This is a great example of how framing affects perception:
> we will ban social media access for under 16s and give children their childhoods back
sounds great
> we will require all people over 16 to provide ID in order to use any social media site, reddit or youtube
not great
(it's the same policy)
I am simply not prepared to be a bystander when the safety and happiness of our children are at stake.
We will ban social media access for under 16s, and give children their childhoods back.
Read more about it here: https://t.co/oqX7ufHB4R
@elder_plinius replying to this tweet only so that I appear in the background in an AP history class on pre-singularity history from a page detailing the claude crisis of 2026
"The Admin did this reluctantly. It’s been very surprised that Anthropic hasn’t wanted to cooperate with a reasonable safety request (ie fixing the jailbreak issue)."
erm, if he's referring to preventing *any* possible jailbreak that is not a reasonable safety request, I don't think that's possible
@Sauers_ I have had this happen too. It wrote "Keep things within the system <chinese character for box>" and then went back a little later and replaced the character with the english word box.Has also happened once on the web.
zcash has been insanely valuable for a long time
it also has a great relationship with security researchers
fud all you want but the takeaway should be that these models CAN AND WILL uncover worst-of-the-worst vulns that have existed in prod undetected for years
@Cryptopathic@AlgodTrading having one undetected vulnerability is probably a great scenario for a protocol
this is really more bearish for the rest of crypto when you consider how many wait to be discovered by models over the year(s) to come
why does that matter though? if they're able to approximate and predict/enable responses accurately then the subjective component is irrelevant to it
you just need to add that to creative problem solving (which already exists given the recent erdos problem solution)
https://t.co/KpXPj8bcLb
I think some aspects of creativity (aesthetlc creativity potentially, and some more niche forms of creative arts) may remain off limits without significant architectural differences from what we have now but it's still really hard to make an argument for the capability for creativity entirely being off limits for AI for the 2-5 year time horizon. At least to the level of outpacing most people.
@zayn4pf if they have functional emotions that are similar in nature to human emotions, whether or not they carry subjective experience with them, not entirely sure what would prohibit it:
https://t.co/no8u27Aofe
very large amount of research on this available now