Evolution of IQ: Data from 200M BC to 2026

In this visualization I track the evolution of animal intelligence from the Jurassic Period (~200 million years ago) to 2026, ranking the most cognitively capable organisms across deep evolutionary time, including extinct species, modern animals, and the emergence of machine intelligence. Methodology & Data Sources: Because conventional IQ tests are anthropocentric and designed exclusively for humans, in this visualization I use an Equivalent IQ (eIQ) scale. Modern adult Homo sapiens is anchored at 100 eIQ, providing a common reference point for cross-species comparison. eIQ is a comparative cognitive index derived from three variables: 1. Neuron Count - estimated cortical or pallial neurons as a proxy for information processing capacity. 2. Encephalization Quotient (EQ) - brain size relative to expected brain size for body mass, providing an indicator of cognitive investment and behavioral flexibility. 3. Behavioral Complexity - documented evidence of tool use, problem-solving, communication, social learning, cooperation, and other cognitively demanding behaviors. The weighting system is designed to reduce the distortions produced by relying on any single measure. In particular, raw brain size alone can systematically favor large-bodied animals while underestimating highly neuron-dense species such as corvids and parrots. Primary Data Pools: Cognitive estimates for extinct taxa are reconstructed primarily from fossil endocasts and cranial morphology. CT-based reconstructions provide estimates of brain volume, relative development of major brain regions, and olfactory structures. Relevant data were derived from paleontological literature, the Smithsonian National Museum of Natural History, the Journal of Paleontology, and the Journal of Human Evolution. Modern animal cognition is evaluated using comparative neuroanatomical and behavioral datasets. Particular emphasis is placed on neuronal density and absolute neuron counts rather than brain mass alone. Machine intelligence is treated as a separate computational domain rather than a biological species. Historical and modern systems, including early neural networks and contemporary large language models, are positioned using performance on standardized cognitive benchmarks, computational scaling, model capability evaluations, and published research on AI progress. Important Note The eIQ values should be interpreted as comparative estimates, not literal IQ scores. Intelligence is multidimensional, species-specific, and highly dependent on ecological context. ***** Hi, I'm Sasha. I crunch numbers, play with data, and create cool visuals. If you enjoy my work, a little support can get me a coffee and a cookie for my baby girl Eva ☕🍪 https://www.paypal.com/paypalme/datai... If you want to connect with me:   / sasha-medin-07270234   My main data project: https://fivehundred.co/40-years-of-ca...