The bitter lesson is finally coming for the self-driving cars. The vision stack, 3D maps, lane selection grammar, occupancy networks, it’s maybe all about to give way to a single GPT looking at camera feeds and predicting the next steering wheel adjustment.
It’s mostly a latency problem at this point. The models are too big to run locally, but given that open-weight models like Qwen already exist, an open-weight, low latency equivalent to Astra can’t be too far out.
Yeah cause every car needs 8xH200 pulling 10kW to run a VLM at realtime speeds. Would be unfortunate if 4G dropped out under some trees while using the API after all.
valine · · focus · HN ↗
It’s mostly a latency problem at this point. The models are too big to run locally, but given that open-weight models like Qwen already exist, an open-weight, low latency equivalent to Astra can’t be too far out.
moffkalast · · focus · HN ↗
Marha01 · · focus · HN ↗
When the models stop improving, we will get model-specific ASICs that are much more power-efficient.
moffkalast · · focus · HN ↗
Soo, never? Granted Cerebras is a thing, if the process can be commoditized.
At the moment the area of edge inference at speed seems pretty bleak though.