How to remedy today's AI/ML/DL/ANNs: Trans-AI as Real Machine Intelligence and Learning
AI with its derivative terms looks as one of the worst misconceptions and misnomers in modern science and technology, widely used and abused, while arresting the real development of real machine intelligence.
And A. Turing had never used the term of AI in his paper COMPUTING MACHINERY AND INTELLIGENCE.
The lethal misconception with the 70+ years mainstream paradigm, AI/ML/DL/ANNs is a simulation of human intelligence/mind/brains/behavior, instead of reality, causality and mentality.
As a result, it is ending up with a Fake Intelligence, namely:
The Lethality of Narrow Superhuman AI/ML/DL
The most concern of humanity must be the current accelerated growth of Big Tech’s Narrow and Weak AI of Machine Learning, ANNs and Deep Learning, as a Fake AI vs. Real World AI.
It is fast emerging as narrow-minded automated “superintelligences” outperforming humans in any narrow cognitive tasks, and implemented as
LAWs or military AI,
ML/DL drones, killer robots,
humanoid robots,
self-driving transportation,
smart manufacturing machines,
RPAs,
cyborgs,
trading algorithms,
smart government decision makers,
recommendation engines,
medical AI system., all what is so scaring Musk.
The whole idea of Anthropomorphic and Anthropocentric AI (AAAI) as the narrow or general ones, aimed at simulating human intelligence, cognitive skills, capacities, capabilities and functions, as well as intelligent behavior and actions in computing machines is raising a number of undecidable social, moral, ethical and legal dilemmas.
The narrow and weak Deep-Learning AI programs classify tremendous amounts of data without any understanding of the world and meaning of their inputs or outputs (e.g., the recommendation to treat or a risk score or behavioral changes).
It consequences could be much worse than human cloning, which is prohibited in most countries, and massive technological unemployment without any compensation effects is just the beginning of the end.
This is what good minds forewarned humanity about the the possibilities and possible perils of AAAI, mimicking human learning and reasoning by machines and humanoid robots:
“The development of full artificial intelligence could spell the end of the human race…?It?(https://race….It)?would take off on its own, and re-design itself at an ever increasing rate. Humans, who are limited by slow biological evolution, couldn’t compete, and would be superseded.”
— Stephen Hawking told the BBC
“I visualise a time when we will be to robots what dogs are to humans, and I’m rooting for the machines.”
—Claude Shannon
“I’m increasingly inclined to think that there should be some regulatory oversight, maybe at the national and international level, just to make sure that we don’t do something very foolish. I mean with artificial intelligence we’re summoning the demon.”
—Elon Musk warned at MIT’s AeroAstro Centennial Symposium
All what we need, its is a radically new kind of AI, Real and True MI, or Real World AI, which is to simulate and understand or compute reality, causality and mentality in digital reality machines.,
This is becoming clear for some smart industrialists, as E. Musk, who understands that without the Real World AI no really intelligent machine is possible.
“Self-driving requires solving a major part of real-world AI, so it’s an insanely hard problem, but Tesla is getting it done. AI Day will be great”. “Nothing has more degrees of freedom than reality.”
From a lethal AAAI to Trans-AI as Real Machine Intelligence, Cyberintelligence, Technological Intelligence, or Computing Intelligence
Today's AI's biggest lethal misconception consists in a widely established cognitive bias that AI must have the same nature as human intelligence or human mind or human brains, while "most work in AI involves studying the problems the world presents to intelligence rather than studying people or animals".
Its misnomer comes from a strong desire of its initiators to label it differently from cybernetic intelligence or machine intelligence or computing intelligence or techno-intelligence by any ways, regardless of logical inconsistency from its original definition:
领英推荐
The AI synonym ring (synset) includes:
Machine Learning
Deep Learning
Artificial Neural Networks
Cognitive Computing
Narrow AI
Strong AI, General AI
Artificial Superintelligence, Superhuman Intelligence…
This all comes from the massive?misconceptions on intelligence, based on cognitive biases, poor understanding and/or in-group reasoning.
Any AI which is Artificial Human Intelligence (AHI), Psychological or Cognitive AI or Behavioral or Biological AI or Statistical AI, is scientifically defective as being subjective and anthropomorphic.
Anthropomorphism in AI is the attribution of distinctively human-like feelings, mental states, and behavioral characteristics to computing machinery and robots.
The tendency in popular culture to conceive of AIs as people-like (both behaviorally, emotionally, cognitively, and morally) is badly influenced by fictional narratives (literary science fiction, films and tv shows), media coverage of AI and robots, and AI research community.
Real Machine Intelligence is NOT after comprehending human Psychology through formal logic or statistic ML algorithms. Real MI is NOT about mimicking human brain, intelligence, mind or behavior, all conscious or unconscious phenomena, from perception to desire or personality.
So-called, classical, logical or Symbolic AI had failed due to its defective assumptions to replicate the cognitive thinking of humans with all accuracy and precision.
The intelligent algorithms, bots, cobots, robots, humanoids, or digital humans, unlike human intelligence, need to be constantly fed with real-world data, and human minds never could beat real AI machines in terms of data processing, computational power and speed of execution.
Again, there is no biological plausibility or psychological and mental similarity between natural (e.g., human) and artificial intelligences.
AI as Statistics-Driven Machine Learning Deep Neural Networks
Today's AI of ML, DL and Deep Neueal Networks is driven by statistics, big data analytics, its processes, methods, techniques and algorithms.
Statistics and science are from two different worlds: "There are three kinds of lies: lies, damned lies, and statistics".
The more massive data sets, the more biases and prejudices, disinformation and social engineering, etc.
“Studies show that biases and mistakes color many of the libraries used to train, benchmark, and test models, highlighting the danger in placing too much trust in data that hasn’t been thoroughly vetted — even when the data comes from vaunted institutions”.
What Does It Mean for MI to Understand?
How can we know or determine in practice whether a machine can understand? In 1950, the computing pioneer Alan Turing tried to answer this question with his famous “imitation game,” now called the Turing test.
While never using the term of AI in his seminal paper COMPUTING MACHINERY AND INTELLIGENCE, machine intelligence was reduced to the human language understanding.
"The crux of the problem is that understanding language requires understanding the world, and a machine exposed only to language cannot gain such an understanding.?
Humans rely on innate, pre-linguistic?core knowledge?of space, time and many other essential properties of the world in order to learn and understand language. If we want machines to similarly master human language, we will need to first endow them with the primordial principles humans are born with. And to assess machines’ understanding, we should start by assessing their grasp of these principles, which one might call “infant metaphysics.”
Training and evaluating machines for baby-level intelligence may seem like a giant step backward compared to the prodigious feats of AI systems like Watson and GPT-3. But if true and trustworthy understanding is the goal, this may be the only path to machines that can genuinely comprehend what “it” refers to in a sentence, and everything else that understanding “it” entails".
Still, it is missing the core point: "It’s simple enough for AI to seem to comprehend data, but devising a true test of a machine’s knowledge has proved difficult".
This is the whole issue of real and true MI, to comprehend/understand, intelligently process/compute DATA to be reached by the Trans-AI paradigm.?