The AI Effect, ChatGPT, and True AI
"Knowledge is Power, World Knowledge is World Power" ["Scientia potentia est, mundi scientia mundi potentia est"]
“Man is a transitional animal. He is not the climax of creation.” ―Carl Sagan, Cosmic Connection: An Extraterrestrial Perspective
What is the AI Effect?
The AI effect occurs when onlookers discount the behavior of an artificial intelligence program by arguing that it is not "real" intelligence. "AI is anything that has not been done yet."
The reason is that AI science and technology is without any precise, universally accepted definition of AI. The existent ones as different as below:
Artificial intelligence (AI), the ability of a digital computer or computer-controlled robot to perform tasks commonly associated with intelligent beings.
Artificial intelligence (AI) is the intelligence of machines or software, as opposed to the intelligence of humans or animals.
“Artificial intelligence is that activity devoted to making machines intelligent, and intelligence is that quality that enables an entity to function appropriately and with foresight in its environment.”
The AI effect might refer to all specialised artificial intelligences, rather than the networks of multiple specialised AIs.
As far as technological development is a never-ending process, the AI effect will also continue without end.
Then the AI effect relates to humans as well; for once future AI robots take over, they might consider human intelligence as not real intelligence.
What is really AI and just Software Automation
Artificial Intelligence is distinct from that of a human being. I is not about mimicking or paralleling or simulating human activity, intelligence, and learning.
It is about activity, intelligence, and learning for machines which are distinct from that of humans.
The label AI or machine intelligence or machine knowledge or machine learning makes real sense when considering tasks which are not traditionally associated with human (or animal) mental or physical activity.
Machine learning is on probabilistically estimating basic patterns and rules rather than the concepts, principles, and methods of human learning. It is discovering complex data patterns, rules, and relationships which require performing extremely complex mathematical calculations and modeling.
ML is trained to detect that a certain sequence or arrangement of data conforms to a known pattern, finding instances of a known pattern. But it is unable to discover a new pattern, an abstract pattern, to recognize or discover the semantic significance of a newly discovered pattern. Its cause. Its consequences. Its relationship to other patterns.
AI, as machine intelligence (MI) or machine learning (ML) or machine knowledge (MK), is concerned with computing technologies that allow machines to model and simulate reality, to know, learn, infer, interact, and solve problems of any computable complexity in any range of environments.
The AI algorithms working with world's data are discovering new data patterns and infer rules, or WORLD KNOWLEDGE, as objects and properties, concepts and laws, principles and methods.
Whatever without encoded a priori world or background knowledge and higher level intellectual capacities, as knowing and effectively interacting with the world, developing complex concepts or data structures, principles, and methods, discovering causal patterns and inferring deductive, inductive or causal rules from the information from the environment, is not true AI, MI, MK or ML.
Some forms of advanced software or hardware automation impersonated as machine intelligence or artificial intelligence or machine learning include:
The mere complexity would not automatically confer the label of AI or ML. It is rather the degree that the algorithms working with the environment and its data discovering new objective patterns and rules, or classes, principles, and methods.
If a complex device or system, hardware or software is to automatically and autonomously recognize new interaction patterns and objective rules, concepts and principles, all what implies higher-order intellectual activity, then AI and machine intelligence or machine knowledge and learning might become appropriate labels.
The difference between any forms of automation, an arithmetic calculator or GPT-4, and true artificial intelligence or a human brain is not of scale and speed, but of world knowledge and deep learning, autonomy and generality and interactivity.
What is ChatGPT?
ChatGPT (Chat Generative Pretrained Transformer) is a chatbot that produces human-like AI-generated content based on the input it is given by a user. It was developed by Open AI and released in November 2022.
ChatGPT uses ‘transformer architecture’, a deep learning technique that works through terabytes of data containing billions of words in order to create answers to questions or prompts entered by a user. It is a version of machine learning Natural Language Processing models called Large Language Models (LLMs).
Since its launch, ChatGPT has gone viral as a human-like chatbot that responds to users based on what they input.
The tool is able to answer questions and produce responses based on a dataset of 300 billion words and 175 billion parameters.
Its extension, ChatGPT-4, is sold as what is really missing the world/background/general knowledge as a pre-trained or embedded or encoded library of scientific knowledge and frameworks of ontological categories and classification algorithms.
"GPT-4 is OpenAI’s most advanced system, producing safer and more useful responses... GPT-4 can solve difficult problems with greater accuracy, thanks to its broader general knowledge and problem solving abilities...GPT-4 surpasses ChatGPT in its advanced reasoning capabilities."
领英推荐
Now, reading the statistics below, I recall C. Sagan's wisdom: “Man is a transitional animal. He is not the climax of creation.”
"ChatGPT as the iPhone moment for AI"
Here is a breakdown of the lists, prompts-collections, resources, articles referenced in the ChatGPT story:
30x: Intro to ChatGPT and foundation models; 14x: Free Prompt Engineering Courses, Masterclasses & Ressources; 500+: Marketing Content Creation; 50x: AI Art Prompts; 16x: Music ;1,500+: Business — from accounting to zoology; 280+: Coding; 100x: Development tools, Extensions; 30x: Autonomous AI Agents built on top of ChatGPT; 1,000x: Funny, amazing, interesting; 20+: Passing exams; 50+: Jailbreaking; 50x: ChatGPT is a fail; 100x: Superintelligence will kill us all
The ChatGPT website currently receives an estimated 1.6 billion monthly website visitors (a 9.73% decrease on the previous month’s 1.85 billion visitors, but a 66% increase from February 2023 and an increase of around 1 billion from January 2023), with an estimated 100 million active users
However, out of the top 1000 websites, 12% are already blocking GPTBot, and the number is rising fast.
OpenAI’s revenue predictions for ChatGPT are $200 million by the end of 2023 and $1 billion by the end of 2024.
ChatGPT contains 570 gigabytes of text data.
It is estimated that training the model took just 34 days.
The tool costs approximately $100,000 per day or $3 million per month to run on Microsoft’s Azure Cloud, with each word generated costing $0.0003.
The majority (62.52%) of OpenAI’s site visitors are aged between 18 and 34, and 65.68% are male compared to 34.32% female.
An average of 53% of people can’t tell that ChatGPT content was generated by an AI.
When GPT-3 launched, it caused AI token cryptocurrency prices to jump by up to 76.7%.
ChatGPT is currently supported in 161 countries according to OpenAI.
Why Is AI so Important?
AI technology could be the last boom or bust for humanity. There is a massive existential, worldwide struggle around AI, Real/Authentic AI vs. Fake/Imitating AI, while most people are trying to deal with propaganda and disinformation, poverty and unemployment, spiraling inflation and falling economies, geopolitical strife and painful recovering from the global pandemic.
The secret sauce of Real AI vs. Fake AI is the causal/scientific/conceptual/mathematical/interactive models of reality, as the world data learning and inference and interaction framework.
The big problem is that “Our Brains Don’t Understand Artificial Intelligence” (Medium, 11 April, 2023) in the wise ways. Our brains, minds really, are struggling to truly comprehend and understand the nature of Artificial Intelligence.
It seems only a few alien intelligences are aware of a simple life-critical truth:
There is no AI tool in existence today, if only as counterfeit AI tools, as listed below, that has no pieces of awareness and consciousness, sentience or real intelligence, to interact with the world in the most effective and efficient, smart and sustainable ways.
In the right ways, these fake AI companies should be prosecuted for a commercial counterfeiting (a fraudulent imitation) of the human body/brain/mind/intelligence/behavior, as "to imitate something authentic, with the intent to steal, destroy, or replace the original, for use in illegal transactions, or otherwise to deceive individuals into believing that the fake is of equal or greater value than the real thing".
And it is naive, if not stupid, to expect The New Society of AI (Medium, 6 April, 2023) when with more connections than the number of synapses in the human brain, GPT-5, or whatever may match or supersede real human intelligence.
Resources