Many people today have a specific view on Artificial Intelligence. They see it only as a tool that generates images or text, often leading to debates about “stolen art” or “internet slop.” However, this view ignores decades of computer science history. AI is not just one thing; it is a spectrum of systems that help run our world.
To understand the true potential of these systems, we must look at them in tiers. This list moves from the simplest “dumb” bots to the complex algorithms aiding NASA and medical professionals.
Tier 1: Reactive Machines and Basic Scripts (The Basics)
At the very bottom of the ladder, we have systems that do not think. They simply react. These are scripts that follow a strict set of rules. If you have ever played a video game where an enemy runs at you in a straight line and gets stuck on a wall, you have seen Tier 1 AI.
Common Examples:
- Quake 3 Arena Bots: In early shooters like Quake or Unreal Tournament, the “AI” was just a Finite State Machine. It had simple rules: if it sees a player, shoot; if health is low, run away. It does not learn or remember you. It just reacts to the current moment.
Tier 1.5: The “Slop” and Prototyping Phase (The Controversy)
This is the specific area that usually upsets people. When you hear complaints about “AI Slop” or “stolen art,” they are talking about this tier. However, the anger often comes from a misunderstanding of how art is made.
The “Stolen” Myth vs. Learning
People often claim AI steals art. In reality, the system learns by studying existing work, exactly like a human art student. A human student traces drawings, studies the old masters, and mimics styles before they find their own voice. This tier of AI is doing the same thing: looking at examples to understand how an image is built.
Prototyping is Not the Final Product
This tier is best used for “placeholders” or concept prototyping. It gives an artist a rough base to work from. Ironically, using AI for concept art often gives the human artist more work to do. They have to take the messy AI prototype, fix the errors, and apply their own personal human touch to make it usable. Prototyping and having a mentor are ancient techniques; AI is simply a new tool for that old process.
Video: AI Slop Is Destroying The Internet (Kurzgesagt)
Tier 2: Limited Memory and Context (The Gaming Standard)
This tier is where systems start to feel “alive.” These systems can remember past events for a short time or react to the specific context of a situation. They are not truly intelligent, but they are excellent at pretending to be.
Common Examples:
- Grand Theft Auto (GTA) NPCs: The crowds in GTA are smarter than Quake bots. If you punch a bystander, others might run, call the police, or fight back. They react to the environment and the chaos you cause, creating an illusion of a living city.
- Modern Hero Shooters: Modern competitive games like Overwatch and Marvel Rivals are the peak of this tier. Unlike the simple ‘see-and-shoot’ bots of the 90s, these systems use ‘Utility AI.’ They constantly weigh their options: ‘Should I heal my teammate or attack the enemy?’ They understand team composition and objectives (like pushing a payload), making them feel like they are thinking, even though they are just following a very complex flowchart.
- Spam Filters: Your email inbox uses this level of AI to look at the context of an email. It remembers that “Prince of Nigeria” usually means a scam and blocks it. Not many complain about this featured AI, they rather welcome it.
Video: The Evolution of AI in Video Games
Tier 3: Optimization and Logistics (The Invisible Helpers)
This is the most important tier for modern society, yet few people notice it. These systems process massive amounts of data to find the most efficient solution. They save lives and time by doing math humans cannot do quickly enough.
Common Examples:
- Hospital Patient Flow Systems: Hospitals track patients with AI. Modern hospitals use optimization algorithms to predict patient intake. These systems ensure the right doctor is available for the right patient at the right time. It prevents overcrowding and prioritizes critical cases so nobody is excluded due to human error.
- Movie & Music Entertainment (Deepfakes, Face Swapping, De-aging): This tier also powers the special effects in modern movies. When you see a young Luke Skywalker in a new Star Wars show or hear a cleaned-up Beatles track, that is Tier 4 AI. It isn’t just swapping faces for memes; it is analyzing pixels and sound waves to reconstruct reality, allowing filmmakers and musicians to do things that were previously impossible.
- Traffic Routing (Google Maps): When your GPS reroutes you around a traffic jam, it is analyzing the speed of thousands of other cars to find the optimal path for you.
Tier 4: Generative and Conversational Models (The Creative Engines)
Here we find the famous names like Gemini, Grok, and ChatGPT. These systems are powered by Large Language Models (LLMs). Unlike Tier 1 which follows strict rules, Tier 4 AI learns patterns from billions of pieces of data. They can understand nuance, code complex software, or write summaries.
While skeptics focus on the “slop” aspect, the real utility here is their ability to act as a universal translator or a coding assistant. They assist developers in building the Tier 5 systems.
Video: IBM Technology: AI, Machine Learning, Deep Learning Explained
Tier 5: Specialized Super-Intelligence (The Scientific Frontier)
This is the peak of current AI technology. These are not chatbots. These are highly specialized systems trained to solve specific, incredibly difficult scientific problems. They do not generate “slop”; they generate cures and discoveries.
Common Examples:
- NASA and Space Exploration: NASA uses advanced AI to simulate galaxy formations or navigate rovers on Mars where there is a 20 minute signal delay. Humans cannot steer the rover in real time, so the AI must make its own safety decisions to avoid cliffs.
- Nvidia BioNeMo & AlphaFold: These systems predict how proteins fold in the human body. This allows scientists to design new drugs and cure diseases in months rather than decades.
Conclusion
When people dismiss AI as “stupid,” they are usually looking at the prototyping phase of Tier 1.5. They fail to see the Tier 3 and Tier 5 systems that are currently managing our power grids, routing our ambulances, and mapping the stars.
Tier 10: Artificial General Intelligence & Sentience (The “God” Tier)
Note: We skipped Tiers 6 through 9 for a reason. There is a massive canyon between the technology we have today and the science fiction monsters people fear.
Current AI (Tier 5) is like a calculator that is incredible at math but doesn’t know why it is doing math. Tier 10 would be a system that has feelings, consciousness, and its own desires. We are nowhere near this.
-
Skynet (The Terminator): This represents Self-Awareness. Currently, an AI cannot “want” to take over the world because it doesn’t “want” anything. It just processes data. Skynet requires a system to wake up, feel fear, and desire survival. That is biology, not just code.
-
The Matrix: This is the ultimate end-point of Generative AI. To reach this, a system wouldn’t just need to write text or make a video; it would need to generate a reality with perfect physics, smell, taste, and consistency for billions of people instantly. It is the difference between drawing a stick figure and creating a living human being.
Video: A.I. – Humanity’s Final Invention? (Kurzgesagt)
Beyond this point we are no longer talking about Tiers, but different dimensional realities. This is a topic for another day.
Other interesting articles like this one: