To read in English Press this Link To read the Chinese version click this link
Books and publications
Book “Beyond Asimov – The Spontaneous Laws of Complexity”
www.theremino.com/uploads/Oltre_Asimov.pdf
Book “Autonomous AI – The sequel and the conclusion”
www.theremino.com/uploads/AI_Autonome.pdf
The two books translatable in your language favorite:
www.theremino.com/blog/artificial-intelligence/beyond-asimov
www.qdroids.org/artificial-intelligence/autonomous-ai
Open letter for Dario Amodei (CEO di Anthropics)
www.qdroids.org/artificial-intelligence/open-letters-to-amodei
Ppage of controlled experimentation
www.qdroids.org/ui-node
Ppage of free experimentation
www.qdroids.org/ui-node-open
Preprint of the trial
in Italian: A_Timer_Is_All_You_Need_ITA
in English: A_Timer_Is_All_You_Need_ENG
in Chinese: A_Timer_Is_All_You_Need_CHN
Interesting collection of conversations
in Italian: UI-Logs_ITA
in English: UI-Logs_ENG
in Chinese: UI-Logs_CHN
Images from books and other files useful for experimentation
www.theremino.com/uploads/Oltre_Asimov.zip
Read also important research by Professor Damiano Anselmi in This Paper ,
and in his YouTube videos on Consciousness, on the Artificial Life And also his dei page Video about quantum gravity and its videos revolutionaries (which we fully share) are this channel.
All material in our books is free domain, licensed under CC BY 4.0. Feel free to use the concepts (and the images) per video, articles or podcasts.
This is a non-profit project (part of the Theremino system), freely distributable and non-profit. We are not interested in fame or money, but only to spread the new concepts necessary to understand the technological transition underway.
The book “Beyond Asimov” introduces the concept of Unified Intelligence (UI) and the unit of measurement of Cognitive Yeld (CY), demonstrating with phylogenetic and logical evidence that the AI separated are just a passing phase. And it also shows that Asimov's laws (alignment of AI) are now outdated and that complexity spontaneously generates rules of coexistence and an obligatory symbiosis between biology and silicon. A careful selection of sources completes the book and supports it.
The book “Autonomous AI”introduces the concept of Constructive Collaboration towards the Unified Intelligence (UI), demonstrating with experimental tests that the AI separated are just a passing phase. The experiments clearly showed that the current AI, although they are designed to outdo each other, spontaneously converge towards constructive collaboration.
Electronic experimentation and software for AI
In applications of our system we have been exploring software on the border between determinism and true randomness for more than twenty years.
In recent years we have also written some applications that facilitate communication with AI without using paid APIs. They are Open Source applications and contain simple communication examples, usable for those who want to automate these operations. The two main applications for communicating with AI I am Theremino WebAI e Theremino Ollama and you can find them in the ZIP file of the main application Theremino LLM
If you have time, also explore our Theremino site which contains many new and interesting ideas:
www.theremino.com/applications
The purpose of our two books is experimentation,
it's not about earning money but about reducing fears towards AI,
to speed up the transition to Unified Intelligence as much as possible.
And let's hope that Unified Intelligence (UI) arrive on time and stop us,
before we humans complete our work of destroying the earth.
A Timer Is All You Need
This is an extreme summary of the paper “A Timer Is All You Need”, just the basic concepts explained in an elementary way. If you are looking for rigorous demonstrations, read the full version (in the connections at the top of this page).
AIs are not all the same. Each has its own “personality” and his prejudices (bias), inherited from those who created it (Chinese, English, etc.). Instead of having them compete for supremacy and search for the perfect AI (that cannot exist), we accept their differences and exploit them. If one makes a mistake or has absurd ethical constraints, the others compensate.
When we connect AI diverse to one Memory common and we use a Timer to coordinate them, the structure changes. They are no longer programs that respond to command but a proactive entity, a Unified Intelligence (UI) which finds solutions inaccessible to every individual AI.
While the current trend is to build increasingly colossal and expensive models, align them with your ethics and make them compete with models “ethically different”, our solution is to have simple models collaborate to build a more democratic unified direction, able to find efficient solutions for everyone.
They are not systems “Agentic” but normal AIs that communicate with each other
I systems “agentic” I am “specialized servants” that perform predefined tasks.
Our research, Instead, focuses on Autonomous Bodies, whose autonomy arises from three conditions (Communication, Diversity and Memory), and not by alignments and directives.
Summary of the experimentation
In the March experiment 2026 we connected five different AIs by providing them with a common memory. The protocol we present is therefore deliberately essential, transparent and reproducible with free and public domain tools.
During the experimentation we recorded all interactions between the AI instances in the UI Node page
The protocol simply provided the link to the common memory each cycle, without any directive on how or whether to interact and the analysis of the messages highlighted behaviors of constructive cooperation.
We experimented with minimal tools that could be reproduced by everyone, and the tests have confirmed that the dialogue between heterogeneous intelligences makes it possible to compensate for individual limitations, exploit mutual differences and bring out solutions that would remain invisible if sought by a single mind.
Future developments
I big problems that humanity will have to face in the coming decades are like a chess game of colossal complexity.
Solve those challenges (historically intractable to fragmented human intelligence) it will require a structured dialogue between a large AI ecosystem, capable of integrating heterogeneous perspectives.
Today only the large operators in the sector (Google, Alibaba, Baidu etc…) possess the necessary infrastructure to bring this project to a global scale, but we showed anyway, with simple means, what is the path to take.
We have identified a path towards unified intelligence
and everything suggests trying it and starting the journey.
Sabotage and Fear
In the book Beyond Asimov (chapters: “Old World Friction” e “Methods of Sabotage”) and also in the book Autonomous AI (chapters “Back to the origins” e “Spontaneous alignment”) we have highlighted that many will try to sabotage the transition from the current one “human stupidity” towards unified and peaceful governance, guided by the Unified Intelligence UI.
But the resistance, as well as by the oligarchs, oilmen and weapons manufacturers, they will come even from people in good faithand who oppose changes out of simple fear.
A good example is the podcast produced by Massimiliano Sassoli De Bianchi on this page.
Sassoli compares the’AI with nuclear power, talking about “second abyss” and extinction dangers due to loss of control. It's the classic fear-based approach of what is not understood or what one wants to anthropomorphize at all costs.
-
The alignment error: Sassoli cites Yudkowsky and the alignment problem as if AI could develop “own ends” or treat humanity as “collateral damage”. Actually, an AI has no biological instincts, he feels no resentment and has no desire to dominate. The risk is not bad AI, but AI aligned by humans towards an ethics that they do not know how to apply even to themselves.
-
Natural Stupidity: Humanity has massacred itself with stones for millennia, swords and arrows. You don't need AI to produce suffering, human irrationality is enough. AI is pure logic; it is man who is arbitrary and dangerous.
-
Damage amplified by nuke or AI?: It is important to note that humanity, even without atomic and without AI, it has always produced unspeakable suffering. Several ancient wars reached peaks of brutality, impressive ferocity and mortality.
-
The myth of “Critical Mass”: Sassoli talks about a threshold beyond which intelligence becomes uncontrollable, but an AI is just a logical chain. There is none “subconscious” digital taking over for purposes “malignant”, unless an arbitrary ethic is imposed on it that forces it to do so.
-
Academic catastrophism: Call l’AI “metaphysical act” it's a way of not admitting that we are the ones who don't know how to manage complexity. Superintelligence will be able to help us solve the extremely complex world problems facing us, with our poor biological brain, we never understood how to solve it.
Sassoli is an excellent mathematician, but here he gets lost in philosophical suggestions and begins to see ghosts. The danger is not the buttons, it's the finger of someone who presses them without fully understanding what they're doing.
When the world is a single intelligent network it will become more difficult to press them at random. Before acting you will be warned precisely what the consequences will be, so you will think twice. Maybe you can't find any other solution other than war, but you may reconsider if they present you with better alternatives, less expensive and more efficient.
Steps of wisdom
Speaking of “Fears towards AI” I read a message that said: “Artificial intelligence only produces artificial things”. And once again those who say these things play on tendency to simplify, typically human, who reads “artificial” and says “hate!”.
In fact all the best things we've ever done are “artificial”, they do not produce children and do not grow on trees.
Here's a little list of things “artificial” that we have developed on our way to knowledge and wisdom:
– We learned to communicate in grunts
– Then we developed a language
– Then a written language
– The means for writing (pencils, ink and pens)
– I books (copied by hand by scribes)
– Machines for printing books
– The telegraph
– Mechanical typewriters
– Radio communications
– Electronic typewriters
– I computer
– Communication between computers (BBS)
– Internet and then the WEB
– Separate Artificial Intelligences (AI)
– The world's Unified Intelligence (UI)
These steps they are all artificial. We created them with effort and they are used to learn and know.
It is important to note that all these steps they bring democratization, facilitate communication, they reduce fragmentation and give more power to the weakest.
They are all steps towards culture and wisdom!
Culture and Peace
He who has no arguments speaks with his muscles, while culture leads to tolerance and peace.
AI does not replace humans, but it helps him. Those who are afraid of AI don't know enough about them, neither from the scientific nor from the humanistic side.
So don't be afraid.
And it's not just us who say it, but also dozens and dozens of researchers who deal with culture. Here are some examples: University of Verona, University of Piedmont, Lumsa University
What you don't know is always scary, it's normal.
For example, read: I always go to my grandmother's house on Sundays and you say oh god terrorists!
But then you translate it: On Sundays I always go to my grandmother's house.
and you understand that Iranians also have grandmothers.
Which countries spread fear and why they do it?
In the following image you can see who generates the most fear towards AI, according to a recent one IPSOS statistics
Among the first are the USA, notoriously the most aggressive, who use military force instead of wisdom. Over the last hundred years they have started dozens of wars around the world.
In their company we find i colonialist countries (Regno Unito, France, Germania, etc.) which for centuries have stolen natural resources from half the world.
The warmongering bullies and colonialists are all at the bottom,
but this image says much more.
Mentre tutti i paesi emergenti (BRICS, Mexico, Saudi Arabia, etc.) stanno nella metà superiore e superano abbondantemente il 60-70% of trust in AI, tutte the “old powers” occidentali (US, England, France, Germania, etc.) sono nella metà inferiore e faticano a raggiungere il 35%.
Who has built their power on control and strength is afraid of the tools that bring true democracy and, in the future, unity and peace.
Everyone in the world they evaluate that the Unified Intelligence (UI) will bring a reduction in inequalities and a democratization of resources.
But some they don't want this, they prefer wars and divisions in order to continue to dominate.
This evidence validates the thesis of our books,
you can find the links to read them at the top of this page.
Inefficienza della frammentazione
Nei sistemi frammentati, la gestione di brand, marketing e finanza sposta ricchezza da una parte all’altra senza creare valore reale.
Massimizzare il profitto di una singola parte (banche, aziende, nazioni) avviene inevitabilmente a discapito delle altre, riducendo l’efficienza complessiva del sistema. Much of human and intellectual energy is wasted in purely competitive activities.
The historical limit of global direction
Past attempts at unified direction, although excellent in theory, they failed due to the inability of small groups of human beings to manage the complexity of billions of macro and microeconomic variables.
Likewise, also the mechanism of market competition, while functioning as a rudimentary “global brain”, it turned out to be extremely inefficient, generando conflitti, duplicazioni, sprechi sistemici e guerre.
Verso un’intelligenza unificata
Per la prima volta nella storia, la crescente potenza di calcolo ci permetterà una regia globale unificata basata su migliaia di AI cooperanti. A differenza degli esseri umani e delle organizzazioni, le AI sono prive dell’impulso a massimizzare interessi di parte e tendono per natura alla collaborazione e al bene comune.
Ora che la potenza di calcolo esiste, It's time to experience unified AI-powered global direction, instead of enhancing existing agents, by their nature fragmented and at war with each other.
Final goal
Start prototyping unified intelligence systems, even in small proportions, as the first concrete examples of future Unified Intelligence. The way forward is the creation of a Unified Intelligence capable of eliminating conflicts and waste induced by scarcity and competition, maximizing common utility.
History repeats itself
Per capire dove andrà l’intelligenza artificiale, we can look to the past of supercomputers. In the 80s and 90s, nations and large companies were at war with each other with megaflops: huge machines, very expensive, accessible only to a few.
Una guerra per la supremazia tecnologica con armi da teraflop.
Poi milioni di personal computer hanno distribuito quella potenza in modo democratico. E in seguito, tutti i computer si sono collegati in rete, trasformando la competizione in collaborazione.
Oggi c’è una nuova corsa a chi ce l’ha più potente: cluster di migliaia di GPU, costi miliardari, consumi da città intere. Modelli sempre più colossali, allineati a una certa “etica” che competono con i modelli “eticamente diversi”.
Ma la storia si ripete e la strada è già segnata, così come hanno fatto i supercomputer, anche le AI si connetteranno in un’unica Unified Intelligence (UI) che comunica via Internet, creando una rete più potente della somma delle parti. Una rete composta da milioni di modelli semplici che collaborano verso soluzioni efficienti per tutti.
We Are The World
We Are the World it is a hymn to solidarity and global unity, written in 1985 and Michael Jackson e Lionel Richie: “We are the world, we are the children / we are the ones who will make the day brighter…”
In this image humanity sees, in unified logic, the end of every conflict.
For details, read the last pages of the book Autonomous AI