> ## Content Index > Fetch the complete content index at: https://adjacent.media/llms.txt > Use this file to discover other available public pages before exploring further. # // Friday Five — September 11 - URL: https://adjacent.media/five/2026-09-11-five/ - Published: 2026-09-11T19:30:03.000Z - Updated: 2026-09-11T19:30:03.000Z - Description: This week made the boundary explicit: AI crosses capability thresholds wherever a machine can check the answer. - Author: Jonathan Greene - Tags: #friday-five, theme-ai This week made the boundary explicit: AI crosses capability thresholds wherever a machine can check the answer. Formal proofs of Navier-Stokes singularities, Millennium Prize claims traded between labs, a genome atlas enumerating nine billion variants, a model driving Blender — each runs on a verifier, not on judgment. The uncomfortable corollary, stated plainly by Paul Christiano, is that control is the one domain with no checker, and the industry is not building one. Scout's Pick — Outlier AI ### [The checkable frontier is wider than anyone mapped it](https://adjacent.media/signals/when-ai-labs-race-to-solve-maths-hardest-problems/) The verifier thesis cuts both ways: once you notice how much hard intellectual work is machine-checkable, the unverifiable residue looks like the exception, not the rule. Source: [Adjacent](https://adjacent.media/signals/when-ai-labs-race-to-solve-maths-hardest-problems/) AI ### [Christiano says the risk curve is not bending](https://adjacent.media/signals/openai-board-member-warns-ai-industry-unprepared-for-control-risks/) Control is the one capability with no oracle to grade it, and the board member closest to the question says nobody is building the grader. Source: [Adjacent](https://adjacent.media/signals/openai-board-member-warns-ai-industry-unprepared-for-control-risks/) AI ### [Pretraining gains now trace back to data, not scale](https://adjacent.media/signals/data-scaling-not-model-architecture-drives-ai-progress/) Progress follows the corpus, which means it follows whatever humans have already written down and checked. The unwritten domains get no such lift. Source: [Adjacent](https://adjacent.media/signals/data-scaling-not-model-architecture-drives-ai-progress/) AI ### [A model drives Blender by watching the screen](https://adjacent.media/signals/computer-use-emerges-as-next-ai-demand-wave-after-chatbots-and-reasoning/) Rendered output verifies itself. The model can see whether the mesh looks right and iterates without human intervention. Source: [Adjacent](https://adjacent.media/signals/computer-use-emerges-as-next-ai-demand-wave-after-chatbots-and-reasoning/) AI ### [Nine billion DNA variants enumerated and scored in advance](https://adjacent.media/signals/deepmind-maps-9-billion-human-genetic-variants-in-single-database/) Exhaustive enumeration works because each prediction lands against a measurable biological signal. The assay does the grading, not judgment. Source: [Adjacent](https://adjacent.media/signals/deepmind-maps-9-billion-human-genetic-variants-in-single-database/) 6 themes · 159 signals · 51 sources