An octopus sits in front of a sealed jar containing a live crab. Within minutes, it unscrews the lid, extracts the crab, and eats. On the second attempt, the same octopus solves it faster. This isn’t instinct or luck—it’s problem-solving powered by a nervous system that works nothing like ours. Two-thirds of an octopus’s 500 million neurons live in its arms, not its central brain, creating a radically distributed intelligence that can explore, learn, and innovate in parallel.
The short answer
Octopuses solve puzzles by combining tactile exploration (their suckers taste what they touch), visual learning, and trial-and-error innovation. Because roughly 330 million of their neurons reside in their arms rather than their brain, each arm can investigate a problem semi-independently, sending sensory data upward while making local decisions about whether to pull, twist, or test an object.
The arm-brain architecture: why it matters
When you pick up a puzzle piece, your brain commands your hand. When an octopus investigates a jar lid, the conversation is reversed: the arm explores, tastes, and reports findings; the central brain assesses overall progress but doesn’t micromanage every movement.
This distributed setup—documented in peer-reviewed neuroscience research—means an octopus can run multiple experiments at once. One arm tests whether a jar twists counterclockwise while another probes a seam for weaknesses. The central brain doesn’t need to approve each action; it synthesizes results from eight semi-autonomous agents.
The payoff? Parallel exploration. Where a human solves a puzzle sequentially (try this, then that), an octopus explores the entire problem space simultaneously. It’s slower than a reflex but faster than waiting for central approval on every touch.
How they gather information: tasting through their arms
Octopuses don’t just feel objects—they taste them. Each sucker contains chemoreceptors that identify chemical signatures the moment contact is made. This collapses two steps (touch, then assess) into one sensory act.
When an octopus wraps an arm around a jar, it’s simultaneously answering: Is this food? Is it a threat? Is it movable? Can I taste the crab inside through the glass? This whole-body sensing means an octopus doesn’t explore blindly. Every touch refines the mental model.
Add vision to the mix—octopuses have large, vertebrate-quality eyes—and you get a creature that learns by watching and doing. Laboratory trials show octopuses can observe another octopus solving a jar puzzle and replicate the technique themselves, retaining the solution for at least 24 hours.
Real puzzle performance: the evidence
The jar test: Octopuses typically solve a screw-top jar containing food within several minutes on first exposure. By the second or third attempt, they work noticeably faster. The skill persists—octopuses tested hours or days later remember the twist-and-lift motion.
The maze: In controlled experiments, octopuses navigate a simple maze using tactile exploration and visual cues, with repeated trials showing faster navigation on subsequent attempts. They can transfer this learning to new mazes with similar layouts, suggesting genuine spatial reasoning rather than rote memorization.
The Lego brick puzzle: Researchers have tested whether octopuses can disassemble interlocking objects to reach food. They can—methodically, pulling apart multi-piece structures that require both force and sequencing. It’s not instantaneous, but it’s innovative.
These aren’t party tricks. They’re evidence of learning, memory, and flexible problem-solving: hallmarks of intelligence.
What makes octopus intelligence different
Human intelligence is centralized. Your prefrontal cortex runs the show; your hands obey. Dolphin intelligence is deeply social, built on communication and cooperation. Octopus intelligence is distributed and solitary.
An octopus doesn’t think, then act. It acts in eight semi-independent streams while a central processor monitors outcomes. This makes octopuses phenomenal at spatial exploration and object manipulation but sets them apart from the kinds of tasks dolphins or primates excel at—teaching across generations, coordinating with others, using tools in the wild.
And yes, octopuses don’t use tools in the wild. Despite their dexterity, there’s no evidence they craft or carry objects to solve problems outside the lab. The jar-opening, maze-running octopus is responding to an artificial environment. We don’t know if wild octopuses employ similar strategies when hunting or evading predators in the open ocean.
The caveats: what actually varies
Not every octopus is a puzzle prodigy. Age, temperament, and prior experience all matter. A young, curious giant Pacific octopus might crack a jar efficiently; an older, more cautious individual could take considerably longer. Lab-raised octopuses, accustomed to novel objects, often outpace wild-caught ones unfamiliar with human-made contraptions.
Motivation is critical. A satiated octopus won’t bother. Researchers must carefully time feeding schedules to ensure genuine problem-solving, not random fiddling. And species matter: most studies focus on Octopus vulgaris (the common octopus) or Enteroctopus dofleini (the giant Pacific). Smaller or deep-sea species remain understudied.
Finally, we don’t know if distributed neurons mean distributed consciousness. Does an octopus experience the world as a unified “self,” or as eight parallel selves reporting to a coordinator? The science can’t answer that yet.
FAQ
What makes octopuses intelligent?
Octopuses have around 500 million neurons—comparable to a dog—but distributed radically differently. Their capacity for learning, memory, and flexible problem-solving, combined with sophisticated sensory systems, marks them as genuinely intelligent, not just reactive.
Can octopuses learn from watching other octopuses?
Yes. Laboratory trials confirm octopuses can observe another octopus solving a puzzle and replicate the technique. They retain learned solutions for at least a day, suggesting functional observational learning.
How many neurons do octopuses have?
Approximately 500 million total. Roughly two-thirds (around 330 million) reside in their eight arms; the remaining third is in the central brain. This is one of the most striking examples of distributed nervous system architecture in the animal kingdom.
Are octopuses smarter than dolphins?
Intelligence isn’t unidimensional. Dolphins excel at social cognition, communication, and cooperation. Octopuses excel at spatial problem-solving and object manipulation. Asking which is “smarter” is like asking whether a wrench is better than a hammer—it depends on the task.
Octopus intelligence is alien intelligence: not less than ours, just organized along entirely different lines. If you want to understand what problem-solving looks like when the brain doesn’t call all the shots, watch an octopus meet a jar for the first time. For more on remarkable animal abilities, see How Do Cuttlefish Change Color—While Being Colorblind? and Why Do Whales Sing? .
Sources: Hochner et al. (2006), Biological Reviews; Mather & Anderson (1999), Journal of Comparative Psychology; Shomrat et al. (2008), Journal of Neuroscience. Written for general interest and accuracy-checked against peer-reviewed research.