Fumfer Physics 48: Kyiv, COVID, Computation, and Cosmology

Author(s): Scott Douglas Jacobsen

Publication (Outlet/Website): Vocal.Media

Publication Date (yyyy/mm/dd): 2026/05/25

Photo by Alexandre Debiève on Unsplash

In this wide-ranging interview, Rick Rosner and Scott Douglas Jacobsen examine Russia’s May 23–24 attack on Kyiv, including missile and drone tactics, civil preparedness, and depleted-uranium concerns. They then discuss COVID wastewater surveillance, gastrointestinal shedding, and contagiousness estimates before pivoting to paint-by-number technology, computational optimization, the travelling salesman problem, and cosmological questions about the Big Bang, time, and speculative prior-universe models with scientific caution and humour across an unusually eclectic exchange.

Kyiv Attack: Facts, Attribution, and Scale

Rick Rosner: You were part of one of the biggest Russian attacks on Kyiv. What are the key facts of the attack on Kyiv on May 23rd and 24th? Is this desperation because Putin is losing the war, or what is the deal?

Scott Douglas Jacobsen: There is an attribution problem. We can speculate about what is in Putin’s head, but it is better to begin with what Russia actually did and what officials have confirmed. Moscow framed the attack as retaliation for recent Ukrainian strikes, while Ukrainian and European officials described it as an escalation against civilian and symbolic targets. That does not prove a single motive, but it does show a deliberate mass-strike strategy.

Overnight from May 23 to May 24, Russia launched a large combined missile and drone attack on Ukraine, with Kyiv and the Kyiv region as the main targets. Ukraine’s Air Force reported that Russia launched 90 missiles and 600 drones of various types. The weapons reportedly included one Oreshnik intermediate-range ballistic missile, two Kh-47M2 Kinzhal air-launched ballistic missiles, three 3M22 Zircon missiles, Iskander-M and S-400 ballistic missiles, cruise missiles, and drones.

Ukrainian air defences shot down or jammed 55 missiles and 549 drones. Sixteen missiles and 51 drones still hit 54 locations, while debris from intercepted weapons fell at 23 locations. Kyiv was the primary target, and damage was reported across the city’s districts.

At least four people were killed across Ukraine, including two in Kyiv and two in the Kyiv region. More than 80 people were injured in Kyiv, and Ukrainian reports put the broader injury count at around 100. The attack damaged residential buildings, schools, a market, a water supply facility, government buildings, and cultural institutions. The National Art Museum, Kyiv Opera Theatre, Ukrainian House, Valeriy Lobanovskyi Dynamo Stadium, and the Chornobyl Museum were among the damaged sites.

The Chornobyl Museum should not be described simply as “destroyed” unless later reporting confirms that. The more accurate wording is that it was damaged, with reports of serious losses. Al Jazeera reported that Zelenskyy said more than 40 percent of the museum was “irretrievably lost.”

The Oreshnik missile reportedly targeted Bila Tserkva, south of Kyiv, rather than Kyiv city itself. Its use was significant because Russia confirmed it had used the missile, and Ukrainian and Western officials described it as a nuclear-capable weapon and a form of escalation.

For me personally, this was the most severe attack I have experienced during my trips to Ukraine. I slept in the bathroom, and other people I know did too. That says something about how serious the night felt on the ground, but the broader factual record is this: Kyiv faced a major combined Russian missile-and-drone assault, one of the largest of the full-scale war by scale and damage footprint, with civilian infrastructure, residential areas, and cultural sites hit.

Sheltering, Civil Preparedness, and the Fifth Year of War

Jacobsen: Some people went to shelters, and others just toughed it out. I checked in with each of them, asking, “Are you safe? Are you okay?” Thankfully, as far as I know, nobody I knew was injured or killed last night. That is still bad.

Rosner: I am glad you made it through this.

Jacobsen: This is now the fifth year of the full-scale war. The full-scale invasion began on February 24, 2022; the broader Russo-Ukrainian war began in 2014. People have learned how to take shelter. Russia can launch an enormous attack on a city, and because of air defences, shelters, warnings, and civil preparedness, the casualties can still be lower than the scale of the attack might suggest. But it is still terrible.

Rosner: You mentioned a hypersonic missile. I do not know exactly how fast this one goes.

Jacobsen: The Oreshnik is generally reported at around Mach 10, or roughly 12,000 kilometres per hour. It is a ballistic missile with hypersonic re-entry speed, so its terminal phase is extraordinarily fast. The speed alone generates enormous kinetic energy, in addition to the explosive payload.

Rosner: It is also difficult to intercept because of its speed and flight profile. It is a terror weapon, too. It is not something Russia is going to fire casually in large numbers. It is probably extremely expensive, likely costing millions of dollars per missile, so part of the purpose is psychological: use enough of them that they get reported on and scare people. The main body of these attacks still consists of missiles and large numbers of drones.

Radioactive Debris, Depleted Uranium, and Plutonium Risks

Jacobsen: Yes. There are also reports about elevated radiation in some Russian weapon debris, but the accurate point is depleted uranium, not depleted plutonium. Ukraine’s Security Service reported elevated radiation on fragments of an R-60 air-to-air missile mounted on a modified Geran-2 drone in Chernihiv Oblast after an April 7 attack. Tests reportedly identified depleted uranium components, including uranium-235 and uranium-238, in the warhead.

The Ukrainian Air Force later cautioned against treating this as a new category of “radioactive drone.” They explained that the R-60 is an old Soviet missile with a depleted-uranium alloy warhead. So, the issue is serious, but the wording matters. It is not depleted plutonium, and it is not, on current evidence, a new fleet of radioactive drones.

Rosner: When people talk about nuclear weapons, they usually mean weapons that produce a nuclear explosion using fissile material, such as highly enriched uranium or plutonium. Depleted uranium is different. It is what remains after much of the fissile uranium-235 has been removed. It is mildly radioactive, but its main danger is chemical toxicity, especially if particles are inhaled or ingested after impact or burning. So, it is dangerous, but it is not a nuclear bomb. It is more like a toxic and radiological contamination hazard. That is bad enough without giving it comic-book supervillain physics.

Jacobsen: This is the first reporting I have seen on that specific finding so that I would hold off on any conclusive claims. What we can say is that elevated radiation levels were reported in the wreckage, and if radiation levels are elevated, then something radioactive is present relative to the normal background level. The cause, scale, and broader pattern still need careful verification.

If that is abnormal compared to regular Shahed electronics and components, then something is suspicious. If investigators conduct a chemical analysis and find uranium, plutonium, or another radioactive isotope, then they have evidence.

Rosner: Even the United States has used depleted uranium in some munitions. Depleted uranium is a by-product of uranium enrichment. Enrichment increases the proportion of uranium-235, the more fissile isotope used in reactor fuel and nuclear weapons. What remains is depleted uranium, which is mostly uranium-238, with much less uranium-235 than natural uranium.

Depleted uranium is used in some armour-piercing munitions because it is extremely dense and can punch through armour effectively. It is not a nuclear explosive. It is mildly radioactive and chemically toxic, so inhaled or ingested particles can be dangerous.

If they were using plutonium, that would be worse. Plutonium is not “purely explosive,” but plutonium-239 is used in nuclear weapons, and several plutonium isotopes emit alpha particles. The main health danger is inhalation: plutonium particles can lodge in lung tissue, damage cells, and increase the risk of lung disease and cancer.

A tiny particle can contain a huge number of atoms, and those atoms decay over time. If that material is inside your body, it becomes a little internal radiation source. So, yes, radioactive contamination in weapon debris is bad enough without exaggerating it into science fiction.

All right. I have some dumb topics, including “butt COVID.”

COVID Wastewater, “Butt COVID,” and Transmission Uncertainty

Jacobsen: What is butt COVID?

Rosner: That is my term for it. The United States is currently at relatively low COVID levels. Wastewater and clinical data in spring 2026 showed a lull, with national wastewater levels reported by the CDC as very low and some surveillance averages at their lowest levels since spring 2022.

The question is whether COVID ever gets low enough to lose its potential to cause large spikes. In past years, the United States often saw two major waves a year: a winter wave and a summer wave. The winter wave usually began after Halloween, rose through Thanksgiving and Christmas, and peaked around early January, when people were gathering indoors. Then it would decline through February and March, reaching lower levels in spring or early summer.

Then, as people travelled, gathered, and returned to school, another wave often appeared in late summer, especially around August and September. Johns Hopkins has noted that the United States has had a COVID wave every summer since 2020, though it remains unclear whether COVID will eventually settle into a more predictable seasonal pattern like influenza or RSV.

This winter was not wave-free, but it appears to have been smaller than several earlier COVID surges. In January 2026, wastewater levels were still moderate by CDC measures and high by WastewaterSCAN, but some analysts noted that if that was the winter peak, it was much lower than peaks in prior years.

We still have the potential for spikes. As of May 19, 2026, CDC modelling estimated that COVID infections were growing or likely to grow in 4 states, declining or likely to decline in 14 states, and not changing in 19 states. That means the current burden is relatively low, but the virus has not gone away.

I have not been masking at the gym, mostly because the incidence in California appears low. One estimate I saw put contagiousness at around one person in 250. I think that may be high, especially at a gym, where people are less likely to work out if they feel awful. I estimate the gym risk is closer to one in 500; this is a rough personal estimate, not a formal surveillance number.

When you look at the peaks and valleys of COVID during the first two years of the pandemic, wastewater levels sometimes dropped close to zero in some places.

Now, in the valleys, wastewater levels drop very low, but they do not always reach zero. That means some SARS-CoV-2 RNA is still entering the wastewater. It does not necessarily mean that everyone contributing to that signal is actively contagious. It may reflect ongoing low-level transmission, prolonged fecal shedding after infection, or other dynamics of wastewater sampling. CDC wastewater data for mid-May 2026 classified national COVID-19 wastewater viral activity as “very low,” not zero.

Some people can continue shedding SARS-CoV-2 RNA in stool after they have cleared the virus from the respiratory tract. That is the serious version of what I am jokingly calling “butt COVID.” A Stanford-led study found fecal viral RNA in about 12.7 percent of participants four months after diagnosis and about 3.8 percent seven months after diagnosis, even when ongoing oropharyngeal shedding was no longer detected. Fecal shedding was also associated with gastrointestinal symptoms.

So, even when transmission is low, wastewater can still detect viral material. It may be coming from people who recently had COVID, people with low-level ongoing infection, or people with prolonged gastrointestinal shedding. Every time they use the bathroom, they may be putting viral RNA fragments into the wastewater. That is different from saying they are breathing infectious virus onto other people.

I follow one analyst who tracks this closely and seems mathematically competent. He estimates that even today’s very low levels could still correspond to many infections per day in the United States. I would treat that as a model estimate rather than an official CDC count. But even if the exact number is uncertain, the broader point stands: “very low” does not mean “none.”

COVID is best known as a respiratory infection. It infects the nose, throat, airways, and lungs, which is why most testing focuses on the respiratory tract. But SARS-CoV-2 can also affect the gastrointestinal tract, and viral RNA can appear in stool. That does not automatically mean the stool contains infectious virus, but it does mean the gut can be involved.

When I was a kid, we had terrible novelty-store toys. Places like Spencer’s Gifts sold prank items such as itching powder. The key point is that the itching powder was not necessarily fibreglass. Many forms use irritant plant hairs, such as rose hips or Mucuna pruriens, which cause mechanical or chemical irritation. Some materials irritate the skin because tiny fibres or spicules are present.

That is a rough analogy for what lingering viral material can do. If SARS-CoV-2 or its fragments persist in the gut, they may irritate tissue or correlate with gastrointestinal symptoms. But someone with only gastrointestinal shedding may be much less likely to infect others through ordinary breathing, because they are not necessarily exhaling infectious virus. The virus signal is going from their gut into the sewage system, not from their lungs into the air.

That is bad news for people with prolonged gastrointestinal symptoms, but it may be better news for population-level transmission. It means some fraction of the SARS-CoV-2 detected in wastewater may come from people who are not highly contagious in the usual respiratory sense. Wastewater remains a useful surveillance tool, but translating wastewater levels directly into the number of contagious people is not straightforward.

I can look at his number: 180,000. He thinks COVID could disappear if daily infections in the United States ever drop to around 10,000. But, by his estimate, we are still at around 180,000 infections a day. I would treat that as a model estimate, not an official case count.

I also think you can subtract some portion from that 180,000 figure, perhaps many tens of thousands, because some of the viral material detected in wastewater may come from people who are shedding SARS-CoV-2 RNA but are not necessarily contagious in the usual respiratory sense. That is my speculation, not a settled measurement.

There are other statistics. CDC’s ensemble forecast for the week ending May 30, 2026, estimated that COVID-19 accounted for about 0.11 percent of new emergency department visits in the United States. The most recent reported week, ending May 16, was 0.12 percent. So, roughly speaking, only about 1 in 1,000 emergency department visits were diagnosed as COVID-related.

The cleaner statistic is not “1,500 people a week show up at emergency rooms with COVID.” CDC’s same forecast estimated about 1,200 new laboratory-confirmed COVID-19 hospital admissions for the week ending May 30, 2026. That is hospital admissions, not emergency room visits.

Either way, severe clinical indicators are very low compared with earlier pandemic periods. Wastewater levels are also very low, but not zero. CDC notes that wastewater monitoring can detect infections even before people seek medical care and can also detect infections without symptoms.

So, I think you can probably subtract an uncertain number from this analyst’s wastewater-based infection estimate, since a fraction of the SARS-CoV-2 RNA in wastewater may come from people who are not very contagious. I wouldn’t put a hard number on that. It is a reasonable hypothesis, not a clean accounting line. Comments?

Jacobsen: No.

Paint-by-Number Kits and Image Optimization

Rosner: I have one more topic. You are probably not old enough to remember the first paint-by-number craze. You have probably heard of it, though. Paint-by-number kits became a major popular hobby in the early 1950s, especially through Max Klein of the Palmer Paint Company and artist Dan Robbins. The Smithsonian identifies Klein and Robbins as central figures in that fad.

When I was a kid, you could buy paint-by-number kits. You would get a kit with a simple canvas or board. Printed on it was a pattern of shapes forming a picture. Each shape had a number printed inside it. There might be eight different numbers, and you would get eight pots of paint.

If it were a watercolour kit, you might get little discs or dishes of watercolour. You would touch the paint to the water and then fill in the corresponding numbered areas. The pictures you got in the late 1960s were not great. It might be a horse in a field next to a fence or something similarly simple. It was not always aesthetically satisfying.

As you might guess, there has been a modern resurgence of paint-by-number kits, and higher-resolution printing and better manufacturing can make some of them look much better. Carol has been doing these kits, and many of them come from China.

With the kit, you get a nice canvas, maybe 17 by 20 inches. Instead of eight colours, the number of colours depends on how much you pay. You can buy the same image at different price points. The more you pay, the more colours you get: maybe 24, 36, or 48 colours. You get the paint pots and an intricate printed image.

There are hundreds, probably thousands, of different regions to fill in. With that many different colours, by the time you are done, you get a decent-looking image.

The technology involved is impressive. You have to start with the source image. The one Carol did most recently was a fluffy dog surrounded by flowers. You can go online and see the original painting. Then they have to scan or digitize the painting and optimize it for the number of colours you bought.

They have to figure out a near-optimal colour distribution. If you bought the 36-colour version, they have to reduce the original image to 36 workable colours. Then they have to provide those 36 paints.

Maybe they have thousands of paint colours already available. Or maybe they have fewer base colours and mix the little paint pots from those in varying proportions. In any case, they either have to custom-mix the colours or keep a huge range of colours on hand.

The Travelling Salesman Problem and Practical Computation

There is a famous problem in computing. The travelling salesperson problem is this: a person must visit a set of cities, and you are trying to find the shortest possible route that visits each city exactly once and returns to the starting point.

It is famous because the number of possible routes grows extremely quickly. For a simple version with n cities, the number of possible orderings is roughly factorial. With 10 cities, you have thousands or millions of possible routes, depending on how you count starting points and reversed routes. With 100 cities, the number becomes unimaginably large.

So, brute-forcing the absolute shortest route becomes computationally intractable very quickly. People have developed clever algorithms and approximations, but the general problem remains hard because the search space explodes.

Big Bang Cosmology and the Problem of “Before”

Jacobsen: My friend has a question for you. What was before the Big Bang? They just looked at me and smiled. They are very happy, very giddy.

Rosner: Most cosmologists would say that, in the standard Big Bang model, asking what came “before” the Big Bang may not be meaningful in the ordinary sense, because time itself is part of the universe. Time is not necessarily something sitting outside the universe with a stopwatch, waiting for the universe to begin.

So, in that standard picture, there may not have been a “before” in the usual sense. The Big Bang was not an explosion inside pre-existing space. It was the early hot, dense state from which space, time, matter, and energy evolved.

That said, we do not know with certainty what happened at the earliest moment. Our current physics breaks down near the Planck era, where quantum gravity would matter. Some speculative models propose a bounce, a cyclic universe, eternal inflation, or a larger multiverse framework. But those are not established in the same way as the evidence for the hot early universe.

So the careful answer is: in standard cosmology, “before the Big Bang” may be an ill-posed question, because time as we understand it begins with the universe. In speculative cosmology, there may be something prior, but we do not yet have decisive evidence.

You can argue that there may have been some prior physical state, or some previous universe, in which something happened that led to the creation of a new universe. Some speculative models, such as black-hole cosmology, explore ideas in that general territory. But that is speculative, not established cosmology.

You can argue both ways. You can argue for some previous physical state that set up the Big Bang, or you can argue that the Big Bang marks the boundary of meaningful time in our universe. In the standard picture, time, space, matter, and energy are all part of the universe, so it may not make sense to speak of time “before” the universe in the usual way. NASA’s public summary puts the caution plainly: scientists are not sure what came before inflation or what powered it.

Without space-time, there is no free-floating time sitting outside the universe. There is nothing ordinary language can clearly refer to as “before,” because the framework for before and after is itself part of the universe.

Paint-by-Number as Semi-Optimization

To get back to paint-by-numbers, there seems to be a lot of computation-heavy semi-optimization involved. I do not know the exact technical term. In the travelling salesperson problem, mathematicians seek the optimal path. Most practical users are often satisfied with a route that is close enough.

The travelling salesperson problem asks for the shortest possible route that visits each city exactly once and returns to the starting point. The brute-force method grows factorially and becomes impractical very quickly, although there are sophisticated exact algorithms, heuristics, and approximation methods.

In practice, finding the single shortest absolute path often matters less than finding a very good path. If you have to visit 80 cities, you might run an algorithm many times and gradually improve the route. After a thousand runs, maybe you find a path that is 1 percent shorter. After ten thousand runs, maybe you improve it by another fraction of a percent. At some point, the improvement may not matter much practically.

Paint-by-numbers seems full of these optimization problems. If you order the 24-colour version, the system has to process the image and divide it into many regions based on 24 colours. The 36-colour version will have different regions optimized for 36 colours.

I think the technology is impressive. For $40 or $45, you can send them your own picture. For Mother’s Day, I sent in a photo and bought the expensive 48-colour version. For some reason, they gave us 50 colours. Their optimization must have concluded that the image needed a couple of extra colours. We got it back in about a week.

They took a regular photo and optimized it so the final painting would match it, using only 50 colours. They divided the image into many regions and gave us 50 pots of paint, each customized to that region.

Carole has done a few sets of these. You do not get the same colours every time. It does not seem as if they are working from 100 fixed colours and selecting the 48 closest. They appear to be able to generate hundreds of colours, or at least a very large palette. The technology is impressive.

Maybe I am extra impressed because I remember how limited paint-by-number kits used to be when they were done manually. The original craze began in the early 1950s through Max S. Klein of the Palmer Paint Company and artist Dan Robbins, and the early kits were much simpler than many custom kits today.

Jacobsen: Thank you very much for the opportunity and your time, Rick.

Last updated May 3, 2025. These terms govern all In Sight Publishing content—past, present, and future—and supersede any prior notices. In Sight Publishing by Scott Douglas Jacobsen is licensed under a Creative Commons BY‑NC‑ND 4.0; © In Sight Publishing by Scott Douglas Jacobsen 2012–Present. All trademarks, performances, databases & branding are owned by their rights holders; no use without permission. Unauthorized copying, modification, framing or public communication is prohibited. External links are not endorsed. Cookies & tracking require consent, and data processing complies with PIPEDA & GDPR; no data from children < 13 (COPPA). Content meets WCAG 2.1 AA under the Accessible Canada Act & is preserved in open archival formats with backups. Excerpts & links require full credit & hyperlink; limited quoting under fair-dealing & fair-use. All content is informational; no liability for errors or omissions: Feedback welcome, and verified errors corrected promptly. For permissions or DMCA notices, email: scott.jacobsen2025@gmail.com. Site use is governed by BC laws; content is “as‑is,” liability limited, users indemnify us; moral, performers’ & database sui generis rights reserved.