Rick Rosner and Scott Douglas Jacobsen on Kyiv Attacks, Contextual Compute, AI Risk, and Future Toilets

Author(s): Scott Douglas Jacobsen

Publication (Outlet/Website): Medium (Personal)

Publication Date (yyyy/mm/dd): 2026/06/02

How do Rick Rosner and Scott Douglas Jacobsen connect wartime life in Kyiv, contextual compute, AGI hype, labor-market disruption, and the future of personalized sanitation technology?

In this interview, Scott Douglas Jacobsen and Rick Rosner discuss overnight explosions in Kyiv, ballistic-missile threats, Ukrainian air defense, Crimea logistics, and civilian sheltering. They then pivot, with characteristic absurdist precision, to toilet-paper efficiency, sanitation futures, contextual compute, AI valuations, AGI hype, labor-market disruption, and personalized diagnostic toilets. Jacobsen frames computation as energy optimization across CPUs, GPUs, and quantum processors, while Rosner links integration, specialization, and technological risk, with public consequences for economies and health.

Rick Rosner: So, last night was very explody over Kyiv [Ed. June 1–2 overnight].

Scott Douglas Jacobsen: Yeah. The big thing I was paying attention to was that I went to bed at about 1:30 a.m., somewhere between 1:00 and 2:00, and I immediately started trying to determine whether the explosions were caused by Shahed drones or ballistic missiles. Based on the sounds, they seemed much more likely to be ballistic missiles.

Rosner: Which is worse, right? Because they are faster and more powerful?

Jacobsen: Yeah. Ballistic missiles are generally much faster and more difficult to intercept. They also produce a distinctive flash and explosion. I did not hear anti-aircraft fire at the time, which further suggested that some of what I was hearing could have been ballistic missiles. Based on the character of the explosions, that was my assessment.

I reacted to the first two strikes, went into the bathroom, and sheltered there. Then things quieted down for a while, and I went back to bed. After that, I did not even get up. I put on my noise-canceling headphones and eventually fell asleep.

By the next day, reports indicated that roughly a dozen people had been killed and dozens more injured.

The attack was somewhat similar in scale to the May 24 strike on Kyiv. It seems as though Russia increasingly has to accumulate missiles and drones before launching large-scale barrages intended to stress or overwhelm Ukrainian air-defense systems.

The drones themselves are not the primary concern.

Rosner: So they are running short of munitions, is what you are saying?

Jacobsen: That is one possibility. Ukrainian strikes against military facilities, logistics hubs, airfields, and supply infrastructure deep inside Russian-controlled territory may be affecting Russia’s ability to sustain the same tempo of attacks. At the same time, these operations are expensive and resource-intensive.

The recent attacks appear designed to saturate or overwhelm air defenses. Many of the strikes were directed at Kyiv itself, as was the case on May 24.

In my opinion, drones are less concerning than they were earlier in the war because Ukrainian air defenses, electronic warfare systems, and mobile interception teams now destroy a large majority of them. Exact interception rates vary from attack to attack, but it is common for most incoming drones to be neutralized before reaching their targets.

The more serious threat comes from ballistic missiles because they are harder to intercept and require sophisticated air-defense systems such as Patriot batteries and their associated interceptor missiles. Ukraine has improved dramatically in drone technology and drone operations, but ballistic-missile defense remains heavily dependent on advanced Western systems.

Those are the weapons that genuinely worry me. Drone attacks, especially isolated ones during the daytime, generally attract much less concern from residents.

For example, there was a drone overhead earlier today. I could hear it, so I went indoors for a few minutes and waited for it to pass. The general safety guidance is to position yourself behind multiple walls and away from windows whenever possible.

But yes, I am safe and alive. I was not among those killed or injured, and no one I personally know was killed or injured. I am fortunate in that respect.

Some people, however, know multiple victims personally.

Rosner: Are there contractors who go around and fortify apartments?

Jacobsen: That is an interesting question. I do not know. Many Soviet-era apartment buildings were constructed using reinforced concrete and other durable materials. However, even a sturdy building can sustain catastrophic damage from a direct hit or a nearby ballistic-missile strike. Depending on the circumstances, part of a building, or in extreme cases an entire section, can be destroyed.

Rosner: My wife used to work at a private high school that had fairly porous perimeter security. We talked about building her a little hiding place in case a shooter ever came to the school. The cheapest option seemed to be buying sheets of half-inch steel and lining the area under her desk. If a shooter arrived and his weapons were not especially powerful, the steel might stop the bullets, provided he did not walk around the desk and see her hiding there.

We never actually did it. She left the job before we took any action.

Still, it seems to me that there must be some affluent Ukrainians who are building panic rooms fortified against shrapnel.

Jacobsen: It is possible. I would not rule it out.

I suspect some wealthier individuals have reinforced safe rooms or private shelters. However, for most people, the standard recommendation is to use designated shelters, underground facilities, or locations that provide substantial protection from blast effects and shrapnel.

Rosner: Have you ever slept down in the subway?

Jacobsen: Pardon?

Rosner: Sometimes Ukrainians sleep in subway stations during air raids because they offer protection. Kyiv has a metro system. Have you ever done that?

Jacobsen: No. I have never done that myself, but many people do.

Some Kyiv Metro stations are exceptionally deep because they were designed to function as bomb shelters as well as transportation infrastructure. In some of the deepest stations, the escalator ride can take close to two minutes from the surface to the platform.

Rosner: Because they are so deep underground.

Jacobsen: Yeah. Unless someone starts using something like American bunker-buster bombs against civilian targets, which is not even on the menu right now, it is not very likely. So it is still quite safe.

Rosner: Yeah.

Jacobsen: It is a good subway system, too. The only issue is that getting down there takes forever.

Rosner: Yeah. It is like Britain’s Tube. The escalators are maybe a hundred yards long, perhaps even 150 yards. We read over here that Crimea is becoming increasingly cut off from Russian supply lines. What are you hearing over there? Is it true that there is really only one bridge the Russians can use to supply Crimea?

Jacobsen: That is generally correct, yes.

Rosner: So will the Ukrainians eventually destroy that bridge and cut Crimea off entirely from Russian supply routes?

Jacobsen: I would not put it as “we.” I am neither a Ukrainian soldier nor a Russian soldier. I am a journalist reporting from here.

The human rights record is much more favorable to the Ukrainian side. I do not foresee the Ukrainians undermining that moral authority, but if they did, my reporting would reflect it. It is not about which team I support. I am not really into sports analogies anyway.

In general, though, that appears to be part of the strategy. Ukraine has been conducting deeper strikes against supply lines in an effort to slow down and constrain Russian operations. Whether it is limiting Russia’s ability to launch attacks from the sea by targeting ships, or reducing the supplies available for land operations, the goal is to disrupt logistics and slow the overall war effort.

Rosner: I want to move on to another topic.

Jacobsen: Sure. Anything you want.

Rosner: I have developed an exponential model of wiping my butt.

Over the past six months or so, I have become more conscious of the fact that we cut down trees and turn them into toilet paper to wipe feces off our buttocks. It seems wasteful. It strikes me as the sort of thing future civilizations might look back on and say, “You people were remarkably profligate in your use of natural resources.”

As a result, I have been developing more efficient strategies for using less toilet paper, origami-style folding, unfolding, and refolding techniques. In the later stages of the process, I can get three wipes out of two squares of toilet paper.

My exponential model works like this. Sometimes I get lucky. I take magnesium, which occasionally gives me what I call a “Shawshank poop.” The reference is to Andy Dufresne in The Shawshank Redemption, who crawled through a river of sewage and emerged clean on the other side.

A level-zero poop is one that comes out so cleanly that you wipe once and the toilet paper remains clean. That is a Shawshank poop. Unfortunately, that only happens about a quarter of the time for me.

Jacobsen: Is it a redemption experience?

Rosner: Well, it is certainly nice to have a clean poop, but usually it is not that clean. Then we move to the exponential model.

To avoid having an itchy butthole an hour later, you need to reduce the residual fecal matter by at least a factor of 10,000, and perhaps 100,000. In other words, we are talking about exponents of four or five.

The first wipe should get you to E1. It should remove roughly 90 percent of the material. The next three wipes should get you past E2, leaving less than 1 percent of the original amount. Then another three or four wipes should reduce it to E4 or E5, one ten-thousandth or one one-hundred-thousandth of the amount you started with.

So I suppose what I am saying is that I have mathematized wiping my butt.

Jacobsen: What kind of geometry would you employ to calculate that?

Rosner: There is no geometry, except that my butt is not straightforward. It has a lot of scar tissue from hemorrhoids and hemorrhoid surgeries. Wiping is not a simple process. I have to work the toilet paper around carefully to make sure I am getting all the places where fecal matter could become trapped.

At one point, I talked so much about my butthole that people finally said, “Enough of this. Let’s see it.” They urged me to go into the bathroom, take a picture of my butthole, and show everyone what all the fuss was about.

So I did. When people draw a butthole, it usually looks like an asterisk, simple and symmetrical. Mine looks more like the Rolling Stones logo, with the lips and tongue. There is just a lot going on down there.

When I wipe, I have to work my way around all the different structures, the badlands, so to speak, to make sure nothing has been left behind.

Jacobsen: Do you feel that the geometry of ribbed tubes could be applied here?

Rosner: No. It is more that there are lots of little protrusions of scar tissue and distended hemorrhoidal tissue.

The last time I had surgery, the doctors explained that internal hemorrhoids were the source of the bleeding and other problems. Those are relatively easy to treat because they have fewer pain-sensitive nerves. External hemorrhoids, on the other hand, generally cause fewer medical complications but are much more painful to operate on.

As a result, the surgeons removed the internal hemorrhoids and left the external structures intact. Those occasionally become inflamed and contribute to the unusual topography.

As I said, it resembles badlands terrain. Think of those eroded desert landscapes you see in movies like Cars, with buttes, mesas, and winding canyons. My butthole is essentially a miniature version of that landscape, a sort of Grand Canyon with more geological features than most people possess.

Jacobsen: We talked about contextual compute, or contextual computing, many months ago when I was still living in Canada. I remember thinking about it while conducting an interview by a lake and lagoon in a small town.

Perhaps it was the fresh air, or perhaps it was the conversation itself, but I was speaking with a computer-chip expert who was designing a new computational architecture. This was an idea that was pitched to me rather than the other way around.

The architecture combined traditional CPUs, GPUs, and quantum-computing elements on a unified platform. My argument, which we discussed at the time, and which he liked both as a term and as a concept, was what I called “contextual compute.”

The idea is that you have an intelligent algorithmic layer sitting above the hardware stack. Beneath it are the various computational resources and their capabilities. The higher-level system determines the most efficient way to allocate workloads, deciding when to use a CPU, a GPU, a quantum processor, or some combination of the three.

Rosner: This is actually a major topic right now, and it leads directly into what I wanted to discuss.

Jacobsen: My argument for contextual compute is that it represents one of the most significant frontiers in computing because it effectively makes the cost of computation and the cost of electricity converge toward the same optimization problem.

In an ideal system, those two variables become increasingly aligned. The Venn diagram closes until it is almost a circle. Computation becomes optimized not merely for performance but for energy efficiency and resource allocation.

That connects to discussions by Sam Altman, who has described a future in which compute becomes something like a utility. Just as people pay for water, electricity, or internet access, they would pay for computational capacity through a kind of universal spigot.

Many people reacted to those ideas with interpretations that I thought were misguided. They imagined dystopian futures in which access to compute would be tightly controlled. In reality, we already have versions of that model today. Paid AI services provide varying levels of access to models, agents, and computational resources.

The concept is not fundamentally new. He was simply describing it on a broader and more universal scale. A lot of the criticism came from people misunderstanding the proposal.

Rosner: There is a lot of chaos around this right now, financial chaos in particular. But one way to think about contextual compute is that, without it, a system is effectively autistic.

I view autism as an issue of final-stage integration rather than basic capability. People on the spectrum often have highly efficient submodules for things such as mathematics, music, or language. The challenge is frequently the integration layer that coordinates those capabilities.

That is what you are talking about with contextual compute. You have a chip with multiple methods of processing information, and the final software layer determines how best to utilize the available computational resources.

Quantum computing, for example, is only useful for certain classes of problems. For many tasks, conventional logic circuits are more efficient. The challenge is deciding when to use each computational approach.

What is happening in AI right now is related to that. Anthropic and OpenAI are separate, competing companies. One of the competitive pressures in the industry is the race toward enormous valuations and eventual public offerings.

One issue is that task-specific AI is generally more efficient than pursuing artificial general intelligence, or AGI. Training systems for narrowly defined purposes requires far less computational power than building broadly capable intelligence.

At the same time, there has been a great deal of journalism suggesting that the people running AI companies are trying to build “AI gods.” That is a sensationalistic way of describing the push toward AGI.

Jacobsen: If I may, this is on the same level as the “God Particle” discussion we had the other day.

Rosner: Exactly.

Jacobsen: It is the same type of hyperbole.

Rosner: Sensational language designed to attract attention. The reality is that there is a significant push toward AGI, meaning a form of artificial intelligence capable of performing at or above human levels across a very broad range of domains. The idea is a system that can answer almost any question and outperform human experts in many fields.

We have already seen AI systems contribute to solving long-standing mathematical and scientific problems. Whether that reflects genuine general intelligence or extremely powerful specialized systems is a matter of debate.

My concern is that there appears to be a willingness to spend hundreds of billions of dollars pursuing AGI rather than focusing on more targeted and efficient applications of AI. Critics worry that AGI could be less predictable and potentially more dangerous than systems trained for specific tasks. It also consumes vastly greater resources.

At the same time, AI is reshaping labor markets. Investors have rewarded companies that reduce costs through automation, and AI-related firms have driven much of the recent enthusiasm in stock markets.

The problem is that stock valuations can rise because companies are reducing payroll expenses rather than because they are producing significantly more value. If enough jobs disappear without equivalent replacements being created, that could become economically destabilizing.

The concern resembles what happens when a water table is overdrawn. Things may appear stable for a while, but eventually the underlying structure weakens and sinkholes begin to appear.

Some people argue that AI is creating a similar dynamic in the labor market. Workers displaced from traditional jobs move into lower-paying or less stable forms of work, masking the extent of the disruption. Over time, however, the cumulative effects could become more visible.

Whether that leads to a speculative bubble bursting remains uncertain, but it is one of the major concerns being debated right now.

Compounding the issue, many people worry that highly influential technology leaders are making decisions with insufficient oversight. Critics argue that governments have not established clear frameworks for regulating advanced AI systems, creating additional uncertainty about the future trajectory of the technology and the economy.

Comments?

Jacobsen: What do you think the next major innovation in human sanitation will be? For example, humanity probably progressed from using natural water sources, vegetation, and other rudimentary methods, to toilet paper and modern plumbing. More recently, we have seen advanced bidets, particularly in Japan, that wash, dry, and automate much of the process. What comes next?

Rosner: Well, the next step is probably medical diagnostics based on your poop.

I believe that in Japan there are already toilets capable of performing basic health assessments. They do not yet operate at the level of conducting detailed microbial analyses of your stool, but that seems like the logical next step.

For example, we largely stopped widespread COVID testing years ago. Outside of specific situations, such as people with symptoms using at-home tests, or hospital patients being tested for respiratory illnesses, we no longer monitor COVID prevalence through mass individual testing.

Instead, public-health agencies increasingly rely on wastewater surveillance. Scientists analyze sewage samples for viral fragments and other biological markers. By examining wastewater from hundreds or even thousands of treatment facilities, they can estimate COVID prevalence and identify which variants are circulating in the population.

I imagine the next stage of toilet technology would involve applying similar techniques at the individual level. A toilet could analyze your gut microbiome and identify potential health issues before symptoms become severe.

Take ulcers as an example. For many years, people believed ulcers were primarily caused by stress. We now know that many are associated with infection by Helicobacter pylori (H. pylori). A sufficiently advanced toilet might detect elevated levels of H. pylori and alert you before ulcers develop.

So yes, I suspect personalized biological monitoring is the future of toilet technology.

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.