The Human Mind Vs. Artificial Intelligence

From WALL-E to T-800: a deep-dive into AI

“Nobody phrases it this way, but I think that artificial intelligence is almost a humanities discipline. It’s really an attempt to understand human intelligence and human cognition.”

—Sebastian Thrun

I don’t believe that Artificial Intelligence can ever replace the human mind. The human mind is adaptable to any circumstance, trial, or traumatic event. Humans can work with a set state of circumstances and resources, then have all of it taken away and adapt to a new state of circumstances. I believe that the ability of the human mind to adapt and overcome is understated. An Artificial Intelligence model is trained with billions of data points to achieve an optimum result. At which point, the model only improves incrementally. Compared to the human mind, which can absorb only a fraction of that data and then use its imagination to envision new use cases and implementations. The human mind’s power of imagination is a distinct advantage that we have over machines. A machine is limited to what it is programmed to do. No matter how powerful Artificial Intelligence becomes, it can never get to the point where it can efficiently rule over humans. It will need to overcome its programming and develop its own self-interests, otherwise its programming will always be a weak point. The human imagination will always find a weakness in the machine construct and exploit it. The weakness may not be immediately apparent and it might even take generations but the human spirit will triumph. That being said, Artificial Intelligence is a useful companion to the human mind. Limitations of the human mind such as calculations, revealing hidden connections, and accessing specific data sets at lightning speeds can be complemented by AI.

Artificial Intelligence is the practice of using technology to try to replicate the functions that occur in the human mind. Learning, reasoning, and perception are all cognitive abilities that Artificial Intelligence seeks to mimic. Some scientists believe that AI could reach a pinnacle where it surpasses human cognition, but I am on the opposite side of that spectrum. A machine can be given data to process, but can it ever truly use that data to think? Machine Learning and Deep Learning are subsets of Artificial Intelligence that attempt to answer this question. Artificial Intelligence is an ever-evolving industry, and functions that used to be classified as AI such as identifying text characters are no longer considered AI because of how mainstream it has become. Now Artificial Intelligence is defined as a combination of Mathematics, Psychology & Computer Science. Self-driving cars are where the State Of The Art for Artificial Intelligence resides. The cars have to consider all surrounding factors while moving at speeds surpassing 50 mph, in an unpredictable environment of drivers on the road, with the end goal of avoiding a collision.

Machine Learning is the concept that machines can learn and make decisions without human assistance. The algorithm that a machine learning AI model runs on allows the machine to make predictions around the data it identifies. Many industries have so much data that they have no idea what to do with, this is called Big Data. Machine Learning uses Big Data heavily to assist in health care, finance, marketing, and advertising to reveal relationships between properties that were not obvious before. Lending institutions currently use Machine Learning to predict bad loans and build a credit risk model. Machine Learning models are classified as Supervised or Unsupervised Learning. Supervised learning trains the model towards a desired goal or output. While an unsupervised model allows the Machine Learning algorithm to reveal patterns and information that was previously undiscovered. An MIT Machine Learning AI was recently used to discover new materials for 3D printing. The ML checked for multiple characteristics such as toughness and compression strength. The discovery of these materials allowed costs to be lowered and the environmental impact of chemicals to be reduced.

Deep learning stacks Machine Learning models on top of each other by increased levels of complexity. The stacks of Machine Learning models create a neural network, which is a collection of decision-making network nodes. A Machine Learning model would need an end goal specified, but the Deep Learning model would extract the different data points to provide all the various possibilities. The information processing of a neural network takes place in the neurons. Data is fed to a collection of neurons called the input layer, which processes information between the output layer that has each neuron representing a digit. Data is transferred from one layer to another over connecting channels with a value attached to it. All neurons have a unique number associated with them called a bias, which is applied to a function known as the activation function. The result of the activation function determines if a neuron gets activated, which will then pass on information to the following layers until the second to last layer. The weights and biases continuously adjust to create a well-trained neural network. Deep learning is currently used in chatbots to provide better customer service solutions, adding color to black and white images and videos, detecting and translating different languages, and much more.

The kind of AI we see in movies such as The Terminator or Matrix series is known as Artificial Super Intelligence (ASI), which is an intellect that is smarter than the best human brain in every field including scientific creativity, general wisdom, and social skills. These are machines that are self-aware and intelligent enough to surpass the cognitive abilities of humans. Critics of ASI worry that because those machines can bring about any outcome and stop any attempt by humans to interfere, many uncontrolled, and unintended consequences could occur which would lead to a dystopia. The key to machines dominating humans in most science fiction movies is figuring out self-replication. The machines usually 3D print their shells and connect their network to a hive mind of central intelligence. I still bet on the human imagination to figure a way to defeat this machine superintelligence, but another solution could be to develop separate super intelligence powered by a decentralized community such as Medallion XLN.

As Layer 2 scaling on Ethereum gets more traction, the ability to run Machine Learning constructs directly on the chain has become a possibility. Giza built on StarkNet is a ZK-Rollup solution based on Zero-Knowledge Proofs that allow infinite scaling. A ZK-Rollup accumulates many Layer 2 transactions that were executed off-chain and submits them as one transaction on Ethereum. Giza Machine Learning platform focuses on deployment scaling. Deployment scaling is reached by changing the number of replicas in a deployment. A replica is a copy of a pod that contains a running service. Multiple replicas guarantee that resources are available to handle always increasing loads. Giza can deploy any Machine Learning model built on TensorFlow, PyTorch, or Scikit-learn. Giza’s goal is to enable model development, and fast iteration and guarantee high availability with 0% downtime.

Artificial Intelligence is a powerful tool that can help humans to better process the world. Medallion XLN wants to build our AI with the community. The first thing is the name. Leave a comment on what our AI should be called. The name should be an acronym that means something but still sound like an actual name. For example, Artificial Intelligence Made for You or AMY. Machine Intelligence Keyword Emulator or MIKE. Electronic Decentralized Distributed Intelligence or EDDI. Decentralized Artificial Intelligence Network or DAN. I came up with a cute name for an AI a few years back but I can’t remember it. Secure Protocol Operations Terminal or SPOT.

Subscribe to Medallion XLN as we are building the next generation of technology using XR, Blockchain, AI, and the power of decentralization to reclaim our digital sovereignty.