An AI's ability to code complex games and physics simulations is a strong indicator of its overall power. This showcases its deep understanding and ability to handle sophisticated, multi-layered logic required for complex business applications, not just simple tasks.

AI models show impressive performance on evaluation benchmarks but underwhelm in real-world applications. This gap exists because researchers, focused on evals, create reinforcement learning (RL) environments that mirror test tasks. This leads to narrow intelligence that doesn't generalize, a form of human-driven reward hacking.

Static benchmarks are easily gamed. Dynamic environments like the game Diplomacy force models to negotiate, strategize, and even lie, offering a richer, more realistic evaluation of their capabilities beyond pure performance metrics like reasoning or coding.

While language models are becoming incrementally better at conversation, the next significant leap in AI is defined by multimodal understanding and the ability to perform tasks, such as navigating websites. This shift from conversational prowess to agentic action marks the new frontier for a true "step change" in AI capabilities.

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.

As reinforcement learning (RL) techniques mature, the core challenge shifts from the algorithm to the problem definition. The competitive moat for AI companies will be their ability to create high-fidelity environments and benchmarks that accurately represent complex, real-world tasks, effectively teaching the AI what matters.

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.

OpenAI's new GDP-val benchmark evaluates models on complex, real-world knowledge work tasks, not abstract IQ tests. This pivot signifies that the true measure of AI progress is now its ability to perform economically valuable human jobs, making performance metrics directly comparable to professional output.

Standardized AI benchmarks are saturated and becoming less relevant for real-world use cases. The true measure of a model's improvement is now found in custom, internal evaluations (evals) created by application-layer companies. Progress for a legal AI tool, for example, is a more meaningful indicator than a generic test score.