#general artificial intelligence

Singularity vs. Spirituality – Playing with Paradox Pairing

The man vs. the machine, the mind vs. the matter, and the potentiality vs. the actuality, the perception vs. the reality and the brain vs. the mind, the conscious vs. unconscious mind and the paradox between these pairing of words keeps us intellectually busy, otherwise how do we manage our monkey mind. We operate in a world of duality, constantly striving to reconcile between this two world. One…

At present, the only information that these stabs at artificial intelligence get is through our curation. Some subset of humanity decides what goes into the training models and tweaks the algorithms, and then we get them to spit out gobs of text, images, and even video. Personally, I’ve never seen much value in all of that since we produce human intelligence cheaply, with less of a carbon…

Artificial intelligence (AI) is the field of computer science that addresses the cognitive problems typically associated with artificial intelligence, such as learning, problem-solving, and pattern recognition. Mention artificial intelligence in Blockchain (often abbreviated to “AI”) and one might think of robots or futuristic scenarios. However, AI is not limited to robots in science fiction but has also entered the field of advanced computer science in modern non-fiction. Professor Pedro Domingos, an outstanding researcher in this field, divides machine learning into “five schools”, namely the symbolism school originated from logic and philosophy, the connectionist school from neuroscience, the evolutionary school related to evolutionary biology, the Bayesian theorem school that combines statistics and probability, and the analogical reasoning school that originated in psychology.

More recently, the Bayesian school of the theorem has made several advances in a field called “machine learning,” thanks to advances in the efficiency of statistical computation. Likewise, the connectionist school of thought has made progress in a subfield called “deep learning,” thanks to advances in network computing. Machine learning (ML) and deep learning (DL) are both fields of computer science that grew out of the discipline of artificial intelligence.

Use Cases

abnormal detection Find items, events, or observations that don’t match expected patterns or other items in your dataset.

fraud detection Build predictive models to help identify potentially fraudulent retail transactions, or to detect fraudulent or inappropriate item review results.

customer churn Identify customers who are at high risk of churn and allow you to aggressively offer them special offers or wider customer service to keep them.

content personalization Provide a more personalized customer experience using predictive analytics models to recommend items or optimize website traffic based on a customer’s previous behavior.

What is deep learning?

Deep learning is a branch of machine learning that includes various hierarchical algorithms aimed at gaining a better understanding of data. Unlike more basic regression algorithms, these algorithms are no longer limited to creating a set of interpretable relationships. Instead, deep learning relies on these non-linear algorithmic layers to create distributed representations that can interact based on a range of factors. For large training datasets, deep learning algorithms are beginning to be able to identify relationships between elements. These relationships may exist between elements such as shapes, colors, text, etc. From there, people can use the system to create forecasts. In machine learning and artificial intelligence, deep learning is powerful because the system can recognize relationships beyond what a human can actually encode in software, and recognize relationships that a human isn’t even aware of. When sufficiently trained, algorithmic networks can begin to predict or explain very complex data.

human approach

The first definition of an AI system is that they think like a human. Systems fall into this category if they can learn and solve problems like humans. Haugeland defined it in 1985 as a “thinking machine”. Hellman in 1978 pointed out that this category is “automating activities related to human thinking”.

A system that can perform human-like performance will meet the requirements of the Turing Test, which means that the system can behave like a human being, can communicate in English, understand other people’s speech, respond, can evolve, and can draw new conclusions on its own. Kurzweil defines this category as “the art of creating machines to perform functions that require human intelligence to perform.”

rational approach

The second category of AI systems primarily measures their ability to perform rationally, which differs markedly from human behavior, which is sometimes irrational. Likewise, such AI systems fall into two categories: rational thinking and rational action.

Charniak and McDermott in 1985 described a system that can think rationally as “using computer models to study the development of intelligence.” This is also known as the “law of thought” approach.

Aristotle was the first scholar to explain what he described as “right-thinking” or irrefutable reasoning. The example given by Russell and Norvig is “Socrates is a human being, and all human beings are mortal, so Socrates is a dead man.”

A system can also act rationally, exhibiting the skills listed in the Turning Test. Poole describes creating AI systems that can act rationally as requiring “computational intelligence, a study in designing intelligent agents”

weak artificial intelligence

The terms weak and strong are another way of distinguishing between different kinds of AI systems. A more appropriate term for thin AI might be narrow AI or artificial narrow intelligence. This kind of artificial intelligence only focuses on specific tasks, such as Apple’s Siri, Amazon’s Alexa, or Google’s self-driving cars.

powerful artificial intelligence

Strong AI includes two types of AI: artificial general intelligence (AGI) and artificial superintelligence (ASI). AGI is a conscious self-awareness system that can solve problems and even plan for the future. ASI is a system that surpasses human capabilities, but there are still no examples of ASI applications, you only see them in movies. 2001: A Space Odyssey has a computer system called HAL. If you remember this, you know roughly what the ASI system does.

How does machine learning work?

Machine learning uses examples of inputs and expected outputs (“structured data” or “training data”) to continuously improve and make decisions without programming them in a sequence of step-by-step instructions. This approach mimics actual biological cognition: children learn to recognize objects (such as cups) through the demonstration of the same object (such as various kinds of cups). Today, machine learning applications are widespread, including spam filtering, machine translation, and speech, text, and image recognition.

What is the difference between deep learning and machine learning?

AI systems are primarily viewed as learning systems; that is, machines that can become more competent at tasks typically performed by humans, with limited or no human intervention.

“Narrow AI” refers to technologies and applications designed to perform a single or limited task. It differs from “artificial general intelligence” or “general artificial intelligence”, which refers to an artificial intelligence system that can successfully perform any intellectual task that a human brain can accomplish, or a machine that far exceeds the hypothetical capabilities of a human brain.