KM4BI

Astra Reactions, Jobs Report and Cybercab Launch

Astra shifts the AI conversation toward real tasks, as US jobs surprise and Tesla sets out its Cybercab ambitions.

Published

Diet TBPN connects three signals: Astra GPT-6's Blender demonstrations, a stronger-than-expected US jobs report, and Tesla's Cybercab ambitions. The common thread is moving from claims to complete tasks and measurable delivery.

AI should be judged in real tools

The hosts focus on demos in which Astra builds 3D scenes, real-estate walkthroughs, and interactive experiences in Blender. They stress that exceptional output still needs orchestration: iterations, a manager agent, and sub-agents.

In their view, saturated benchmarks are becoming less informative. A more useful test is to give a model a task you know deeply, then assess the result, its taste, and its efficiency. The relevant metric becomes cost per completed task rather than cost per token.

Employment: a positive surprise

The report discussed shows 162,000 US jobs added in August versus 53,000 expected, with unemployment at 4.1%. Restaurants, local education, manufacturing, and healthcare contributed, while information and finance declined.

Cybercab: timing is the test

Tesla is targeting a steering-wheel-free Cybercab below $30,000 and cites roughly 20 cents per mile in operating cost before taxes. The hosts focus on the decisive issues: actual deliveries, usable autonomy, and economic viability—not pre-orders or promises.

Source