AI IQ is here: a new site scores frontier AI models on the human IQ scale. The results are already dividing tech.
A startup called AI IQ is quantifying the intelligence of over 50 AI language models by assigning them IQ scores based on a composite of various reasoning benchmarks, generating interactive visualizations that have sparked both praise and criticism for oversimplifying complex capabilities into a single number. Critics argue that this approach can mislead by masking the "jagged" nature of AI capabilities, while supporters find it a useful tool for comparing model performance in a rapidly evolving landscape.
For enterprise buyers and AI professionals, the most actionable takeaway from the AI IQ project is the "IQ vs. Effective Cost" scatter plot, which reveals that the highest IQ models are not always the most cost-effective. This chart suggests that strategically routing tasks—using high-cost models for complex problems and lower-cost models for simpler tasks—can optimize AI deployment efficiency. This insight is critical for CIOs aiming to balance performance with budget constraints in AI strategy.