Researchers developed AI that challenges Chat GPT under 50 dollars.

The new open-source AI, S1, challenges traditional AI models. The system uses a similar reasoning model. That Gemini uses. The reasoning language model is not, actually another name for LLMs.  And that is a new step for the AI. Open-source solutions challenge traditional AI. Open-source is not the same as free software. In open-source solutions, developers and other experts can investigate the source code. And that makes the code safer. But that requires deep knowledge of the programming. 

And there is a lot of code in complicated applications that R&D workers must check. Free software means software that we can download it free. But we cannot check the source code, and that makes it possible to slip malicious components in the code. So most so-called trojan horses are freeware. That means the user downloads some very useful-looking software from the net. And then there is the component that acts as spyware. 

Above: A curated list of awesome large AI models, or foundation models, for reasoning.(GitHub, Awesome-Reasoning-Foundation-Models)

Gemini AI tells about the LLM and LRM this: 

"Large language models (LLMs) and reasoning language models (RLMs) are related, but not exactly the same. Here's a breakdown:" (Gemini)

"Large Language Models (LLMs): These are deep learning models trained on massive amounts of text data. They excel at understanding and generating human language. LLMs can perform tasks like translation, summarization, and even creative writing. However, they are primarily focused on language and may not have strong reasoning abilities." (Gemini)

"Reasoning Language Models (RLMs): These models are designed to go beyond basic language tasks and perform reasoning. They can handle tasks that require logical thinking, problem-solving, and drawing inferences. RLMs often incorporate techniques that allow them to understand the underlying meaning and relationships within the text, rather than just processing the surface-level language." (Gemini)

"Key Differences:"

"Focus: LLMs focus on language understanding and generation, while RLMs emphasize reasoning and problem-solving." (Gemini)

"Capabilities: LLMs are good at language-based tasks, but may struggle with complex reasoning. RLMs are designed to handle reasoning tasks, but may not be as strong in pure language generation as LLMs. "(Gemini)  

Techniques: RLMs often use techniques like symbolic reasoning, knowledge graphs, or logical inference to enhance their reasoning abilities." (Gemini)

Relationship:

"RLMs can be considered a subset of LLMs. They build upon the capabilities of LLMs by adding reasoning mechanisms. "(Gemini)

Many LLMs are now being developed with improved reasoning capabilities, blurring the lines between LLMs and RLMs.   (Gemini)

"In Summary:" (Gemini)

"While all RLMs are LLMs, not all LLMs are RLMs. RLMs are a specialized type of LLM that focuses on reasoning and problem-solving. "(Gemini)

The S1 is the signal for the world. The AI is cheaper than we expect. Things like morphing neural networks that share their missions over a large number of less powerful microchips. Morphing neural networks makes it possible to run complicated applications without the latest processor technology. And that makes the AI more complicated than ever before. The AI can generate software. That is extremely complicated. The AI and large language models can turn into virus generators. 

Some people want to give AI to every man and woman on our planet. That can make things like programming and mathematical algorithms more powerful. The ability to connect every smart device together into the giant neural network makes the planet safer. But in the wrong hands the systems that make streets safer turn into surveillance tools, that are more dangerous than we expected. Persons like Stalin and Hitler cannot ever be so dangerous as a person who controls global AI. 

The AI can save human lives. It can detect heart problems before the person even notices those things. But the AI can used to create programs for next-generation jet fighters and tanks. And it's easy to hide malicious code in complicated programs. The AI can collect data from every device, filter, and sort it.

https://www.ceotodaymagazine.com/2025/02/open-source-ai-model-s1-developed-for-less-than-50-challenges-industry-norms/

https://www.digitalocean.com/community/tutorials/understanding-reasoning-in-llms

https://github.com/reasoning-survey/Awesome-Reasoning-Foundation-Models

https://www.edps.europa.eu/data-protection/technology-monitoring/techsonar/large-language-models-llm_en

https://www.hpe.com/fi/en/what-is/large-language-model.html

https://www.rws.com/blog/large-language-models-humans/

Researchers trained neurons to believe that they were real brains.

"A 3D-printed brain model using nanopillars is reshaping neuroscience, allowing neurons to grow in realistic ways. This breakthrough could unlock crucial insights into brain disorders. (Artist’s concept.) Credit: SciTechDaily.com" (ScitechDaily, Scientists Tricked Neurons Into Thinking They’re in a Real Brain"

There is also one very interesting model of alien life. That lifeform is an intelligent robot that is controlled by cloned brains. Theoretically, those cloned brains are possible to make using mini-brains. Cloned neurons that are used in medical tests are basically similar to neurons that are in our brains. It's possible that someday researchers grow brains that are as intelligent as human brains.  

In some visions, the brain is connected to the computer that connects it with surveillance cameras and robot bodies remotely. That kind of system where brains live in nutrients and interact with their environment through mechanical bodies and other things are the tools that can revolutionize science. Or that is the biggest mistake that humans make. 

Researchers cheated cloned neurons think that they were "real brains". That thing made those neurons grow in structured networks. The thing is this kind of ability to program neurons makes it possible to create new biohybrid microchips or researchers can grow cloned neurons in the laboratory and program them for repairing the neural damages. That is one thing that the cloned neurons can make. 

"Bipolar mood swings may stem from two brain “clocks” running at different speeds—one biological and the other dopamine-driven. Scientists created similar patterns in mice, showing dopamine’s key role and offering hope for targeted treatments. Credit: SciTechDaily.com"(SciTechDaily.com, Scientists Discover a Hidden Brain Clock Driving Bipolar Mood Swings)

The discoveries uncovered the hidden clock in brains. They noticed that the bipolar mood swings formed because those clocks worked at different speeds. By synchronizing those clocks is possible to affect that thing. 

Manipulating that clock is theoretically possible to create a situation where a person doesn't need sleep. There is the possibility to create nanotechnical medicals that fill the neurotransmitter stores. That allows the person to operate without sleep. 

But the fact is that brain manipulation is very dangerous. If something goes wrong that can cause terrible damage. The ability to manipulate and train individual neurons makes it possible to transport information into the human brain. That thing is possible in the case of brain damage. Or that thing can be used in cases where a person wants to get some new skills easily. The fact is that the technology that is used in tests will sooner or later escalate into everyman's technology. 

https://scitechdaily.com/scientists-tricked-neurons-into-thinking-theyre-in-a-real-brain/

https://scitechdaily.com/scientists-discover-a-hidden-brain-clock-driving-bipolar-mood-swings/

The mirror bacteria are a big threat.

Representative image of bacteria. (Interesting Engineering, ‘Mirror bacteria’ could cause lethal infections, put all life in danger: Scientists).

Sic-trans isomerism is one of the most interesting things in chemistry. That means that. The protein. Or some other molecule can have different. Or opposite effects. When they oppositely meet their actor. This is why that thing is called mirror molecular isomerism. In biology, the molecules that impact cells ion pumps in opposite directions than they should can destroy the entire cell. So. In this version, the nutrient molecule that travels in the cell the wrong way can destroy the cell. 

That's why medical factories test the "mirror organisms". The idea is that the bacteria that have genomes in mirror order than normal bacteria form proteins that can destroy other bacteria. The thing that makes those bacteria so dangerous is that there are no medicals against them. There are no vaccines or other things that can resist the mirror bacteria. 

The mirror bacteria's purpose is to produce those mirror proteins and mirror genomes that destroy the other bacteria. The mirror-DNA or mirror-RNA is the ultimate tool. That can be used as a tool that can destroy bacteria or any targeted cells. The purpose of the m-RNA molecule is to control the cell organelles. If those molecules travel to those cell organelles backward that thing causes a bad situation. 

The cell reads the RNA or DNA in a certain order. The opposite operating DNA or RNA can cause a situation in which the cells will collapse. The cell and its energy production systems will be terminated in seconds. The DNA that is read backward makes crab that the cell cannot use. But what if researchers create the mirror bacteria? The idea is that the DNA is in the right way, but the base pairs are mirrors. 

There is the possibility that those mirror-bacteria can somehow released into nature. When something exists. It's possible. That it was released from the laboratories. The mirror organisms would be a clear and present danger to nature. But could mirror-DNA form the anti-human? 

Then we must realize that genetics is a dangerous tool. If the base pairs in the human genome turn into the mirror position, could that thing create the "mirror personality"?

The idea of the mirror personality or "evil twin" made the SS director Heinrich Himmler search for the Holy bloodline. The idea was that the Himler's men search for the holy bloodline of Jesus. Then Himmler's men would turn that child into the servant of the nazi ideology. The thing is that Himmler was probably one of the most evil persons in history. 

But the thing is that man should be a big warning to everybody. We are lucky because Himmler and his court didn't know the DNA and its purpose. But in some cases, philosophers ask, what if somebody creates the "evil Einstein"? The thing is that the DNA is acting in many new roles.

The thing that this thing can make is the bacteria that is the mirror of other bacteria. Sometimes. It introduced an idea. What if some test-tube-fertilized humans can get the DNA that mirrors the donor? That means that something causes a situation in which the base pairs turn into the opposite or the mirror. We know that the DNA and genomes have some kind of role in shaping people's minds. 

So could the mirror DNA make things like "anti-Einstein" or "mirror-Einstein"? That thing is one of the most terrifying things in genetics. We know that the psychopathy is the result of anomalies in the brain structures. That means there is a possibility that the "mirror Einstein" would be a psychopath like Himmler himself. 

https://interestingengineering.com/science/scientists-call-for-ban-on-mirror-bacteria-research

When the most advanced tool becomes the enemy of advance.

Why...

Above this text is the image that is not made by using AI. I took that image yesterday evening. That thing required a little bit of trouble and time. Walking into that point, taking my cell phone, and taking that picture took time. When I looked at that picture I realized why people use AI in many things. AI offers easy things to get the job done. That easy tool allows people to make thousands of lines of code in minutes. 

That is possible if the person uses some code libraries. And that makes a person effective. But that thing is not good for advancement and innovation. Also, that way of making programs is not good for data security. If some hackers get those code libraries that allow them to break the systems those codes are used. 

We know that thing. But we ever ask why person makes that thing? Why does that person use libraries and copy-paste? Why that person uses code, that somebody gave to the hard disk? The answer is this something forces a person to make things like that. The nice, well-done code in the hard disk makes work effective. And we can say that generative AI, large LLM, or whatever we call that thing is only to those libraries. 

We can be effective and always use that method to make things. That can make an impression. But then we can say that this method doesn't allow us to advance our programming skills. And that means the AI kills productivity and skills. But is the AI guilty? 

Do we bark the wrong tree? Should we prosecute the need to be effective for that thing? People use AI and code libraries because they have no time to learn things. People are so hurry, that they have no time to learn things. And if they make mistakes they are fired. That forces people to use AI. 

The biggest problem with algorithms is they use existing information. They don't call any telescopes or particle accelerators to create experiments that they can produce new things. They just seek existing information and reorder it. If we start to use AI more and more to become effective we are facing situations in which we don't get new innovations. 

New innovations are things that make the world more advanced. But when we make innovation. We process thought into the product. Without thought is not innovation. Because our mind is to bind to memories we are in trouble with new things.  That means we should create a mind that is not dependent on memories. But that thing is not very easy to make. Almost everything that we make is bound to memories. 

Algorithms do not create things. They just connect existent data into new entirety. And the question is this: do we let ourselves turn into the AI's? Mechanical minds can search existing information but cannot create anything new. 

If we don't let people innovate that means that we will not create new things. Innovations require time and space. And that thing is not a thing that fits into the idea of productivity and effectiveness. 

Algorithms are tools that make many things faster than humans. That is the reason why we use them. The question is always this: do we someday make algorithms that rule people? When we make things by using algorithms we can find an easy way to make things. Algorithms make our lives easier. They search for things from databases. Then, AI can make many things better than we can. We are afraid. That AI helps people to make things. 

Like fake doctoral theses. We are always afraid of that. People lose their creativity when they use AI. We ever ask why we leave our work to the AI. We see that things happen, never asking "why"? 

The problem with the algorithm is that AIs don't make anything new.  They just collect information from the networks. AIs build new entireties using information. That already exists. And that is the problem with AI. The problem is that the system requires effective working. During working days people don't cheer programmers and other people to make new decisions. The requirement for effectiveness is the thing that destroys creativity and advancement. 

People are afraid of remote working because it destroys innovation. But how many hours in week do we use for innovations? Do we really make innovations in our offices? Or do we just read newspapers and reports and make things that have been made 1000 times before? 

Many people think. That innovation is the case, where one person talks and the other listens. The problem with innovation and creativity is that people are afraid of failure. Failure is somehow a non-accepted thing, and that means we make many times things that we made many times before. 

Because. People are afraid of things like grammar mistakes they think that they must let the AI make their texts. Programmers are afraid of mistakes.  Fixing mistakes steals their and other worker's working time. That is the reason why the working environment forces people to use AI. And that is the best way to destroy productive thinking and innovation. 

The AI makes people look like effective. But that system uses only existing information. That is the limit of the AI. Even if we want to make images using AI, we must give orders to it. The AI is like the drawing program that follows orders that we give. That allows us to create new worlds. But that thing doesn't advance our painting skills. 

The memory makes humans effective. Imagination is the thing that connects memory bites together new way. That allows us to create new things. Memories and the connections with existing memories are the things that limit the imagination. The requirement for innovation is an idea, that doesn't come from emptiness. 

Memories are stored in our memory cells. When we use imagination we just reorder those things that we can call memory pixels into the new entirety. Those things are like puzzles that we can reorder. 

The big advance would be the mind. That is not connected with memory. The memory-independent mind is something that we cannot imagine. We use memories to create new entities. And it's hard to even imagine a tool that doesn't use memories. That thing is the tool that can make the next step for innovations and society.