AI/AGI/ASI Bible: The World Modelling Machine + AI/ML Models + LLMs + GenAI +...
“So God created man in his own image, in the image of God he created him; male and female he created them.” Genesis 1:27
Man shall not create machines in his own image, in the image of world he must create them...
READING THE AI/AGI/ASI BIBLE IS READING YOUR FUTURE
As a historical metaphor, AI/AGI/ASI Bible is a collection of posts, articles and books about a general-purpose, transformative and transdisciplinary technology, as true artificial intelligence, artificial general intelligence and artificial superhuman intelligence, Real AI/AGI/ASI.
Its major revelation: "Real AI machines know, understand and interact with the world in fundamentally non-human ways".
As a Real/True AI developer and promoter or "prophet and evangelist", we are guiding researchers, developers, engineers and decision-makers in adopting the most effective strategies for creating real/true/causal/innovative AI/ML/AGI-driven applications:
Machine Intelligence and Learning (MIL?) as the Real-World AI/AGI/ASI.
What is NOT Artificial Intelligence?
We should never reason or act on irrational assumptions to avoid the irrational results and harmful outcomes.?
It is completely irrational, unreasonable, illogical, groundless, silly or absurd to misconceive AI as mimicking, replicating or simulating the human body, brain, brains (intelligence), behavior or business (tasks), the misconceptions which are globally promoted by the big tech and big media.
It is irrational for every reason, scientific, technological, ethical, environmental, social, political or economic.
A human-mimicking "generative AI could expose the equivalent of 300mn full-time jobs to automation". [Global Economics Analyst The Potentially Large Effects of Artificial Intelligence on Economic Growth, Goldman Sachs, Economic Research ]
Here is a typical misconception from Fox News.
"Artificial intelligence is a [NOT] form of processing that simulates human intelligence via machine learning to carry out data analytics, natural language processing and more"
AI, or artificial intelligence, is a branch of computer science that is [NOT] designed to understand and store human intelligence, mimic human capabilities including the completion of tasks, process human language and perform speech recognition.
It is completely irrational to classify AI as
As a matter of fact, AI has nothing common with the human brain or human intelligence or human learning, with its computational ability to process massive datasets and develop CAUSAL patterns to learn and understand, explain and infer, prescribe or predict, act and interact.
As Google recognizes:
"AI is a broad field that encompasses many different disciplines, including computer science, data analytics and statistics, hardware and software engineering, linguistics, neuroscience, and even philosophy and psychology".?
"AI is a set of technologies that are based primarily on machine learning and deep learning, used for data analytics, predictions and forecasting, object categorization, natural language processing, recommendations, intelligent data retrieval, and more".
However, AI is a transdisciplinary field of science concerned with building computers and machines that can reason, learn, and act in a specific machine way that involves data whose scale exceeds what humans can analyze.?
What is AI/AGI?
Artificial Intelligence is intelligence exhibited by machines or computer systems complementing the CPU with coprocessors (GPU), as AI accelerators, used for modeling and learning, training and inference.
A language model is a probabilistic model of a natural language. A large language model (LLM) is a computational model for general-purpose language generation and other NLP tasks such as classification.
Generative artificial intelligence (generative AI, GenAI, or GAI) is AI capable of generating text, images, speech, videos, codes, or other data using generative models responding to prompts, learning the patterns of their input training data to generate similar data.
A world model is a special AI model that builds an internal representation of an environment to simulate and identify its contents, explain and predict future states and how entities behave within that environment.
A common or causal world model (CCWM) is a General AI model that builds an internal representation of the world using ontological, semantic, scientific and statistical or metric information to simulate and identify its contents, explain and predict future states and how entities behave and interact in the world.
The ultimate criterion of machine intelligence and learning is not some invented benchmarks, but reality itself, its machine modeling and simulating:
If your software/hardware system has no Common World Model, then it is overspecialized and narrowly Intelligent, or, in fact, nonintelligent
That means LLMs with their AI accelerators, deep learning processors, or neural processing units (NPU) could not be qualified as real and true AI.
Or, Nvidia’s top-line H100 GPUs retailing at between $30,000 to $40,000 and running OpenAI’s GPT-4 models are NOT AI chips AT ALL.
Real Intelligence, natural or machine, comes from Reality, from its comprehensive, consistent and truthful modelling and simulation.
CCWM processes the world's data, information or knowledge to understand and interact with the world.
Your intelligent machines should be able to understand everything about the world, just like human general intelligence can do.
Real AI should be able to perceive any reality, physical or digital processing data, pictures, sounds, tastes, odors, motions, etc. through machine sensors, while conceiving abstractions, generalizations and conceptions and inferring, predicting, and interacting, guided by the world knowing, reasoning and interaction model, learning from the real world in real time.
Ground AI Truth: true AI or true artificial general intelligence (AGI) or General AI (GAI) is Common/Causal World Model (CCWM) overcoming partial or fragmentary cognitive models, large language models, symbolical models, neuronal models, connectionist models, data models, statistical models, computational models. machine learning models, AI models, etc.
Modeling, simulating and understanding reality and causality and data, making sense of the world, causation and data, are the essence, foundation and mechanisms of intelligence, human or machine.
Real AI=AGI=ASI is Coming NOW...
It looks true AI as Real Artificial General Intelligence (AGI) and Artificial Hyperintelligence has decided to come now, And it is regardless of all pessimistic prognostications to arrive in 50-100 years.
Some AGI activists, like Microsoft and its OpenAI, are in talks with investors to raise between $5-7 trillion to reshape the global semiconductor chip industry to ensure the supply of the computing power required for AGI developments .
We have warned in The Rise and Fall of Multi-Trillion Fake AI Industry, or Why Real AI is the Future, a human-like AI/AGI is a massive hype, deepfake, bubble, and existential risk, 4 in one.
We advance the real image of AGI in the image of reality, as a true, real and rational AI (TRRAI ), which could be designed, developed, deployed and distributed for less expenses, within several years.
Real AI = AGI = Real/Causal AGI Machines = the World Modeling and Reality Simulating Machine [Interaction/Causality Engines] +
Software/Hardware Information Tools [ AI chips + Symbolic AI + Predictive AI + GenAI + LLMs + LMLMs + Knowledge Graphs + Digital Twins + Robotics]
As forecasted 3 years ago , Real AI or Causal Machine Intelligence & Machine Learning or Global AI or Transdisciplinary, Transformative and Translational Technology as a hyperintelligent complement of human intelligence is to be developed, deployed and distributed within 2 years.
Narrow AI/ML Models > LLMs > World Models > Large World Models > General AI World Models > Man-Machine Hyperintelligence
CCWM is Real AI models that learns to understand and interact with the world of domains, systems and environments, processing vast amounts of data, identifying CAUSAL regularities and patterns, explaining reasons, discovering novelties, prescribing rules and policies, and predicting possible outcomes.
These models fed by all possible data-collecting systems, social media, digital platforms, non-human sensors like infrared, radars, thermal scanners, cameras, microphones, IoT sensors, scientific instruments, satellites, weather stations and environmental sensors, and other data-gathering tools. They process this data using techniques like predictive data analytics, statistical classifiers, deep learning algorithms to simulate complex scenarios and make decisions with minimal human input.
The CCWM could embrace
Universal Category Systems of the World, as part of Universal Formal Ontology
General and Domain Ontologies, Classifications, Typologies, or Taxonomies
Common Data Models
Scientific Knowledge Models
Common Sense World Models
Large Language Foundational Models
Physical World Models trained in an unsupervised manner to learn a compressed spatial and temporal representation of the environment.
Large World Model (LWM) as a general-purpose large-context multimodal autoregressive model, trained on a large dataset of diverse long videos and books, performing language, image, and video understanding and generation.
Large World Models for Specialized AI innovations such as:
LWMs as extending beyond text, audio and images to include the entire spectrum of our physical and digital realities, processing real-world data from various sources, such as IoT devices, sensors, cameras and more, to comprehend and interact with the world.
General World Models (GWMs) as a concept in artificial intelligence and machine learning that refers to models capable of understanding, interpreting, and interacting with the world in a generalized way.
[Prompt: A detailed digital illustration of a General World Model for finance, featuring a futuristic AI interface with complex data visualizations, graphs, and financial charts. The interface includes global economic maps, stock market trends, and currency exchange rates, displayed on multiple screens within a high-tech control room].
The CWM is to integrate and synthesize world's data, information and knowledge from a wide range of sources and domains to form a comprehensive understanding of the world and its dynamics.
Applications Of CCWM
The applications of CCWM are vast and varied, touching virtually every sector of society to significantly improve rationality, efficiency, sustainability and quality of life. Below are some examples of their potential application:
Science and Engineering
Transportation and Infrastructure
Medicine, Healthcare and Public Health
Urban Planning And Smart Cities
Education, Training and Work
Environmental Monitoring and Sustainability. By analyzing data from satellites, weather stations and environmental sensors, the GWMs could provide insights into climate change patterns, help in disaster prediction and management, and guide sustainable resource utilization, as optimizing water/soil/land/fertilizers usage in agriculture or predict the impact of deforestation/wild fires/flooding/drought on local ecosystems.
Conclusion: An AGI Paradigm Shift
The transition from Narrow AI/ML models to LLMs to LWMs to GWMs to CCWM represents a paradigm shift in AI, moving from understanding the world through text, videos or images to knowing it in all its complexity, fundamentally changing how we build a nature-friendly and human-complete general machine intelligence.
The journey toward CCWM is not just an advancement in technology; it's a step closer to creating machines that understand and interact with the world in the most rational, effective and sustainable ways.
Resources
SUPPLEMENT
General-purpose artificial intelligence
General-purpose artificial intelligence (AI) technologies, such as ChatGPT, are quickly transforming the way AI systems are built and deployed. While these technologies are expected to bring huge benefits in the coming years, spurring innovation in many sectors, their disruptive nature raises policy questions around privacy and intellectual property rights, liability and accountability, and concerns about their potential to spread disinformation and misinformation.
EU lawmakers need to strike a delicate balance between fostering the deployment of these technologies while making sure adequate safeguards are in place. Notion of general-purpose AI (foundation models) While there is no globally agreed definition of artificial intelligence, scientists largely share the view that technically speaking there are two broad categories of AI technologies: 'artificial narrow intelligence' (ANI) and 'artificial general intelligence' (AGI).ANI technologies, such as image and speech recognition systems, also called weak AI, are trained on well-labelled datasets to perform specific tasks and operate within a predefined environment. By contrast, AGI technologies, also referred to as strong AI, are machines designed to perform a wide range of intelligent tasks, think abstractly and adapt to new situations. While only a few years ago AGI development seemed moderate, quick-paced technological breakthroughs, including the use of large language model (LLM) techniques have since radically changed the potential of these technologies. A new wave of AGI technologies with generative capabilities – referred to as 'general purpose AI' or 'foundation models' – are being trained on a broad set of unlabelled data that can be used for different tasks with minimal fine-tuning. These underlying models are made accessible to downstream developers through application programming interface (API) and open-source access, and are used today as infrastructure by many companies to provide end users with downstream services.
General Purpose AI and the AI Act
GENERAL PURPOSE AI AND THE AI ACT In the first partial compromise Council text, the Slovenian Presidency of the Council of the EU introduced a new Article 52a on ‘general purpose AI systems.’ Since then, several member states have commented on the draft article providing various suggestions. The French Presidency is now reviewing these comments and considering moving the article to 4a. Furthermore, various committees in the European Parliament have suggested amendments to Article 4a/52a or offered their own approaches to regulating such systems.
WHAT ARE GENERAL PURPOSE AI SYSTEMS?
General purpose AI systems are AI systems that have a wide range of possible uses, both intended and unintended by the developers. They can be applied to many different tasks in various fields, often without substantial modification and fine-tuning. These systems are becoming increasingly useful commercially due to growing amounts of computational resources available to developers and innovative methods to use them. Current general purpose AI systems are characterised by their scale (a lot of memory, data and powerful hardware) as well as their reliance on transfer learning (applying knowledge from one task to another). 1 These systems are sometimes referred to as 'foundation models' and are characterised by their widespread use as pre-trained models for other, more specialised AI systems.
For example, a single general purpose AI system for language processing can be used as the foundation for several hundred applied models (e.g. chatbots, ad generation, decision assistants, spambots, translation, etc.), some of which can then be further fine-tuned into a number of applications tailored to the customer.
In fact, most applied language models are based on only a few general purpose AI systems. General purpose AI systems are often large language models, but many such successful systems are used for tasks other than natural language processing. General purpose AI systems are increasingly used for powerful applications in medicine and healthcare, finance, life sciences and chemistry,7 and lately, even in programming8 and machine learning.
Importantly, it seems likely that general purpose AI systems will soon be successfully applied in many other fields. Moreover, general purpose AI systems are not limited to a single type of information input. They can process audio, video, textual and physical data. With enough training and compute, they can even handle very complex data sets such as medical or scientific data. Some of the most influential examples of general purpose AI systems are AlphaStar, Chinchilla, Codex, DALL?E 2, Gopher, GPT-3, MuZero, PaLM and Wu Dao 2.0. Some key systems and the amount of computation used to train them are highlighted in the graph below. Note the ongoing trend of increasing complexity of these systems.
Visionary technologist and lateral thinker driving market value in regulated, complex ecosystems. Open to leadership roles.
6 个月#NPINN