AI/AA= #TransAI = #DataUniverse + #MachineOntology + #ScientificComputing: #RealAI vs. #FakeAI
#AIAA , #TransAI , #TrueAI , #TruthAI , #RealAI , #AlternativeIntelligence , #AlienIntelligence , #AugmentedAI , #MachineOntology , #DataUniverse , #WorldData , #WorldModeling , #WorldModelComputing , #UniversalOntology , #MachineOntology , #HumanAI , #AI , #ML , #DL , #ANNs , #GenerativeAI , #LLM , #FakeAI , #FalseAI , #FakeML , #FakeDL , #FakeANN , #FakeLLMs , #DeepFakes #FakeAIChip #FakeAICompany ?
Humans are finitely smart and infinitely stupid, while AI is infinitely smart and finitely stupid, thus completing each other. Author
“I’m going to start something which I call TruthGPT or a maximum truth-seeking AI that tries to understand the nature of the universe. And I think this might be the best path to safety in the sense that an AI that cares about understanding the universe is unlikely to annihilate humans because we are an interesting part of the universe.” E. Musk
Truly intelligent AI vs. faux AI/ML/DL
Truly intelligent machines are driven by real or true AI, which is trademarked as as AI/AA and hashtagged as #AIAA . It is a ?transformative, transdisciplinary and translational artificial intelligence, machine learning, deep learning, artificial neural networks, NLP/NLU/NLG, data science and engineering (#TransAI or #Trans -AI or #MetaAI ), going beyond and across the knowledge, assumptions, thinking, data, capability, methods, paradigms, techniques, models and practices of science, technology and engineering.
Real AI models learn causal world models while fake AI models learn/memorize surface statistics or spurious correlations.
Trans-AI involves all the major modes of science:
Formulaically, these two contradictory types of AI could be stated in the following way.
Real AI Formula = Data Universe [World Information] + Machine Ontology [General World Model] + Scientific Computing & Engineering/Computational Science & Technology [Algorithms, models and simulations, mathematical, scientific, computational + computer hardware + computing infrastructure, parallel, grid, supercomputing, quantum computing, high-performance, the internet]
Fake AI Formula (LLMs) = massive amounts of training data [structured/formatted, PDF, HTML, or JSON, or unstructured/unformatted data sets of examples, ?text (words and numbers), images, video, or audio, "teaching the machine learning algorithms" or fitting the parameters (weights) of classifiers how to interpolate "make predictions"] + training algorithms [regression, classification, optimization, decision trees, supervised learning, unsupervised learning, reinforcement learning, neural networks] + massive computing power [specialized weak/narrow AI chips computing from TSMC, Nvidia, Google, AMD, Amazon, Microsoft, Cerebras, or IPU of Graphcore ]
Fake/Faux [not genuine, fake or false] Artificial Intelligence is the simulation/replication/mimicking/ of human brain/mind/cognition/intelligence/action/tasks by machines, especially computer systems.
Specific applications of FAI include expert systems, natural language processing, speech recognition, machine vision, machine learning, deep learning, ANNs, as typified by #deepfakegenerativeAI , run by #fakeAIneuralnetworks , which compute power and training data scales are growing just exponentially.
Fake AI cloud services.?All the leading cloud providers are rolling out their own branded #FakeAI as service offerings to streamline data prep, model development and application deployment.
Top examples include?#AWSFakeAI Services,?#GoogleFakeAI Cloud,?#MicrosoftFakeAI Azure AI platform,?#IBMFakeAI solutions,?#OracleFakeAI Cloud Infrastructure AI Services, #NvidiaFakeAIaaS .
Here is the #FakeAI Chip Landscape
#FakeAI 's ethical challenges include #MLbias , due to improperly trained algorithms and human bias;?misuse, due to deepfakes and phishing;?legal concerns, including AI libel and copyright issues;?ghost work , as data labelers, delivery drivers and content moderators, elimination of jobs; and?data privacy concerns, particularly in the social media, banking, healthcare and legal fields.
"Tech companies that have branded themselves “AI first” depend on heavily surveilled gig workers like data labelers, delivery drivers and content moderators. Startups are even hiring people to?impersonate AI systems ?like chatbots, due to the pressure by venture capitalists to incorporate so-called AI into their products. In fact, London-based venture capital firm MMC Ventures surveyed 2,830 AI startups in the EU and found that?40% of them ?didn’t use AI in a meaningful way.
Far from the sophisticated, sentient machines portrayed in media and pop culture, so-called AI systems are fueled by?millions ?of underpaid workers around the world, performing repetitive tasks under precarious labor conditions. And unlike the “AI researchers” paid six-figure salaries in Silicon Valley corporations, these exploited workers are often recruited out of impoverished populations and paid as little as?$1.46/hour ?after tax". [The Exploited Labor Behind Artificial Intelligence . Supporting transnational worker organizing should be at the center of the fight for “ethical AI.” ]
AI/AA as Sustainable [Universal] Intelligence
We all heard of alternative energy as clean energy from renewable energy sources replacing fossil fuels, coal, oil, natural gas and nuclear fuel: biomass [wood and wood waste, municipal solid waste, landfill gas and biogas, biofuels]; hydropower; geothermal; wind; solar.
Extending the idea of alternative energy, I innovate the concept of alternative intelligence (AI/AA), real or true AI, as overruling a human-mimicking, energy-hungry artificial intelligence in the forms of machine learning and deep learning or generative AI and large language models.
Such Real AI System is to govern critical [system] software systems, like as air traffic control systems, military control and command systems, national voting systems, telecommunication networks, power grid systems, transportation systems, supply chain management systems, industrial automation systems, smart web systems, cloud ML software platforms, online learning platforms, etc.
Currently, there are two categories of human-like AI: 'artificial narrow intelligence' (ANI) and 'artificial general intelligence' (AGI), marked with all kinds of severe limitations, from being nature-destructive, the biggest threat to humanity, to all sorts of risks and biases, as involving algorithms, models and training data. It is beside of or because of "counterfeiting the human brain or human intelligence or human behavior or human tasks ".
As a next-generation technological replacement, there emerging a Truth-seeking, scientific AI, which is universal learning machines?driven by a true, real, authentic AI sustainably interacting with the world while effectively operating with general and fundamental concepts (universals) and world's data (data universe).
Like as renewable energy flows involve natural phenomena such as?sunlight,?wind,?tides,?plant growth, and?geothermal heat, the AI/AA involves the world of reality, its modeling and simulating and effective and sustainable interacting, in terms of digital universals and data structures, advanced hardware and software systems.
Universal Ontology as the essence of intelligence and learning
Universal ontology is about coding and organizing everything in the world in terms of universals and properties, categories and relationships.
It provides a standard world model for humans, as metaphysical ontology, and computing machinery, as machine ontology, offering a uniformly organized human knowledge of the world as well as the world data reference computing framework. [Reality, Universal Ontology and Knowledge Systems: Toward the Intelligent World ]
Ontology, as the first science, is whenever you ask deepest questions about the nature of things, reality or universe or world or existence or being, its entities and relationships, as what-why-how is it all? what is really their? what is out their? What is existence? and What is the nature of existence??What is AI? and What is the nature of AI??
It is an ontological inquiry, which is about the nature of existence, about a mental or computing encoding of reality, what is making humans and computers really intelligent.
Ontology is the science of sciences, a comprehensive study of reality, being, existence, or the world as a whole. Its universe of discourse, or universe is a collection that contains all the entities in the world.
It studies the common features and relationships of real entities, ontological, physical, biological, mental, social, informational or digital entities.
There are theoretical and applied ontology. The former covers all sciences, from physics to the humanities.
It creates the most fundamental theories of the universe, as
As Metaphysics, it started Philosophy, as Natural Philosophy, modern natural sciences.
In AI, Computer Science and Data Science, it goes as an ontology, the specification of conceptualizations, basing on
There are top-level, domain, task and application ontologies.
Thus, ontology is to organize world's knowledge, transforming data into information, knowledge or intelligence.
It is mistaken as vocabulary, knowledge representation language, taxonomy, etc.
The purpose of ontology is to update itself to data ontology, thus organizing world's information as a machine-readable content. Ontology engineering, learning and encoding are hot topics today for ML/DL to enable sharing and reuse of domain knowledge for general learning transfer.
In all, universal and global ontology is like a master theory of everything, a universal knowledge and reasoning framework integrating both top-level ontologies and domain ontologies.
Universals or concepts or abstractions are the essence of intelligence, human or machine
It is essential to know that universals are as real as the universe itself, but not individual, existing as the world of reality, the totality of entities and interactions, in all possible forms and kinds, scopes and scales.
The concrete-abstract distinction or the universal-particular opposition is the greatest bias and prejudice dating back to the ancient?Greek philosophy.
It is commonly presumed that particulars are concrete, universals are abstract, while particulars could be abstract and universals are concrete.
Philosophy capitalizes on such biases and prejudices, dividing itself in different parties and ideologies, as realisms, extreme?and moderate, idealisms, extreme and moderate, or nominalism?and materialism.
The conception?of universals varies from metaphysics to culture:
All who think that universals are just common or abstract names or concepts in the mind, need to think twice. Democracy, Equality, Peace or War, Humanity and Morality are supposed to be real universals. Otherwise, we are all "fake idea?makers".
Nothing is more real than universals,?as types or classes or categories, properties or states or conditions, changes or actions, relations or interactions, which are applied to everything, to all instances and cases, in the world or in some domains.?
Bottom line. There is one emerging, general-purpose technology, which is both universal, all-purpose, all-round and fundamental, basic and foundational, which is AI/AA Technology.
All-Purpose Universal Intelligence vs. General-Purpose AI
We have to differ a human-like, human-level AI, dubbed as AGI, strong AI or full AI, from the real and true or general AI.
The first one is a techno-fiction which is neither possible as a mixture of artificial narrow techniques, as machine learning, DNNs, NLP/NLG, computer vision, etc., nor as a human-like and human-level AI.
It is a fake AGI, going as a mass market fraud.
Real and True AGI is NOT the intelligence of a machine or computer that enables it to imitate or mimic human capabilities.
Real AI is the intelligence of a machine or computer that enables it to model and simulate and understand the world of reality, including causality and mentality, data universe and computation.
Driven by the world model engine, Real AI systems effectively interact with environments, navigating the world, recognizing objects, learning from experience, discovering regularities, communicating, solving complex problems, predicting, decision-making and acting upon the data, rules and algorithms and encoded scientific models of reality and mentality.
AI/AA as a Synthesized Man-Machine Intelligence and Learning (MMIL) could be one of the greatest strategic innovations in all human history. It is fast emerging as an integrating general purpose technology (GPT) embracing all the traditional GPTs, as electricity, computing, and the internet/WWW, as well as the emerging technologies, Big Data, Cloud and Edge Computing, ML, DL, Robotics, Smart Automation, the Internet of Things, biometrics, AR (augmented reality)/VR (virtual reality), blockchain, NLP (natural language processing), quantum computing, 5-6G, bio-, neuro-, nano-, cognitive and social networks technologies.
But today's narrow, weak and automated AI of Machine Learning and Deep Learning, as implementing human brains/mind/intelligence in machines that sense, understand, think, learn, and behave like humans, is an existential threat to the human race by its anthropic conception and technology, strategy and policy.
领英推荐
The All-Purpose, Universal, True or Real AI is to merge Artificial Intelligence (Weak AI, General AI, Strong AI and ASI) and Machine Learning (Supervised learning, Unsupervised learning, Reinforcement learning or Lifelong learning) as the most disruptive technologies for creating real-world man-machine super-intelligent systems.
Again, AI must be a real and true and autonomous, self-learning and scientific technology, not a human-like and human-level artificial AI, or fake AGI.
AI/AA as an All-Knowing, Omniscient Man-Machine Intelligence
AI/AA is an omniscient man-machine intelligence, which should not be mixed with a general-purpose or generative AI.
As mentioned, there are two categories of human-like AI: 'artificial narrow intelligence' (ANI) and 'artificial general intelligence' (AGI) vs. a true, real or authentic universal AI.
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.
General-purpose AI tools are now reaching the general public. In March 2023, Microsoft launched a new AI-powered Bing search engine and Edge browser incorporating a chat function that brings more context to search results. It also released a GPT-4 platform, giving businesses and governments the ability to generate text, images, code, videos, audio, and to build their own applications. Developers are using these 'foundation models 'to roll out and offer a flurry of new AI services to end users. General-purpose AI tools have the potential to transform many areas, for example by creating new search engine architectures or personalised therapy bots, or assisting developers in their programming tasks. According to a Gartner study, investments in generative AI solutions are now worth over US$1.7billion. [General-purpose artificial intelligence, European Parliament ]
All-know AI is an universal/encyclopedic AI of machine intelligence and learning (MIL), which is universally contextual due to an encoded universal knowledge of philosophy and science, technology and engineering.
Three ways to build the omniscient MIL:
Empirical, inductive or statistic, from bottom-up, as a multi-set of narrow and weak AI/ML/DL technologies, techniques, methods, algorithms, models, applications, platform, like Google AI Platform or Microsoft Bing/ChatGPT Platform;
Theoretical, deductive or ontological, from top-bottom, a system of narrow and weak AI/ML/DL technologies, techniques, methods, algorithms, models and applications, framed by a general intelligence and knowledge;
Scientific, inductive-deductive, combining data and generalizations, theories and statistics, a global AI driven with "the theory of reality and mind", world modelling and master algorithms, and operationalized as general, narrow and weak AI/ML/DL technologies, techniques, methods, algorithms, models, applications, platforms.
Thus, AI’s omniscience refers to MIL’s state of knowing everything, complete knowledge of all things, actual and potential, as mostly contained in human encyclopedias.
The Future of AI is RealAI/AA, #AIAA
The future AI is a real, true, genuine, authentic or general purpose AI, which is following the development path:
AI/AA = Symbolic AI >
ML > DL > ANI >
Artificial General Intelligence (AGI) >
Artificial Super Intelligence (ASI) >
Real AI, Causal AI, Man-Machine Superintelligence and Learning >
Trans-AI, Meta-AI
Narrow AI is too specialized and focused only on task-specific commercial algorithms, problems and applications in specialized areas, instead of integrating all technologies, techniques and models:
Artificial Narrow Intelligence
Experts systems, RPA
Machine Learning
Artificial Neural Networks
Deep Learning
Machine Perception/Vision
NLP
NLG
NLU
Robotics
Automation
The Internet of Things
As a result, its best exponents, as ChatGPT and large language models, mix good insights with utter stupidity (idiotic mistakes) and/or toxic content.
ChatGPT and LLMs are dumb and dull and deficient while having access to a petabyte amount of data and verbal information available on the Internet, including Wikipedia, Web sites, social media networks, and electronic books.
They merely predicts the next token/thing without any causal model of how the world actually works. ChatGPT & LLMs have sophisticated syntax with no semantic connection with reality, making them incapable of explaining what/why/how things happen.
Such narrow or weak or fake AI models in the forms of machine learning programs and deep learning algorithms are simple to create. You need to identify your special problem or specific task, such as object/text/speech/facial recognition, language translation, or playing board games, gather relevant data, choose the right tools, develop narrow AI models, train and evaluate them, and deploy.
Again, “Modeling techniques that today power many AI applications, such as deep learning and neural networks, are inherently more difficult for humans to understand. For all the predictive insights AI can deliver, advanced machine learning engines often remain a black box.” [Why businesses need explainable AI—and how to deliver it]
Conclusion
AI/AA is the summit of all human evolution with human knowledge having a revolutionary history:
Homo Sapiens Sapiens >
Mythology > Religion > Philosophy >
Ontology > Logic > Epistemology > Semantics >
Mathematics > Physics >
Science & Technology >
Computing Machines > the Internet/WWW >
Emerging Technologies (Robotics, Quantum, Genetic, Bio-, Neuro-, Nano-, Cognitive and Social Engineering) >
NAI/ML/DL >
Causal AL/ML/DL >
Human Intelligence > BMI/MMI > Digital Reality/Cyberspace >
Digital Superintelligence >
TransAI = MetaAI = TrueAI = Real AI = Human-Machine Superintelligence > Machina Sapiens >
I-World = Smart and Sustainable World