![rw-book-cover](https://substackcdn.com/image/fetch/$s_!T89d!,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F943bc4ac-da7d-486f-887f-b6eb85f47b53_1268x685.png) ## Metadata - Author: [[Benn Stancil]] - Full Title:: Get Out of the Token Path - Category:: #🗞️Articles - URL:: https://benn.substack.com/p/get-out-of-the-token-path - Read date:: [[2026-06-07]] ## Highlights > Second, despite their relative simplicity, agents can still be valuable products. According to a [research paper from Meta](https://arxiv.org/pdf/2603.28052), “changing the harness around a fixed large language model can produce a 6x performance gap on the same benchmark.” Though a small startup can’t outbuild Anthropic’s or OpenAI’s models, it can build a harness that *improves* their models. ([View Highlight](https://read.readwise.io/read/01ktezr7hgq8sgq23j9qnd09rk)) > It is 2026. [It is the beginning of the harness era.](https://tomtunguz.com/harnessing-ai/) Build an agent. ([View Highlight](https://read.readwise.io/read/01ktezsc5k1f0c4mjr32t3jnxy)) > so they built claude code, and they priced it to win. … once a lab decides a profit pool is theirs, they burn the surrounding economics down until the people sitting on it leave on their own. ([View Highlight](https://read.readwise.io/read/01ktf010y4nj9wacbs4sh22195)) > Specialized agents might work at first, but they do not last. Eventually, a powerful model running inside a general-purpose harness that is designed alongside that model will catch up. ([View Highlight](https://read.readwise.io/read/01ktf58q4xkcz5w67hyvftgwn5))