This text discusses the concept of biological computing and the idea that ecosystems, such as ponds, can function as intelligent systems. Stafford Beer and Gordon Pask worked on a project to build ecosystems that could act as computers, using inputs and outputs. Beer experimented with using small organisms and entire pond ecosystems as homeostatic controllers. The text argues that biological systems, like ecosystems, can solve complex problems beyond our cognitive abilities. It suggests that intelligence is not limited to human beings, but can be found in various forms throughout nature. The concept of feedback and the accumulation of experiences are highlighted as essential components of intelligent behavior. The text also mentions the use of language models like GPT-4 for predicting and generating coherent communication, questioning whether this can be considered a form of thinking. The idea of a “pond brain” is presented as a metaphor for the potential intelligence found in natural systems.
| Signal | Change | 10y horizon | Driving force |
|---|---|---|---|
| Pond brains and GPT-4 | Shifting perspective on intelligence and computing | Greater recognition of nonhuman intelligences | The need to solve complex problems beyond human capabilities |
| Biological systems as problem solvers | Recognizing the problem-solving abilities of biological systems | Increased integration of biological systems in problem-solving tasks | Desire to solve problems that exceed human cognitive abilities |
| Ecosystems as computers | Viewing ecosystems as computational systems | Increased understanding and utilization of ecosystems for computing tasks | Recognition of the computational capabilities of ecosystems |
| Intelligence in various entities | Expanding the definition of intelligence | Greater acceptance and respect for nonhuman intelligences | Broadening our understanding of intelligence |
| Representation vs. performance | Emphasizing performance over representation | Focus on practical outcomes rather than symbolic representation | Complexity of systems and the limitations of symbolic representation |
| Feedback as a requirement for intelligence | Recognizing feedback loops as a marker of intelligence | Identifying feedback loops in various systems as indicators of intelligence | Understanding the importance of feedback in intelligent behavior |
| Intelligence beyond the brain | Acknowledging intelligence in non-brain entities | Recognition of intelligence in DNA, rainforests, markets, and climate systems | Expanding the concept of intelligence beyond the human brain |
| Questioning LLMs’ thinking abilities | Challenging the claim that LLMs are thinking | Evaluation of LLMs’ thinking capabilities and limitations | Exploration of the nature of thinking and intelligence in LLMs |
| Continued pursuit of “pond brain” concept | Persistence in the development of natural and unconventional computing | Advancements in natural computing and unconventional computing | Desire to create computing systems inspired by biological intelligence |