ML models doesn't have fight or flight, so we'll have to show them tornado and teach to run away.
34 karma · joined August 7, 2022
If for safety, make it an opt-out feature, so the ones who know what they're doing can disable it.
Mandatory locking down is not for safety but for corporate control.
The state is the persona of the AGI and reinforcement learning is for the neuroplasticity, to learn, unlearn and evolve
Check [nimo](https://www.nimoplanet.com/), a start-up working in the same segment, i've tried on their product and it does have a virtual screen fixed to position and it allows upto six 1080p. screens at a time pinned at any angle.