Generic evaluation metrics like "helpfulness" or "conciseness" are vague and untrustworthy. A better approach is to first perform manual error analysis to find recurring problems (e.g., "tour scheduling failures"). Then, build specific, targeted evaluations (evals) that directly measure the frequency of these concrete issues, making metrics meaningful.

Public leaderboards like LM Arena are becoming unreliable proxies for model performance. Teams implicitly or explicitly "benchmark" by optimizing for specific test sets. The superior strategy is to focus on internal, proprietary evaluation metrics and use public benchmarks only as a final, confirmatory check, not as a primary development target.

Standard automated metrics like perplexity and loss measure a model's statistical confidence, not its ability to follow instructions. To properly evaluate a fine-tuned model, establish a curated "golden set" of evaluation samples to manually or programmatically check if the model is actually performing the desired task correctly.

The critical challenge in AI development isn't just improving a model's raw accuracy but building a system that reliably learns from its mistakes. The gap between an 85% accurate prototype and a 99% production-ready system is bridged by an infrastructure that systematically captures and recycles errors into high-quality training data.

The primary bottleneck in improving AI is no longer data or compute, but the creation of 'evals'—tests that measure a model's capabilities. These evals act as product requirement documents (PRDs) for researchers, defining what success looks like and guiding the training process.

Building a functional AI agent is just the starting point. The real work lies in developing a set of evaluations ("evals") to test if the agent consistently behaves as expected. Without quantifying failures and successes against a standard, you're just guessing, not iteratively improving the agent's performance.

When models achieve suspiciously high scores, it raises questions about benchmark integrity. Intentionally including impossible problems in benchmarks can serve as a flag to test an AI's ability to recognize unsolvable requests and refuse them, a crucial skill for real-world reliability and safety.

When selecting foundational models, engineering teams often prioritize "taste" and predictable failure patterns over raw performance. A model that fails slightly more often but in a consistent, understandable way is more valuable and easier to build robust systems around than a top-performer with erratic, hard-to-debug errors.

Instead of generic benchmarks, Superhuman tests its AI models against specific problem "dimensions" like deep search and date comprehension. It uses "canonical queries," including extreme edge cases from its CEO, to ensure high quality on tasks that matter most to demanding users.