Sarah Guo is flagging a genuinely interesting technical result: scaling laws for hybrid architectures, which until now lacked the predictable training behavior that made transformers so easy to bet on. If h-nets really scale efficiently — with learned tokenization as a key lever — that is a credible path to cheaper real-time voice and on-device models, squarely Cartesia's lane. The caveat: the source is a VC hyping the work, so 'huge compute efficiency gain' should be treated as a claim until independent benchmarks land. Still, the transformer monoculture cracking at the edges is a trend worth tracking closely.

When a Meta executive openly admits that AI negativity will only intensify, that candor is worth more than a dozen optimistic forecasts. The report captures a real shift: resentment toward AI has spread beyond creatives and young people to the broader public, and the graduation booing trend is its most visible symptom. In our view, Mosseri is being pragmatic rather than pessimistic — platforms keep inserting AI features users never asked for, so friction is guaranteed. The open question is whether companies will adjust their rollouts or simply brace for the sentiment to worsen.