STORY · MODELLER_
Open and closed models—the large performance gap that nobody quite talks about correctly
Nathan Lambert analyzes why proprietary AI models from OpenAI and Anthropic still outperform open models on standard benchmarks, and what the complexity in these evaluations actually obscures. He discusses how this gap is likely to shift as evaluation methodology and open source models mature.
WHY IT MATTERS
This affects investment decisions, career choices, and strategy for companies evaluating how much they can rely on open models versus proprietary solutions. Understanding what a benchmark *actually* measures is critical as the industry decides which tools to build on.
SOURCES
MACHINE-GENERATED SUMMARY This summary is written by machine from the sources below. We sort and explain — but we are a way into the field, not the final word. Check the source when something matters to you.