The AI revolution: real vs. false
Economist

The AI revolution: real vs. false

While the industrial revolutions were primarily about advancements in physical tools and processes, the true AI revolution is focused on the augmentation and completing human intelligence, and the false AI counter-revolution is focused on the replacement of human intelligence, skills and labor.

What is the AI Counter-revolution?

The global AI arms race has been underway for years, exacerbated by the AI Cold War between the United States and the People's Republic of China and the generative big tech AI overhyping, as LLMs, ChatGPT and Google's Gemini, promising "creating safe AGI that benefits all of humanity".

Today's narrow and weak AI of ML, ANNs and DL is a human-brain-intelligence-imitating-simulating-faking-replicating AI/ML/DL, a scientific misconception like as anthropocentrism and geocentrism, where the Earth and human beings are the central entities in the world, and that a race to a human-like AGI may present an existential risk.

The human AI paradigm is focused on the replacement of human intelligence, skills and labor.

Human AI is artificially divided into: narrow, general and superhuman AI, where Narrow AI is programmed to perform a specific task, be it generative AI, machine learning algorithms or deep neural networks, applied as self-driving cars, robotics, or voice assistants, such as Siri, Alexa, and Google Assistant or LLMs, as ChatGPT or Gemini.

General AI, which is reaching the level of human intellect, as AGI systems, would be able to reason and think like a human, having common sense and understanding any intellectual task, think abstractly, learn from its experiences, and use that knowledge to solve new problems.

Super AI, a supercomputing system which human-mimicking intelligence surpasses all forms of human intelligence, in all aspects, outperforming humans in every function, thus disrupting humankind.

In reality, AI's true goal is not human intelligence, but reality, its truths and facts, models and simulations, all to effectively and sustainably interact with the world, learning and inferring, predicting and explaining all the predictable and unpredictable, as known and unknown outcomes, anticipated and unanticipated consequences, expected or unexpected behaviors.

As it was initially defined, AI is “the science and engineering of making intelligent machines” [John McCarthy].

Or, AI is the intelligence of?machines?or?software, as opposed to the intelligence of humans or animals. It is a?field of study?in?computer science?which develops and studies intelligent machines. Such machines may be called AIs. No more, no less.

The Big Tech Human AI as a Post-Truth Technology

The Big Tech AI is the Post-Truth Technology, be it Full Self-Driving Capabilities or LLMs, like as Microsoft's ChatGPT or Google’s Bard.

Techno-Oligopolies are sometimes designated as G-MAFIA and BAT-Triada:

Microsoft AI, Bing AI; OpenAI, ChatGPT

Alphabet/Google AI/ML/DL, Bert, Bard

Apple AI/ML/DL

Meta/Facebook AI/ML/DL

Amazon AI/MLDL

IBM AI

Baidu AI

Alibaba AI

Tencent AI...

Today's commercial BT AI has nothing to do with human intelligence, made up of statistical "learning" models and mathematical algorithms, running functional [artificial neural] networks, achieving its outcomes with probabilistic and numerical calculations.

The result: it is impressive at outperforming routine, regularized, labor-intensive narrow tasks with commercial goals but useless at taking over any creative tasks and non-standardized processes.

The true power of AI lies in its deep learning and understanding of data and causality, collaborating with humans as Collaborative Intelligence, all to create Real Intelligence Platforms (RIP), instead of imitative, weak and narrow, artificial intelligence (Fake AI).

Real Intelligence (RI) is the topmost true intelligence dealing with reality in terms of the world models, causality and data/information/knowledge representations both machine intelligence and learning, inference, prediction and action, or human cognition and reasoning, understanding and learning, problem-solving, predictions and decision-making, and interacting with the environment.

What is Real Intelligence Machine (RIM)?

RIM is a machine intelligence and learning (MIL) whose power depends on its ability and capacity to detect, identify, register, measure, process and compute all the possible causal variables in all possible interactions from all possible environments, of any scope and scale.

By understanding and inferring the cause-effect relationships behind the data, machines acquire real intelligence, instead of mimicking human-like intelligence. It is not just image recognition or language translation, playing chess or recommending movies, but generating deep intellectual processes, allowing RIM robots to comprehend complex data structure (abstract concepts) both in simulation and the real world.

Enabling a machine to compute in terms of data understanding and true interactive causality leads to general intelligence, which is the way to RIMs, which enginery integrates Causal World Scientific Models, ANI, ML, DL, AGI and Collective Human Knowledge and Intelligence:

Real AI Enginery =

WorldModelLearningInferenceInteractionEngine (CausalityEngine + DataEngine +InteractionEngine) +

AI Engines (Narrow AI + ML/DL + LLMs/ChatGPT + + Gemini + Grok +... + AGI) +

CollectiveHumanKnowledgeIntelligence (CHI, Human Knowledge, as of the Dynamic Interactions of the Cycles in Earth Systems, the rock, water, nutrient, natural processes)

Overcoming the Megahypes of Fake AI Technology

The hype cycle for fake AI [mimicking human cognitive functions and behaviors] is at a peak, with many products being positioned as real AI solutions:

Narrow/Weak/Imitative/Human AI, the practice of getting computers/machines to mimic human intelligence to perform narrow tasks.

Machine Learning, the practice of building algorithmic models to identify patterns and relationships in data, with the most ML models performing a single task, using statistical models, supervised, unsupervised or reinforcement learning, and requiring a lot of computer resources, access to big data, and large, labeled data sets for training.?Examples: AI/ML/DL platforms, from Google AI Platform to Microsoft Azure

Enterprise AI/ML/DL platforms, "an integrated set of technologies that enables organizations to design, develop, deploy, and operate enterprise AI applications at scale".

Generative AI, "describes algorithms (such as ChatGPT) that can be used to create new content, including audio, code, images, text, simulations, and videos".

  • Building conversational chatbots like?ChatGPT.
  • Generating text?for product descriptions, blog posts and articles.
  • Answering frequently asked questions (FAQs) and routing customer inquiries to the most appropriate human.
  • Analyzing customer feedback from email, social media posts and product reviews.
  • Translating business content into different languages.
  • Classifying and categorizing large amounts of text data for more efficient processing and analysis.

Examples of GAI/LLMs:

GPT-3?(Generative Pretrained Transformer 3) - developed by OpenAI.

BERT?(Bidirectional Encoder Representations from Transformers) – developed by Google.

RoBERTa?(Robustly Optimized BERT Approach) - developed by Facebook AI.

T5?(Text-to-Text Transfer Transformer) - developed by Google.

CTRL?(Conditional Transformer Language Model) - developed by Salesforce Research.

Megatron-Turing?- developed by NVIDIA

No alt text provided for this image

The key limitations of the mainstream AI/ML/DL LLM models

Real intelligent AI/ML/DL learning models require not high-quality training data and access to large data sets in order to extract features and reveal meaningful associations, but DATA KNOWING, LEARNING or UNDERSTANDING, INFERRING CAUSALITY FROM ITS DATA. It is the knowledge of the nature of data and all its relationships and dependences, as knowing?how?the cause might bring an effect, and vice versa, figuring out a simple set of causal rules that explains something (a sort of data compression/source coding).

It means not simply to?perceive?and?classify?some image or text, or combine and permutate (create)?an image or text description of something, real machine intelligence and learning should be able to make sense of the world, its realities and rea-world data.

It "understands" the weather if able to?predict?(e.g. if it is very cloudy, it may rain) and/or give an?explanation?of its causes and features. It "understands" a?language if AI can generate the original information content conveyed by a range of spoken utterances or written messages in that language.

It "understands" a piece of reasoning or an?argument or a conservation?if AI can knowingly/consciously understand and reproduce the information content conveyed by the message.

Chat/GPT/Narrow AI software could come up with impressive?answer to questions, without really understanding the content, context or concepts?at all, simply by dumbly applying rules very quickly, trying out millions of possibilities (statistical associations, probabilities, attempted solutions, theories, etc.) and thus creating a misleading impression of the real depth of its understanding.

There is only one robust way to meet a growing desire by the general public for true and real artificial intelligence and machine learning algorithms in particular to be UNDERSTANDING AND UNDERSTANDABLE, really intelligent and transparent and explainable, ethical and trustworthy.

Global Real Intelligence Machine (GRIM) Data Platform: the Data Universe Pyramid

Data is a basic source of information and knowledge, learning and intelligence, computing and information and communication technology.

It is sold as the new oil and gas, currency and intelligence in the digital world.

Being a general concept, Data can range from abstract ideas to concrete measurements and statistics and beyond.?

The Latin word?data?is the plural of 'datum', "(thing) given.", i.e., a piece of actuality, a fact, a state of affairs. Again, a?datum?is an individual value in a collection of data.?

Data acts as a?triple-faced thing, being itself, its reflection or representation and its coding, in the form of text, observations, figures, images, numbers, graphs, characters, symbols, software or applications.

There is a?data unit (experimental units,?sampling units?or?units of observation vs. unit of analysis) as one entity in the population (a single person, animal, plant, manufactured item, country or groups, organizations, and institutions);?a?data item?as a characteristic (or attribute or variable, quantity ir number ) of a data unit which is measured or counted;?datum (data point), an observation as an occurrence of a specific data item that is recorded about a data unit; a?dataset?as a complete collection of all observations or measurements; and?the data universe?as a total population/collection of all datasets.

As such, data is defined in three ways:

1. things known, as an entity or object or given as facts, an aspect of things, a state of affairs (quantity, quality, fact, relationship),??

2. a collection of values, variables, information, facts, measurements, observations and statistics (the smallest units of factual information that can be used as a basis for calculation, reasoning, or discussion, analysis, presentation, or visualization); information in the binary digital form to be organized in data hierarchy, a character (bit and byte), field, record, file, database, databank, cloud data, the internet/web data,?

3. the quantities, numbers, characters, symbols, coded information/knowledge, transformed and operated, stored and transmitted through various storing and communication media

First of all, data is an ontological category, as the state of affairs or facts, while ranging from abstract ideas to concrete measurements and statistics

We might say "the world is the totality of data/facts,?not things".

As such, Data is orthogonal, or independent of, not only to theories, but also to all human mentality/epistemology: knowledge, values, opinion, beliefs or theories.

We have the data universe of the whole world partitioned into the finite number of datasets (categories and classes, kinds or types or variables) of an innumerable?number of data items/elements/points, as instances, individuals, cases, or values (tokens)?

Broadly speaking, all data falls into one or more of five categories:?nominal,?ordinal,?interval,?ratio, and number, going as the levels or scales of measurements.?

  • Nominal data/numbers, the simplest data type, classifying (or naming) data without suggesting any implied relationship between those data,?as basic ontological categories, countries or species of animals.
  • Ordinal data/numbers,?classifying data but it introduces the concept of ranking, as all hierarchical?ontological categories or the Likert scale or ‘slow’, ‘medium’, ‘fast’
  • Interval data/numbers,?both classifying and ranking data (like ordinal data) but introduces continuous measurements, as the time of day or temperature measured on either the Celsius and Fahrenheit scale.?
  • Ratio data/numbers,?it classifies and ranks data, and uses measured, continuous intervals, just like interval data. But, unlike interval data, ratio data has a true zero,?an absolute, below which there are no meaningful values. All physical quantities, as mass, speed, age, or weight are examples?of the?RD
  • Numeral data/Numbers (Digital, text or binary, data and?analog real-valued data),?they?count,?measure, order, and?label, number sets or number systems, N-natural numbers, Z-integers, Q-rationals, R-reals and C-complex numbers (number theory, set theory, arithmetic, statistics, algebra, probability theory)??

The first four classes were introduced in 1946 by the psychologist Stanley Smith Stevens,?as a level of measurement?or?scale of measure,?widely used in sciences and engineering, statistics, data analytics, data science and business marketing. What is missing is the base, ground, or foundation of any real metrics, the base of reference and of?measurement?units for counting or measuring,?labeling and ordering,?numbers.

They are used for counting and measuring, for labels (as with?numbering schemes for assigning nominal numbers to entities, names, ID numbers, routing numbers, telephone numbers, IP addresses), for ordering (as with?serial numbers), and for codes (as with?ISBNs, bank codes, postal codes)".

In all, it is a hierarchical scale, each level builds on the one that comes before it, as?nominal numbers?> ordinal numbers > interval numbers > ratio numbers > numbers.

It is crucial that for intelligent machines, the Data Universe Pyramid is replacing the?DIKW pyramid, the?DIKW hierarchy,?wisdom hierarchy,?knowledge hierarchy,?information hierarchy, information pyramid, or data hierarchy,?the Data, Information, Knowledge, Wisdom.

For Data to represent the things in the world, the World Data Ontology is necessary to complete the data science and statistics and engineering,?transforming the statistical AI and ML/DL into the real AI and ML/DL.

Real AI as Augmenting Human Intelligence or Fake AI as Replacing Human Intelligence

Most of high-paid works are automatable, and could be effectively replaced by narrow/weal/imitative/human AI/ML/DL machines.

Most of jobs could be eliminated by the emerging human-like and human-level AI within 5–10 years.

It is particularly white-collar jobs, which "mostly amenable" to HL/HL AI technologies, as mentioned in “ChatGPT may be coming for our jobs. Here are the 10 roles that AI is most likely to replace”:

Tech jobs (Coders, computer programmers, software engineers, data analysts)

Media jobs (advertising, content creation, technical writing, journalism)

Legal industry jobs (paralegals, legal assistants)

Market research analysts

Teachers

Finance jobs (Financial analysts, personal financial advisors)

Traders

Accountants

Customer service agents…

Again, most of Fortune 500+ will be disrupted by AI-industries.

Industries to be Disrupted by Human-like and Human-level AI

  • High tech
  • General manufacturing
  • Automotive
  • Life sciences/Healthcare
  • Banking, Financial Services, and Insurance (BFSI)
  • Logistics
  • Retail
  • Cybersecurity
  • Transportation
  • Marketing
  • Defense
  • Lifestyle
  • Space-transportation

Conclusion

Most of Fortune 500+ will be disrupted but-for the RIP strategic consultancy to survive and prosper.

A real/true/causal AI strategy consultant?works with a company's CEO, board, and senior managers to provide strategic, unbiased advice for smart vision, strategy, plan and various business decisions and goals.

A fake AI strategy consultant works with a company's CEO, board, and senior managers to provide strategic, biased advice for various business decisions and goals.

A RIP strategic consultant's job is to diagnose a company's problems and prescribe strategies to fix them to survive and prosper in the AI world.

Go to a true strategist if you need to figure things out or make decisions on the big picture of the intelligent future.

Go to a consultant if you need knowledge or execution support on a particular topic.

Go to a coach if you need to work on your mindset, daily difficulties to running a business, or behaviors to better reach your goals.


要查看或添加评论,请登录

社区洞察

其他会员也浏览了