Physical Intelligence API Beta: Frontier Models for Every Robot (pi.website) 1 points by modeless 15d ago ↗ HN
[–] sgvwthry 14d ago ↗ Having to make an API call to run your mission-critical robot seems like a non-starter. [–] fzysingularity 14d ago ↗ Yeah, it’s surprising that this is the runtime model they went with as opposed to a hybrid edge/cloud model. I’d love to see a system 1 on edge, system 2 on cloud offering. They’re likely the best positioned to do this. [–] sote8361 14d ago ↗ I agree, but isn‘t it basically impossible to do any kind of IP protection for LLMs, if someone dumps the disk or RAM - they have your weights.I wonder how the cloud players get around this, maybe Antrophic trusts AWS and Google not to look, but do they really just trust Musk? [–] RockRobotRock 14d ago ↗ Bring your weights in an AWS-snowball-like briefcase with an armed guard handcuffed to it.
[–] fzysingularity 14d ago ↗ Yeah, it’s surprising that this is the runtime model they went with as opposed to a hybrid edge/cloud model. I’d love to see a system 1 on edge, system 2 on cloud offering. They’re likely the best positioned to do this.
[–] sote8361 14d ago ↗ I agree, but isn‘t it basically impossible to do any kind of IP protection for LLMs, if someone dumps the disk or RAM - they have your weights.I wonder how the cloud players get around this, maybe Antrophic trusts AWS and Google not to look, but do they really just trust Musk? [–] RockRobotRock 14d ago ↗ Bring your weights in an AWS-snowball-like briefcase with an armed guard handcuffed to it.
[–] RockRobotRock 14d ago ↗ Bring your weights in an AWS-snowball-like briefcase with an armed guard handcuffed to it.
4 comments
[ 3.0 ms ] story [ 24.2 ms ] threadI wonder how the cloud players get around this, maybe Antrophic trusts AWS and Google not to look, but do they really just trust Musk?