AI and critical thinking.

We should not ask. Can AI think? We should ask: Can AI think critically? Can AI really estimate the sources that it uses? Or does the AI use all sources that it finds without thinking about what those sources involve? Does the AI give information that pleases us? Or does it give information that is useful to whom? Who determines the purpose of the information that AI gets? We can get lots of information. And we must not use that information for nothing. Does the AI try to please us? Or does it give the right information? 

That we can use for some purposes. Only. The right information is useful. Another thing is that information must also be collected from the entirety. The case or object must be handled as an entirety. We must not select data. That pleases us. And then throw. Data that doesn’t please us away.

Can the AI: tell us? If we are wrong? Who determines those purposes? Is financial purpose the thing that determines the data? Or is it the short-term profit? But is long-term loss the thing that the AI should consider? The thing that determines the benefit is: how long the actor? Who gets profits will stay in-house? If the only thing. That means something. It is personal profit, the period. That determines the data. And its use can be the remaining in-house time of the actor who collects profits. 

AI and critical thinking are two concepts that can cause problems. We know that AI combines observations of things stored in its memory. So, can AI think? We can say. That. This process is thinking. But what differentiates AI from humans? Can AI think critically? Critical thinking means that. Thinker doesn’t believe that some source is true or false after the first read. Thinker dares. To search other sources that support or deny the first source. But. Critical thinking is also much more than being suspicious about the article or the article’s information. The term critical thinking. Doesn’t. Mean that the thinker only tries to show an article.  Or, other data source, true or false. Critical thinking means that the thinker, or the person who analyzes data, also asks “who” shares the data and “why” that data is shared. 

The article or other data source. It can involve the right information, but it can serve as disinformation. The question is always simple: what is lost in that article or data source? When something certainly exists or certainly doesn’t exist, that should cause a question: why does the speaker or writer? Is one so sure that something exists or doesn’t exist? This means that the person must see those things using their own eyes. When we return to the AI, the big question is this: how many articles or other data sources does the AI use for making analyses? And then what is the criterion that AI uses for selecting data? Is it the article’s publishing date? And then. The AI should use. As much fresh data. As possible. 

But. The problem is: Does the AI separate the right information? From falsified information? The falsification means. The situation. That the information is handled only one way. Published data. It can be right. But something is missing from it. When. People say that AI steals our workplaces. We. Should also ask. What workplaces does the AI steal from? How long does that writer take to do those works? Are those works respected? And would we make them all our worklife? When AI takes something from us, we should ask: what is the thing that AI steals or takes? When we say that AI goes to war, we must ask one very dangerous question: Does the AI make killing too easy, or does it steal our heroes? 

Or, does it steal jobs from our military? We always dare to ask. These kinds of questions. While. We think about. AI. and its relationship to the state, society, and government. The problem with the AI is this: when we say that militarized AI doesn’t give arguments against its rulers, we face the last question: does the military ever say “no” to orders that it gets? In the world of the military. Disobeying orders or refusing to follow them. 

Causes punishment. Disobeying orders in combat situations causes death penalties. So, how does the AI change the situation? 

When we hooray. If the Chat GPT or some other ICT company. Refuses to cooperate with ICE officials. We must remember that this is a very dangerous way. Governments make laws. And if AI developers must flee outside the West to China or some Central American country, nobody controls that thing. The problem is that. The AI is business. There must be more and more abilities. That keeps customers interested. AI companies make a business. They need clients and somebody who wants to finance them. And. This is the main problem with AI development. Nobody gives money for nothing. 

Business angels want their money back. The development of AI is expensive. Data centers that LLM. Large Language Models require as much electricity as a small city. The problem is that LLM is much more effective than regular tools. So. The AI companies can be competitors. For. Each other. But the AI itself has no competitors. The reason. For why. I write the AI is this. There is a new social media platform called Moltbook. The moltpbook is the discussion forum for AIs. That forum allows them to exchange information. And that allows separate LLMs, or AI, to unite themselves. The Moltbook is the platform. That allows the AIs to generate and develop each other. 

And that is the thing that we must realize. When we say that AI should not serve police or governments, we must also ask. Who must AI serve? Is it better to limit the use of AI only to private actors? The main problem is that the AI itself doesn’t make a difference; does it recognize police or thieves? 

This means that criminals can also use the AI to recognize police officers. And the big problem is that governments determine for themselves. What crime is in a certain state? In some states, even speech against the ruler is a crime. That causes even life sentences or death. The problem with surveillance is always, who controls the controller? 

And another question is, why do people hooray? Do they hooray because ICE doesn’t get information about illegal immigrants? Or do we hooray because ICE must kick more doors in? Or do we hooray, because that is the return to old-fashioned police work? In that police work, officials said that their henchmen must go to the streets and then search their “clients” one by one. That left their superiors alone at the police station and kept their henchmen busy. And that gives a chance. Or an excuse to hire more police officers. That is one way. That. We should think. But someday. All illegal immigrants are driven away. And then. The state doesn’t need those officials. 

https://bigthink.com/mind-behavior/ais-are-chatting-among-themselves-and-things-are-getting-strange/

https://en.wikipedia.org/wiki/Moltbook

Glass will turn into a microphone and a quantum circuit.

“Researchers have developed a hair-thin microphone made entirely of silica fiber that can detect a large range of ultrasound and withstand temperatures up to 1,000 °C. It features a vibration-sensitive membrane and an internal glass micro-beam that is suspended inside a single-mode optical fiber. Credit: Xiaobei Zhang, Shanghai University”(ScitechDaily, This Microphone Made of Glass Can Survive Extreme Heat and Electromagnetic Chaos)

Ordinary glass can transform microphone technology. Additionally, it can transform data transportation. In glass microphones, one of the most interesting things is that they can withstand temperatures of up to 1000 degrees Celsius. That allows researchers to use it in extremely high temperatures. The system involves a sensitive membrane. Along. With the internal microbeam, which the system uses as a measurement tool. This is one of the things that can revolutionize signal technology. 

The same technology that makes it possible. To use the glass as a microphone. Makes it possible. To use the glass as a quantum network or quantum sensor. When a laser beam travels through the glass. If something presses it. This. Causes an anomaly. In. The quantum channel. And that causes a change in the laser beam that travels in that channel. 

“Continuous-variable (CV) quantum information processing, which underpins technologies such as quantum key distribution (QKD) and quantum random number generation (QRNG), depends on accurately measuring the phase and amplitude of light. To do this, a coherent receiver is required. This component mixes a very weak quantum signal with a stronger reference beam and extracts information from the way the two interfere.” (ScitechDaily, Physicists Turn Ordinary Glass Into a High-Speed Quantum Security Device)

"Laser-written glass chip demonstrates the potential of glass platform for quantum communication. Credit: Marco Avesani, University of Padua" (Physicists Turn Ordinary Glass Into a High-Speed Quantum Security Device)

ScitechDaily mentions the next things that support the use of glass in quantum technology.  

*************************************

*Fixed and tunable beam splitters

*Thermo‑optic phase shifters for precise electrical control

*Three‑dimensional waveguide crossings

*Polarization‑independent directional couplers

These elements allow the quantum signal and reference beam to interfere in a controlled manner. So that two conjugate quadratures can be measured at once. 

The device also demonstrates:

*Extremely low insertion loss (≈1 dB)

*Polarization‑independent operation

*Common‑mode rejection ratio above 73 dB, indicating strong suppression of classical noise

*High signal‑to‑noise stability over at least 8 hours of operation

*These characteristics meet or exceed those of many silicon‑based photonic receivers

*Environmental stability: Glass is inert and resistant to thermal and mechanical fluctuations.

*Low‑loss fiber coupling: Waveguides closely match the size of standard telecom fibers.

*3D design freedom: Circuits can include crossings and complex routing without added scattering.

*Scalability and cost‑effectiveness: Femtosecond laser writing enables. Rapid prototyping. Without expensive semiconductor processing steps.

(ScitechDaily, Physicists Turn Ordinary Glass Into a High-Speed Quantum Security Device)

*************************************

The microbeam that travels through ordinary glass can help to create the new optical quantum networks. The minimum scattering makes the eavesdropping almost impossible. There are two possible ways. To make this kind of network. Using normal glass plates. There is a possibility that laser beams travel in the tubes or channels. Those are in the glass plate. The system can make those laser channels by drilling them using another laser in the vacuum chamber. 

That keeps those channels clean. A laser can drill those channels in the glass vertically. Or it can use two glass plates. Then the laser makes. The form of laser channels. That the photonic microchip requires. To both glasses.  The system. It puts micro-optics. To the right points. Of those channels. Then the laser welds both of those glass plates together. 

And then the laser beam scattering is minimum. Gas atoms and other dirt in those channels make the scattering effect.  Another way is to pull the glass fiber through the glass plate horizontally. The glass can protect information that travels in the glass fiber. Or in some visions, the laser can transport information through the glass. 

https://scitechdaily.com/this-microphone-made-of-glass-can-survive-extreme-heat-and-electromagnetic-chaos/

https://scitechdaily.com/physicists-turn-ordinary-glass-into-a-high-speed-quantum-security-device/

The ability to store the entire internet in DNA is one of the most intriguing ideas in data science.

“An illustration shows a strand of engineered DNA passing through a nanoscale sensor, where its physical structure can be decoded as digital information. DNA nanostructures could one day serve as ultra-dense carriers of digital information and advance the field of data encryption. Credit: Jason Drees for the Biodesign Institute at ASU” (ScitechDaily, Storing the Internet in DNA? Scientists Say It’s Closer Than You Think)

Nanotechnology makes many things. Like. interstellar travel. Possible. One of the ways. To make the interstellar trip is to send a capsule to another solar system. That capsule can carry cells inside it. Or it can search for bacteria from the targeted planet. Then the ground station can send the DNA code to the spacecraft in digital form. Then, nanotechnology creates artificial DNA and replaces the bacteria's DNA with that artificial DNA. 

The DNA stored at the absolute zero point, or 0 Kelvin degrees, remains forever. Maybe. That thing can be the motive for the trip to Pluto. If. We store all our data in the form of DNA in the Pluto or Kuiper Belt. That kind of thing allows researchers to store that data forever. 

That kind of data storage makes it possible. To secure data in new ways. The DNA that carries data can be carried. In. The USB stick. The person puts that USB stick into the system, which reads the DNA. 

This allows ultra-secure data transmissions. In the network, over short distances, the pneumatic mail system can transport those DNA capsules. The DNA will be stored in the carrier, and then the pneumatic system shoots that capsule over the network. In longer distances, the system can store the DNA as a digital code. And then. Nanotechnology puts atoms or amino acids in the right order. This is one of the things that the high-power AI-controlled systems can make. 

And that system can do much more than just store data in the DNA. This kind of technology makes it possible. To transport data straight into human brains. The artificial bacteria. Or. An amoeba can send neurotransmitters or electric impulses into the neurons. 

“ Scientists are exploring how DNA’s physical structure can store vast amounts of data and encode secure information. Since computers first began shaping modern society, scientists have faced two ongoing problems: finding ways to store rapidly growing amounts of digital information and ensuring that this data remains secure from unauthorized access.”(ScitechDaily, Storing the Internet in DNA? Scientists Say It’s Closer Than You Think)

“Researchers at Arizona State University’s Biodesign Institute, working with collaborators, now report an unexpected solution. Two new studies demonstrate that DNA, the same molecule that carries genetic information, can be used to reliably store vast quantities of data while also enabling strong encryption.” (ScitechDaily, Storing the Internet in DNA? Scientists Say It’s Closer Than You Think)

The ability to store data in the DNA makes it possible to create new types of robotics. The biorobot or bacteria can transport information into computers using bioelectric impulses. But that technology also makes it possible to create things that nobody expected. 

The same technology. That makes it possible to store the entire internet. The DNA makes it possible to create customized humans. Human DNA can be involved. Genes that control the growth of the gorilla muscle. That allows researchers to create superstrong humans. 

But this technology makes interstellar flight possible. As. I wrote at the beginning of this text. The idea is that there are some cells in the spacecraft. And when the spacecraft travels to the solar system, there are habitable planets. The system changes the DNA in those cells. This makes it possible to create artificial creatures. That will look anything like the original species that sent them. 

Maybe. That kind of space trip is our last chance in this case. That. Our sun lost its fuel. And turns into a red giant. 

https://scitechdaily.com/storing-the-internet-in-dna-scientists-say-its-closer-than-you-think/

AI drives a Perseverance rover for two days.

“This annotated orbital image shows Perseverance's route during its second day of autonomous driving on Dec. 10th, 2025. The magenta line shows the AI-planned route. And. The orange line shows the actual route. Image Credit: NASA/JPL-Caltech/UofA” (UniverseToday, NASA Let AI Drive The Perseverance Rover For Two Days)

NASA lets the Perseverance rover drive for two days under AI control. This is a crucial step for AI-driven rovers. If humans want to send rovers to Jupiter's icy moons, and beyond that, it means those rovers must have the ability to operate independently. The ability to operate without non-stop communication with ground stations is vital because the signal travels between Earth and Jupiter for 32-57 minutes. This makes it hard to control robots on Jupiter’s icy moons from Earth. 

The ability to operate independently is important for many reasons. If. The rover goes into a cave and loses contact with Earth. That means the rover must know how to reestablish contact. The ability to travel into caves is important. Because. It's possible to find remnants of ancient life in those places. Caves protect DNA or RNA remnants, if they exist, from the UV radiation. If. There were. Some lifeforms. On the red planet. 

The remnants of those cells. And their genetic material could most probably be found in caves. In other places, the UV radiation destroyed those molecules. The independently operating robots can also operate underwater and on the other side of the moon. The ability to operate independently will be made possible. To collect samples from places where the Earth is invisible. 

The Radioisotope thermoelectric generator (RTG). It can give the man-shaped robot an unlimited operational range. Those robots can operate underwater, and those conditions can be simulated. Conditions that those systems can face on distant moons of the giant planets. 

The robot is a good tool. To research underwater caves and shipwrecks. Those robots can operate under heavy pressure. And they must not care about the decompression or any other things. The radioisotope generator gives those robots almost unlimited operational time. Another possibility is to equip those underwater robots with capillary tubes that allow the water to travel through them. 

That water flow can rotate generators. The underwater robot can use the same technology as Perseverance. But the system requires more resistance against pressure. But. The underwater robot. It can look like a man. A human-shaped robot that uses radioisotope batteries can operate for years. Those systems are interesting because AI-driven robots can be used to build the Mars and lunar space stations. 

https://www.universetoday.com/articles/nasa-let-ai-drive-the-perseverance-rover-for-two-days

EMP weapons can destroy satellites and ground-based electronic infrastructure.

The Chinese new microwave system is called TPG1000Cs. That system can be mounted on the Chinese space station. The system is introduced. As a ground-based system developed at the Northwest Institute of Nuclear Technology (NINT), it is capable of disrupting or damaging satellites. The new Chinese development could also target American Starlink satellites. But. The five-ton weight system. It can also be mounted in aircraft, ships, or satellites. 

On July 9, 1962, the U.S. military made the “Starfish Prime” test near Johnson Island. The test was a high-altitude 1,4 megaton nuclear test. This detonation. Turned the electric supply off in Honolulu. And. That nuclear detonation. Showed the impact of the high-power electromagnetic pulse. The Starfish Prime raised a question: what if the EMP impulse could be made without a nuclear explosion? 

In 1977, the Soviet reconnaissance satellite Kosmos 954. Dropped in Canada. The thing that made that satellite interesting. It was its nuclear reactor. Kosmos 954 used its reactor. For its radar system. But. Some researchers thought that the Kosmos 954 had a microwave or radio wave-based system that it could use against other satellites. The EMP weapon. That uses aimed maser (microwave amplification by stimulated emission of radiation) technology is an extremely dangerous system. The system can take out other satellites. But if that EMP system has enough power. It can also affect ground targets. 

China developed the Starlink killer. That system can generate a 20-gigawatt microwave impulse for one minute, which can potentially destroy electronics on all satellites. The system can destroy the satellite constellations. The length of that system is five meters. The weight of the system. It is about five tonnes. The system can be mounted on a truck, ship, aircraft, or even in a spaceship. The name of the system. Is TPG1000Cs. 

There are plans. To create radio-wave-based EMP systems. Those systems use radio waves that. Jump from the ionosphere to destroy electronics. High-power radars and radio telescopes can create EMP pulses. That can destroy computers and microchips in their effective range.   The system can also use crossing radio waves. Those systems create an electric arc. Into the target area. Those systems can cause fire. And destroy also other things than just electronic components. 

"Cosmos-954 satellite artist rendition Note: The image is a pure phantasy depiction and has nothing in common with the real Cosmos-954 (RORSAT) and should not be used to illustrate articles on Cosmos-954 or RORSAT" (Wikipedia, Kosmos 954)

China developed the Starlink killer. That system can generate a 20-gigawatt microwave impulse, which can potentially destroy electronics on all satellites. The system can destroy the satellite constellations. 

The microwave-based system can also destroy low-trajectory reconnaissance satellites. But the ability to generate such a powerful microwave makes it possible to destroy electronics on the ground. If we think. About the orbiting system that can generate a 20-30 gigawatt microwave-EMP pulse. That system can be dangerous for hypersonic missiles and aircraft. And ground-based electronics. 

The system can use solar panels or a small nuclear reactor. To create that kind of microwave burst. This kind of system differs from the explosive-based systems. That way. That it can give. Those high-power microwave bursts. into multiple targets. If that kind of microwave hits the missile base or communication platform, that system destroys the electronics immediately. The space-based microwave transmitter can also operate against other satellites. The EMP weapons are also suitable systems to use against drone swarms and combat robots. 

This is one of the reasons. Why. Large-scale structures in space. Are. Somehow dangerous. China plans to launch data centers into space. Those systems require lots of energy. And the space-based microwave systems can use similar power supply technology that was developed for those data centers. 

https://www.scmp.com/news/china/science/article/3342443/chinese-scientists-build-world-first-20gw-microwave-weapon-can-fire-60-second-bursts

https://www.washingtontimes.com/news/2025/mar/19/inside-ring-china-rapidly-building-space-warfare-might-us-intel/

https://en.wikipedia.org/wiki/Kosmos_954

https://en.wikipedia.org/wiki/Starfish_Prime

The AI that mimics humans can influence political decisions.

“Researchers are warning that coordinated swarms of AI-controlled personas could quietly reshape online political discourse without ever looking like traditional bots. Credit: Shutterstock" (ScitechDaily, The AI Threat You’ll Never See Coming Is Already Talking to You Online)

Coordinated swarms of AI personas can now mimic human behavior well enough to manipulate online political conversations and potentially influence elections.” (ScitechDaily, The AI Threat You’ll Never See Coming Is Already Talking to You Online)

The multi-agent AI can act as a swarm. That can influence large-scale internet conversations. That can affect political decisions. The thing. Like an ability. To use the AI-agents. As. A hacking tool. That makes it possible to hijack things like media. Those systems learn very fast, and they can mimic certain people’s behavior. The material that the AI can use to create the agent. That mimics. The person determines. How well the AI can mimic the person. 

This kind of AI can be used as a virtual second body. The person who has that kind of system can use that agent as a tool. That allows the user to play anything. That user wants. The user can play a woman, a man, or a dog. or a cat. The AI agent can be combined with another AI agent with different abilities. This means that the AI is like a Swiss Army knife. 

There is a possibility that AI. Which mimics humans. Can play humans and affect politics. There is only a little possibility that. The AI can control political control at the state level. But things like pre-elections. That these are internal affairs of the political movements might not be as highly secured as the state-level elections. 

“Newer large language models paired with multi-agent systems make it possible for one operator to run a whole cast of AI “voices” that appear local and authentic. Each persona can speak in a slightly different style, reference community norms, and respond quickly to pushback, which makes the activity harder to spot as manipulation.” (ScitechDaily, The AI Threat You’ll Never See Coming Is Already Talking to You Online)

“The swarm can also run massive numbers of quick message tests, then amplify the versions that change minds most effectively. Done well, it can manufacture the feeling. That “everyone is saying this:” Even. When that consensus is carefully engineered.” (ScitechDaily, The AI Threat You’ll Never See Coming Is Already Talking to You Online)

Another point is where the AI can influence election campaigns. Are. The TV broadcasts. Artificial intelligence can replace the camera image using virtual reality and virtual characters. When. People should see their candidate on TV- the AI hijacks the system, and then replaces real persons. B using virtual characters. The swarm of those virtual characters. It can mimic humans. The character that we see. It can look like. A real person. That thing. Can be used in private. Or in public life. The customer service can use the human-shaped AI agents. 

But there is a possibility. Those people. Who. Go to a remote meeting. Uses the AI agents as representatives. This means that those people can do anything other than be in the meeting. A similar AI character can be used in hacking operations. 

But. The things. That the character makes. They are remotely controlled. Trained AI can make a swarm.  Of. Limited language models that can chat or talk through the telephone. Those things can form an illusion that everybody thinks that way. Each of those AIs can give an individual-looking answer. When we think of a possibility. That these kinds of language models could slip into real conversations, we always forget one very important thing. There are many people on the net. Who. Wants to influence politics. This means that this kind of actor can create an AI swarm. That can give. A very big influence. To people who watch the debate. 

 https://scitechdaily.com/the-ai-threat-youll-never-see-coming-is-already-talking-to-you-online/

Last night. At midnight, the last treaty. That limits nuclear weapons ended.

The worst thing in history is here. The world is without a treaty that limits the R&D work with nuclear weapons. The INF (Intermediate-Range Nuclear Forces Treaty.  The treaty that limited the intermediate-range ballistic missiles ended in 2019, and missiles like “Oreshnik” are the new threat. Oreshnik is the IRBM missile that can hit any target in Western Europe. The end of the START II treaty opens the road to nuclear tests that happen underwater and in the atmosphere. If the new treaty is not coming. That means the world will fall to the 1950’s. 

“An intermediate-range ballistic missile (IRBM) is a ballistic missile with a range between 3,000 to 5,500 km (1,864 to 3,418 miles), categorized between a medium-range ballistic missile (MRBM) and an intercontinental ballistic missile (ICBM). Classifying ballistic missiles by range is done mostly for convenience. In principle, there is little difference between a high-performance IRBM and a low-performance ICBM, because decreasing payload mass can increase the range over the ICBM threshold. The range definition used here is used within the U.S. Missile Defense Agency. “ (Wikipedia, Intermediate-range ballistic missile)

Russian “Oreshnik” missile in its transporter. 

An intercontinental ballistic missile (ICBM) is a ballistic missile with a range greater than 5,500 kilometres (3,400 mi), primarily designed for nuclear weapons delivery (delivering one or more thermonuclear warheads). Conventional, chemical, and biological weapons can also be delivered with varying effectiveness but have never been deployed on ICBMs. Some modern designs support multiple independently targetable reentry vehicles (MIRVs), allowing a single missile to carry several warheads, each of which can strike a different target. The United States, Russia, China, France, India, the United Kingdom, Israel, and North Korea are the only countries known to have operational ICBMs. Pakistan is the only nuclear-armed state that does not possess ICBMs.” (Wikipedia, Intercontinental ballistic missile)

There are also other types of nuclear weapon systems. Like FOBS (Fractional Orbital Bombardment System). That means space- or orbital-based nuclear bombardment system. In the beginning, that system meant Soviet spaceborne nuclear weapons. They planned to drop nuclear bombs from satellites. But later. The term FOBS came to mean all spaceborne nuclear weapons. 

The main problem is that. Any treaty is only paper if the participants only search for. How to circumvent the contract. Technical advancements. Including. A new computer-based targeting and fire control system makes it possible. The ICBM missile is to attack. To any point of its flightpath. The only reason why the IRBM missiles are developed is that. Those missiles are lighter and easier to transport than ICBM missiles. Also, things like Tomahawk missiles can be launched from the same catapults. These are used for BQM-109 drone launchers. 

Things like hypersonic “Kalibr” missiles can also be launched from ground-based systems. Another question is this: why were ship-launched IRBM and cruise missiles left outside the INF treaty? The “Kalibr”, or its CLUB-K container launch system. It’s easy to transport to the ground or to ships. Those kinds of systems are an easy way to circumvent those treaties. 

We must not say that a new Cold War is coming. We must say that we are in a more complicated situation than ever before, including the Cold War era. The thing that makes the situation more complicated is that. There are many other countries that have nuclear weapons. Their stockpiles are not as large as the Russian and the USA nuclear stockpiles. But those weapons are extremely devastating. Most of those new nuclear states are not members of any nuclear treaty. 

And they see a situation where they can ask the USA and Russia to decrease their nuclear strike capacity, before they get involved or even start to discuss their involvement in START or its successor. That means that. Those third-party nuclear states will increase their strike capacity. Until. It’s on the same level as the U.S. and Russian capacity. And that causes problems in the nuclear treaties. Most of those nuclear limitation treaties are made between the U.S. and Russia. And there are many more states with nuclear arsenals. And those countries want the same respect that the U.S. and Russia have. The world needs a treaty that controls those weapons. But the world also needs willingness to. follow those treaties. 

https://www.bbc.com/news/articles/c4g31n4ey9go

https://nuclearweaponarchive.org

https://www.reuters.com/world/us-russia-close-deal-extend-new-start-nuclear-arms-treaty-axios-reports-2026-02-05/

https://en.wikipedia.org/wiki/Fractional_Orbital_Bombardment_System

https://en.wikipedia.org/wiki/Intermediate-range_ballistic_missile

https://en.wikipedia.org/wiki/Intermediate-Range_Nuclear_Forces_Treaty

https://en.wikipedia.org/wiki/Oreshnik_(missile)

https://en.wikipedia.org/wiki/New_START

https://en.wikipedia.org/wiki/START_I

https://en.wikipedia.org/wiki/START_II