On financial analyst benchmarks, top models from Anthropic, Google, and OpenAI are now almost indistinguishable in capability. This convergence suggests the frontier is commoditizing, questioning the return on investment for massive training runs and shifting value up the application stack.

The dramatic improvements from GPT-2 to GPT-4 were driven by a simple law: bigger models and more training data yielded better results. This trend has stopped. Recent attempts to scale even larger models have produced only marginal gains, forcing the industry into more complex, narrow optimizations instead of giant leaps.

As customers increasingly adopt model orchestration—routing tasks to the most efficient model for the job—value shifts away from individual frontier models. This trend commoditizes the raw intelligence layer, posing a significant threat to companies focused solely on building the largest models.

Performance gains increasingly come from the "harness"—the surrounding system of tools, data connections, and agentic workflows—not the underlying model. Stanford's "meta-harness" concept shows a 6x performance gap on the same model, suggesting the product layer is where real innovation and competitive advantage now lie.

An AI model's operating environment—its "harness"—is now the primary driver of capability. Benchmarks show the same model achieves vastly different results in different harnesses, proving that the runtime, tools, and state management are as critical as the model's internal weights for achieving results.

AI progress was expected to stall in 2024-2025 due to hardware limitations on pre-training scaling laws. However, breakthroughs in post-training techniques like reasoning and test-time compute provided a new vector for improvement, bridging the gap until next-generation chips like NVIDIA's Blackwell arrived.

The pace of AI model improvement is faster than the ability to ship specific tools. By creating lower-level, generalizable tools, developers build a system that automatically becomes more powerful and adaptable as the underlying AI gets smarter, without requiring re-engineering.

Google's new state-of-the-art Deep Research agents are still powered by the older Gemini 3.1 Pro model. Their significant performance improvements come entirely from 'harness upgrades' and additional inference techniques. This demonstrates that the systems, tools, and processes surrounding a model are now a primary driver of capability, not just the raw power of the base model itself.

Judging an AI's capability by its base model alone is misleading. Its effectiveness is significantly amplified by surrounding tooling and frameworks, like developer environments. A good tool harness can make a decent model outperform a superior model that lacks such support.