It seems like the world is buzzing lately with news about Artificial Intelligence. More specifically, the thing that has everyone talking is a form of AI called ‘Generative AI’.
Exactly what is generative AI, how does it work and what is all the buzz about?
Generative AI in simple terms
Generative AI broadly describes a type of artificial intelligence that is capable of generating text, images, code or other types of content. It uses ‘machine learning systems’ to get very good at certain tasks and then uses those skills to create new things (images, dialog, computer code, video & audio, etc.) The goal of AI development has been to enable a computer to generate a human-like response.
How does generative AI work?
Generative AI uses a computing process known as deep learning to analyze patterns in large sets of data and then uses these learned patterns to create new data that appears human-generated. It does this by employing neural networks, a type of machine learning process that is loosely inspired by the way the human brain processes, interprets and learns from information over time. Except it is infinitely faster and more precise than human learning.
To give an example, if you feed lots of fiction writing into a generative AI model, it eventually will gain the ability to craft stories based on the literature it’s been trained on with similar elements of plot structure, characters, themes and other narrative devices. This is because the machine learning algorithms that power generative AI models learn from the information they’re fed. The digital equivalent of: “you are what you eat.”
Generative AI models are getting more sophisticated very quickly — the more data they are trained on, the more convincing and human-like their outputs become. ChatGPT (created by OpenAI/Microsoft) and Bard/Gemini (Google) are quickly consuming the contents of the entire internet.
Where is it going?
Microsoft, Google, Amazon and Meta have all launched their own generative AI initiatives, into which they are investing billions of dollars, spawning a new race to build more and more capable AI models. Engineers involved in the efforts have described their role as “ushering in” the age of advanced computer thought as a new form of independent intelligence – a new form of ‘virtual life.’
Types of generative AI models
Various types of generative AI models are being developed, each designed for specific tasks and purposes. These can broadly be categorized into the following types.
Foundation models
Foundation models are described as the bedrock of generative AI chatbots, but could better be understood as the ocean of interpreted data in deep learning systems. Foundation models are trained on vast amounts of data, forming the foundational reference model for AI. Foundation models can be trained to answer general questions, such as answering customer service or technical support questions for a particular business.
OpenAI’s GPT-4, Amazon’s Titan and Anthropic’s Claude are examples of foundation models.
Transformer-based models
Transformer-based models are trained to understand the relationships between sequential information like words and sentences. Using deep learning, transformer-based models are good at natural language processing and understanding the structure and context of language, making them best suited for Chat and writing tasks.
Generative adversarial networks (GAN)
Generative adversarial networks are made up of two neural networks known as a generator and a discriminator, which work against each other to create authentic-looking data. The generator’s role is to generate convincing output, while the discriminator works to evaluate its authenticity. Over multiple cycles this process results in more and more convincing outputs. This method is being used for image generation by AI tools like DALL-E and Midjourney.
Variational Autoencoders (VAE)
Variational autoencoders also use two networks to interpret and generate data — in this case, an encoder and a decoder. The encoder takes the input data and compresses it into a simplified format. The decoder then takes this compressed information and reconstructs it into something new that resembles the original data but isn’t entirely the same.
For example, it can learn from hundreds of images of human faces, simplifying them into a few important characteristics — such as the size and shape of the eyes, nose, mouth, ears and so on — and then use these to create entirely new faces or to increase the accuracy of facial recognition systems.
Multimodal models
This is where most AI development is ultimately heading. Multimodal models can ‘consume’ multiple types of data simultaneously, such as text, images and audio, allowing them to create more sophisticated outputs. Examples already exist of AI-generated video clips that mimic the facial expressions and voices of real people with uncanny accuracy.
DALL-E 3 and OpenAI’s GPT-4 are multimodal models.
Regenerative AI
Regenerative AI refers to AI systems that can fix themselves or improve over time without human help. The concept of regenerative AI is intended for building AI systems that can last longer and work more efficiently, such as in remote locations. Examples have been documented of some AI systems creating entirely new AI entities of their own without human input.
Frontier models
“Frontier model” is a term for hypothetical upcoming AI that could far surpass the capabilities of today’s AI. There is no single definition of a frontier model’s capabilities, except that they will be more extensive and more powerful than the AI available today. These are the models being cultivated in mega-billion-dollar research labs today that will have the most far-reaching impact on humanity in the future.
Benefits, Dangers and Risks
There are certainly benefits to be gained from AI technology, including tremendous potential improvements in Healthcare, Pharma research, and engineering. AI models hold potential for labor and time savings in customer support, transportation, software development and research.
A flip side of these benefits, of course, is the potential for job losses as humans are no longer needed in these roles. One major concern is the potential for using AI to spread fake information to discredit rivals or destroy their reputation. AI systems could be used to manipulate public opinion across social media or to perform widespread surveillance. Cyber security professionals warn of possible ‘AI Bot’ attacks against systems and critical infrastructure. The US Military has already predicted that autonomous robotic weapons will be a reality within eight years, including potential control of nuclear weapons.
In November 2023, 16 agencies including the U.K.’s National Cyber Security Centre and the U.S. Cybersecurity and Infrastructure Security Agency released Guidelines for Secure AI System Development, which promote security as a fundamental aspect of AI development and deployment.
On Feb. 13, 2024, the European Council approved the AI Act, a first-of-kind piece of legislation designed to regulate the use of AI in Europe. The legislation attempts to regulate AI, with some AI systems banned outright. It could be a model for use globally in the future.
In the hands of a globalist government the greatest risk could be to our personal liberty.
AI is Different from other Technical Advances
Advances in technology are nothing new, mankind has been using increasingly sophisticated tools throughout history. Until now these advances have been extensions of human capabilities. They have allowed humans to do what we do faster, better and farther, or to create weapons that are unimaginably lethal. These have ultimately been tools that must be wielded, guided or driven by people.
AI represents a unique distinction from all of history’s prior advances. It is the spawning of a ‘thing’ that is not a tool that is wielded, but rather is, itself, a creator of tools. Soon the tools that it creates will be vastly more sophisticated than any human ability to understand them.
It isn’t hard to see how this opening of a Pandora’s Box could impact the future. Sci-fi writers have been warning us about it for years. Optimists will say that we’re smart enough to manage it and keep it safely within the box. Pessimists ask, what if we’re not?
It begs the most important question of our age – how much longer do we have?
The Bible's glimpse into end times events includes an interesting detail -- that if those days were not shortened then no one would survive. Christ's coming will ultimately prevent mankind from annihilating itself.
The capability for self-annihilation is a modern development. When combined with global AI-based controls and autonomous weapons this risk is likely to become even more precarious.
“Now, concerning the times and the seasons, brothers, you have no need to have anything written to you. For you yourselves are fully aware that the day of the Lord will come like a thief in the night. While people are saying, “There is peace and security,” then sudden destruction will come upon them as labor pains come upon a pregnant woman, and they will not escape.
“But you are not in darkness, brothers, for that day to surprise you like a thief. For you are all children of light, children of the day. We are not of the night or of the darkness. So then let us not sleep, as others do, but let us keep awake and be sober.” ~ 1 Thessalonians 5:1-6
The Apostle Paul told the Thessalonians that the church would have every opportunity to anticipate Christ’s coming by looking at the signs.
He has given us all we need to know.
Here's what the Bible tells us to watch for:
- The dry bones reborn — Israel’s return to nationhood after eighteen hundred years. It’s done!
- Increased earthquakes…
- People will be bewildered by roaring seas and storms…
- Wars and rumors of wars…
- Famines…
- False Christs and False Prophets with cults that lead people astray…
- Evil shall increase…
- Many will hate and devour one another…
- The love of many will wax cold…
- Deception and delusion will come over the world’s population…
- Believers will be hated and persecuted…
- There will be armies massing against Israel.
All of these are now current events. The small remaining few are poised and ready -- these could be fulfilled at any time...
- The Gospel is preached throughout the entire world…
- A global (crypto?) currency that allows control over commerce...
- The coming together of the World’s Western Governments under one banner...