好色先生TV

A research collaboration between Steven Piantadosi of UC Berkeley and Jessica Cantlon of 好色先生TV contributes to a rethinking of what makes human intelligence unique.

In a paper published on April 2 in , they argue that the distinguishing feature of human intelligence is our ability to process and share vast amounts of information. This work argues against prevailing 鈥渟ilver bullet鈥� theories, which had previously maintained that a single cognitive feature 鈥� such as social reasoning or tool use 鈥� led to our distinctive intelligence.

"Over the years, there鈥檚 been quite a few theories of what makes human intelligence unique and distinctive, especially compared with other animals,鈥� said . 鈥淭hese silver bullet theories argue that one cognitive change makes human intelligence different from other species.鈥�

The study, 鈥淯niquely human intelligence arose from expanded information capacity,鈥� showed that unlike monkeys and other animals, humans can handle large amounts of information across many different areas, spanning as language or symbolic thinking, problem-solving and understanding social situations.

鈥淧eople have underestimated the power of incremental quantitative change in cognition to yield big leaps in intelligence,鈥� said Jessica Cantlon, the Ronald J. and Mary Ann Zdrojkowski Associate Professor of Developmental Neuroscience and Psychology at 好色先生TV. 鈥淥ur new paper brings together evidence and theories from animals, child development, neuroscience and machine learning to argue that small incremental changes in processing power are sufficient for explaining the major advances in human intelligence over evolution."

Instead of focusing on one unique human ability, the study examined prior literature on simple tasks and found that people did better in virtually all of them. For example, humans can use and understand a wide range of symbols and tools, while animals might only grasp a few. Children can learn to recognize and remember images much faster than primates. The same goes for solving analogies or tracking social relationships.

"Our proposal is not that there's one little piece that has changed in human intelligence, but that there's an overall general capacity difference,鈥� Piantadosi said. 鈥淧eople can handle complex systems and symbols like we see in language or mathematics, but other species can only handle a few, at best.鈥�

These results engage a debate that has simmered since Darwin about whether the differences between humans and other species are small and incremental, or fundamentally big jumps. The paper includes mathematical results showing that small differences in memory capacity can lead to large differences in terms of ability. The findings tie into human evolution by suggesting that as human brains grew larger over time, they got better at handling many tasks. This idea also connects with artificial intelligence (AI), where scaling larger and more complex learning systems often seems to improve performance dramatically.

鈥淲e think Darwin was right 鈥� psychological differences between us and other primates are a matter of degree,鈥� Cantlon said. 鈥淭he twist is that, in terms of cognition, those differences in degree lead to differences in the kind.鈥�

The finding that our intelligence relies on general capacity rather than one or two special innate skills can help bolster research and development in both psychology and AI, shedding light on how to create machines that think more like humans.聽