Topic
Whether headline performance tasks measure capability that transfers to real work.
The startup fine-tuned two Nvidia Nemotron models to allow, flag or block each action an agent intends, and reports 98% accuracy on a 9,429-trajectory benchmark whose own authors say accuracy is the wrong test.
Reality
Programming a furnace is the eye-catching part, but the piece worth copying is a verification module that refuses any figure it cannot resolve to a logged result, which cuts reported fabrication to 4 percent without telling you what the number means.
Reality
The second World Humanoid Robot Games cut the machines' 100-metre time from 21.5 seconds to 8.64 in a year, while the dexterity events failed even with human operators in control gloves. That puts the bottleneck in the hands.
Reality
Alex Palcuie says he now goes to the model first when Claude pages him, and that the answer to whether Claude fixes its own incidents is no. His evidence is his team's hiring plan.
Reality
A Red Hat post argues top-ranked agents routinely leave business metrics flat, and that swapping the evaluation harness can move rankings more than swapping the model does.
Reality
A replay that held the agent's actions fixed produced different labels before and after delayed operations resolved, and one late write moved the following run's score.
Reality