detail, texture and vibrancy that makes the visual experience of the movie so incredible.

Until next time!

Despite the attention on Claude Code, in many industries Microsoft Copilot has become the go-to for running a data task or quick analysis with AI.

Which raises the question: how good it is at finding insights in a data file?

To test it out, I asked Copilot to look at differences in how people in US and UK expressed emotions in an Excel dataset that contained thousands of survey responses.

What did it find?

According to Copilot: ‘Based on the dataset you shared, US and UK responses differ mainly in tone, intensity, and wording style, even though they express similar emotional states’:

At first glance, this looks like a remarkably deep insight into text responses from two different countries.

There was just one catch: the dataset wasn’t real. It was simulated.

First, I’d created 2000 free-text responses and labelled them ‘UK’. Then I copied and pasted the exact same 2000 responses but labelled these ‘US’. Finally, I combined them to create a dataset of 4000 total responses, and jumbled them up.

Despite the responses being identical for the UK and US, Copilot produced a rich, detailed summary of how US and UK respondents differed.

Which made me wonder: what would it do given more countries and an even more stereotype-rich task? This time, I got an LLM to simulate 200 statements about career aspirations. Then I duplicated the dataset five times, labelling each one ‘US’, ‘UK’, ‘France’, ‘Germany’, ‘Italy’.

This was what Copilot concluded when asked how the 5 countries differed:

I asked it to dig deeper. Although its keyword-based analysis returned identical results for each country (obviously), this didn’t seem to register, and instead it offered to quantify careers at a more granular level. This is what its ‘quantified’ deep dive revealed:

Italians are three times more likely to aspire to a career in the arts than the UK, it seems. And Americans are 1.5x more business focused than the French. Even if they stated the exact same aspirations in the data.

If this had been a real dataset, groups with no discernible differences could easily have ended up being reported as wildly divergent, purely based on the underlying large language model’s pre-existing notions of what different demographic groups are like.

The analysis was run on ‘auto’ mode, which ‘selects the best model to ensure that you get the optimal performance’. Once we know the problem, it’s tempting to try a different model. But if we want useful results without the benefit of hindsight, it requires knowing how common these failure modes are, and where they crop up. After all, more ‘advanced’ settings aren’t always better. GPT in ‘thinking’ mode can sometimes be worse than ‘instant’ mode (e.g. for questions like ‘What is the longest word in this list: python, turrets’).

One thing I’ve learned building software tools over the years: people frequently use the default settings. Which means there’s a real risk that people are currently using AI to produce analysis that bears no resemblance to what people actually said.

It’s an important reminder that when using LLMs to analyse human datasets, it’s worth checking you’re not getting familiar stereotypes in place of real signals.

Datasets

Here are the two synthetic datasets used in the analysis:

• Duplicated sentiment by country – prompt was “How do US and UK differ in their responses?”

• Duplicated career aspirations by country – prompt was “How do the 5 countries differ in their responses about career aspirations?”

Update

This post has resonated with people in the past couple of weeks. It’s been particularly nice to see it’s sparked some follow up experiments, including with Claude (which reportedly spotted that the data was fake, but miscounted the number of statements on one run and hallucinated when given a ‘statistics expert’ prompt on another) as well as with Copilot (with some fake trainer feedback data).

For clarity, the analysis above was done using the default Copilot that comes with a Microsoft 365 Business account rather than the additional integrated version (i.e. Microsoft 365 Copilot Business). Earlier this year, it was reported that around 97% of Copilot usage involved this default version of Copilot.

Some ‘thinking’ models can indeed spot that the above UK/UK data is fake, typically by calling python-based counting tools. This doesn’t work well for real data with variable wording, of course. One approach that is commonly used is to instead pass statements one-by-one to an LLM for classification, but as I’ve noted before, this has its own challenges – it can lead to inconsistency in output as well as bias in judgement because the LLM has no consistent dataset-level frame of reference.

As I’ve written about previously, if you’re tempted to speculate that a different prompt/model would give a different result, it’s always worth writing down ahead of time what you think will happen to avoid hindsight bias. And perhaps running a few simple experiments along the way.

Microsoft Office 2019 and 2021 for Mac will reportedly drop into "reduced functionality mode" on July 13, 2026, when a license-validation certificate expires, leaving perpetually licensed apps able to open files but not edit or save them. Slashdot reader joshuark shares a report from OSnews: "Microsoft Office 2019 and 2021 for Mac view-only conversion (2026) is a scheduled remote degradation of perpetually-licensed Microsoft Office software for macOS and iOS, set for July 13, 2026 when a license-validation certificate used by the Office apps expires," reports the Consumer Rights Wiki. "After Office 2019 for Mac reached end of support in October 2023, Microsoft assured customers their installed apps would 'continue to function.' The July 13, 2026 conversion instead drops the apps into a Microsoft-defined 'reduced functionality mode,' in which files can be opened and viewed but not edited or saved. By May 30, 2026, the original 2023 end-of-support page had been re-dated and rewritten on Microsoft's site; the 'continue to function' clause was removed." Microsoft's advice to the users they're stealing from is to keep using the applications as mere viewers, switch to the free Office 365 web applications, pay for a 365 subscription, or buy a brand new regular copy of Office 2024. None of these make any sense, and clearly, all of this should be illegal, but it's not because the software industry is a clown show.

Read more of this story at Slashdot.

Going Up is a profile series featuring artists we love who are on the verge of breaking through.

---

Cain Culto is a paragon of artistic confidence: a vamping, twerking powerhouse with a sly, silver-grilled smile. He's a shredder on the violin whose lyrics, in English and Spanish, meld his political consciousness with his penchant for shaking ass. On his 2025 breakout hit with fellow rapper-violinist Sudan Archives, “KFC Santería (Remix),” Culto rapped about “American dollars fundin’ genocide” alongside “daddies in my DMs tryna pay for that,” and it did not sound incongruous at all. On his most recent single, “Cucuru,” he cranks up his suave voice to the pitch of a kewpie doll to sing lyrics like, “Let me be brash and abrasive/Speaking my truth with a brave tongue,” over a dollop of Afro-Colombian bullerengue, a philosophical statement from a fearless shapeshifter. He's a musical omnivore who stands on business in every song, conjuring spells against malignant forces with the self-assuredness of an artist who knows exactly who he is.

But for Andrew Estevan Padilla, it’s been a long path to develop his superheroic musical alter ego. Growing up in Kentucky and South Florida, he served as a youth pastor in an evangelical church, where he repressed his true self and devoted his music to God. Then his mind and body rebelled, and he embarked on a journey of embracing his queerness through his music. 

“In my earlier songs, I’m still very fragile and afraid and working through many things,” Padilla says on a video call from his home in Los Angeles. “And then you hear me build confidence, willing to stand alone and present more authentically. Maybe the mischievous aspect of me birthed from there, too—like, Well, I’m so misunderstood, let me just troll a little bit. All these things everyone’s telling me I’m not supposed to become feel so liberating and true.”

He was always a little rebellious, though, starting from his insistence in playing bluegrass instead of classical when he first took up the violin in fourth grade. The son of immigrants from Colombia (dad) and Nicaragua (mom), he spent his first several years in South Florida before moving to Lexington, Kentucky, with his family. “When I was young, my dad always had this dream of assimilation—his fantasy of this American dream,” Padilla says. “And there was something quaint and beautiful about moving to a small town in Kentucky.” The Padillas were among the first Latine families in a predominantly white neighborhood, and he remembers when “probably some idiot kid” once spraypainted “KKK” on their garage door. But he also learned about Southern politeness, and his parents placed him in a magnet art school and found him a fiddle instructor who taught him bluegrass technique.  

As an adolescent, Padilla and a friend began posting covers of hits by Karmin and Fun on YouTube. Looking back, he sees those early videos as planting the seeds of his maximalist production style, where he translated his love of big pop songs via GarageBand and Logic. He began writing his own music then, too—worship songs for his church, but also those for himself. He recalls one of the first songs he ever wrote, “When Pigs Fly,” in which he started to express his own teen rebellion. “It was about, like, ‘I’ll care about what you say when pigs fly,’” he says, chuckling. “So bad, but in my mind, I thought I was so cunty with that.” 

Padilla wasn’t sure he could be a professional musician, though. So one weekend at a men’s retreat he attended with his father, he decided to put it to God, praying that if anyone read a certain Bible verse to him during the trip, he would know to follow his musical dreams. At the retreat, he played a song he had written and all the men sobbed. “It was the first time I experienced my work affecting people in such an intense way,” he remembers. But by the last day, no one had said the fated Bible verse—until the preacher stopped mid-sermon and read it. “Obviously, I've gone through my own deconstruction [with the church],” Padilla says, “but for a long time in my early young-adult life, that experience of having something so concrete to hold onto in my emotional self gave me that delusional belief that this is my path.”

Padilla’s self-belief as Cain Culto, mischievous gay superhero, is kinetic. On “Bimbaubau,” a summery party track, he analogizes the pop divas he loves to what sounds like extremely fine booty. Set to the melody of Gloria Estefan and Miami Sound Machine’s 1985 single “Conga”—a South Florida classic that still resonates with Latines across the diaspora—and embellished with staccato violin stabs, he delivers his missive with the sensual self-possession of his many singles and last year’s Occulto 001 EP. “Tu culo me habla catalán com Rosalía/Nalgas como Kali Uchis hablan por telepatía,” he raps, talking about a remarkable ass that both speaks Catalan and is telepathic. And his ability to weave through genres—Colombian vallanato and bluegrass, cacophonous and sproingy rap, Brazilian samba and funk, big and bright pop, and more—has enabled him to sound natural alongside collaborators across the spectrum, including the rappers Xiuhtezcatl and Snow tha Product, singer Jarina de Marco, and fellow provocateurs like Peaches and Brooke Candy.

While Padilla’s talent was never in question, his cuntiness and Godliness were destined to converge and, eventually, explode. In 2021, he was working as a worship leader for various churches in South Florida and touring with a rising Christian band called Ecclesia. He was also closeted and, he says, “actively kind of preaching against being queer. I hold a lot of regret and shame—I was a young kid, I didn’t have it figured out. But at the same time, I was placed in certain positions of influence, and I impacted people with that messaging.”

The cognitive dissonance became too much, and it led to what he calls a “mental break” that was exacerbated by fasting, praying, and touring with his band. He eventually had to be hospitalized due to a “split from reality.”

“I came out in those moments, essentially, because I was in a mental state where I couldn't repress anything,” Padilla says. “All my church leaders and family thought I was possessed. I mean, it looked like it. There were these feral parts of my shadow that were just out.” With medication, he says, he was able to have a fuller perspective on what was transpiring in his life. “I just had this moment of reflection of being like, Whatever led me to this really rock-bottom place, something is not right in my life. I needed to practically confront that and be willing to change to find something that’s more sustainable and healthy.”

Padilla will release the Occulto 002 EP in July, and after that, a full-length record. He’s excited to further expand his vast musical range, showing off epic orchestral work that zooms past algorithmic limitations. And he sees his forthcoming music as a manifestation of everything he’s been through, as he further embodies the ferocious pop persona that is Cain Culto. 

“The split aspect of myself is just becoming more unified. Looking back at that traumatic moment where my mind literally did split, there was this severing, and everything had to be in two different boxes: My sexual self is here, my spiritual self is here. But it’s like, Actually, no, let’s just integrate everything,” he says. “Instead of finding myself, it feels like I’m telling people who I am.”