//
sign in
Post
by @danabra.mov
PostEmbed
by @danabra.mov
Record
by @jimpick.com
Record
by @atsui.org
+ new component
Post
3/ We're in the earliest days of attackers leveraging aggressive safety features. Wouldn't surprise me if customers handling complex cybersecurity issues demand that models be less safety-blunted. This will be increasingly hard for hyperscalers that feel growing pressure to do safetymaxxing.