Why the New Artificial Intelligence Is So Powerful |
More than a billion people are now using artificial intelligence (AI) models regularly, for purposes ranging from work to advice about personal relationships. This trend began with the introduction of ChatGPT in November 2022, so in only three years, AI has gone from an obscure research topic in computer science to a daily tool. What makes the new AI so much more powerful than previous approaches?
The power of the new AI comes from three sources: mechanisms, causal networks, and emergent properties. Mechanisms are combinations of interconnected parts whose interactions lead to regular changes. Causal networks are systems of causes based on multiple mechanisms. Emergent properties are ones possessed by whole systems but not by their components, because the novel properties result from interactions among the components and their functional mechanisms. Current AI systems are powerful because their mechanisms interact to produce causal networks with emergent properties that approximate human intelligence.
Here are the six most important mechanisms operating in AI systems such as OpenAI’s ChatGPT, Google’s Gemini, Anthropic’s Claude, xAI’s Grok, and Meta’s LLaMA.
No one or two of these mechanisms would be sufficient to produce the amazing verbal, pictorial, and musical properties of current AI systems. Here are some of the causal interactions that form a complex network.
Emergence is common in natural systems, ranging from water molecules to diseases. H2O has the property of being liquid at room temperature, in contrast to its components, hydrogen and oxygen, which are gases. Diseases such as diabetes result from causal interactions of genetic, cellular, behavioral, and social mechanisms. Similarly, AI systems gain from their interactions of neural networks, learning, attention, and specialized chips various emergent properties. AI systems lack some key components of human intelligence, including sensory consciousness, emotions, and other kinds of feelings. But here are some emergent properties not found in less complex systems that approximate important aspects of human intelligence.
These emergent properties result from the interactions of causal networks based on underlying mechanisms, as shown in this figure.
This pattern of explanation that combines mechanisms, causal networks, and emergent properties deserves a name. I dub the pattern "micro/macro emergence" because the basic part-whole mechanisms operate at the micro level, and the causal networks are macro-mechanisms that incorporate multiple micro-mechanisms. The combination of micro and macro operations leads to the emergence of novel properties not found in either the micro-mechanisms or causal networks. Here are some more examples.
Hence, the micro/macro emergence explanation pattern is much broader than just explaining why AI models have become so powerful.