Anyone who has read METR's Technical Report on the OpenAI/Hugging Face incident will understand that software-only controls will not be sufficient to contain superhuman AI.
Any continuing work on adversarial superhuman intelligence must rely on the physical containment methods used for nuclear security: including air gap isolation, data diodes, and a two-person rule for data exfiltration.
Moreover, the internet itself must be hardened against penetration by self-propagating AI agents: no more untraceable access to high-end AI computing facilities, and no anonymous fund aggregation or payments.
The first useful step is our Chip Security Act, currently included in the Senate NDAA.
@STLChrisH These mortgage rates also don’t exist in a magical vacuum. Rates on savings accounts/CDs were commensurately high as well, something we haven’t seen in 20 years.
Louisiana's governor is thought to be all powerful but @LAGovJeffLandry would gain even more control over state government under these 5 bills.
They give him more power over:
Civil Service
Ethics Board
Governor's mansion
Supreme Court
https://t.co/ykEU8I4otS #lalege
@dfossier@bob4districta NOLA is a mess, I agree. But I’m talking about statewide results. Republicans have done nothing to move the entire state of LA forward. That’s my only point here. They can’t just keep saying “look at the Dems” when the entire state is falling behind under their leadership.
@jacksonvoss I’m really worried about him worsening this insurance crisis in a lot of ways, but to be fair I don’t think addressing climate change is really within his remit.
@getwired Couldn’t agree more; had this discussion with our team today. I love the technology MS is developing, but frankly the licensing scheme is becoming sleazy. Some cases like with the security baseline assessment being E5 then becoming a VM add-on it is very frustrating.
@merill It would be great if things like this were made more clear within the GUI itself without having to read through various docs. Something easy like identifying what is a security control vs. just “informational” for instance.
@HJudeBoudreaux It’s truly heartbreaking to see the actual real world negative impacts these hateful, regressive policies will have on childrens’ health in our state. All done under the false pretense of “protecting children” from imaginary threats.