GigaQian
Region: US
Sunday 14 June 2026 23:50:35 GMT
17124
1569
93
190
Music
Download
Comments
Tom :
https://github.com/elder-plinius/CL4R1T4S/blob/main/ANTHROPIC/CLAUDE-FABLE-5.md
2026-06-15 00:04:54
66
Darrel123 :
The government didn’t do this because it was security threat. They did it out of spite because of the problems they had with anthropic earlier in the year.
2026-06-15 03:26:12
17
XXIVandMe :
Soo fable 😂😂😂😂
2026-06-15 06:50:03
0
Mango :
i missed the chance to use Fable 5... it was there but i didnt touch it yet until it wasnt there 😭
2026-06-15 00:41:45
13
boorman :
Prompt Mining
2026-06-15 05:30:05
1
Rob :
Remember when Bill Clinton blocked any encryption stronger than 40 bits. Yeah they learned the hard way with that one. Looks like Trump has to also learn the hard way. Time for Asia and Europe to step up and fill the gap
2026-06-15 03:00:22
6
davidedwards6916 :
Your just a huge fountain of knowledge and I absolutely love it !!👍👍👍💪💪💪😂😂😂
2026-06-14 23:56:30
13
MORRDOK :
bro THIS is my favorite Tik Tok account
2026-06-15 01:55:06
5
Split Wood :
Claritas??
2026-06-15 02:32:38
4
katie.gallery :
This is my favorite post about this thank you
2026-06-15 00:57:48
3
Moleta :
We know these systems are all prompt 😏
2026-06-15 02:29:09
3
Aaron Litherland :
It's inevitable that any model that comes out will have vulnerabilities exposed. Prompting is definitely part of it or skills wouldn't be so valuable. Of course, gaurdrails is the biggest thing. Eventually, we will have local models just as powerful and I believe it's closer than we are told.
2026-06-15 00:30:03
3
virtualuman_ :
how do we implement?
2026-06-15 01:20:21
1
Bumi Kita :
whole world should follow your post.
2026-06-15 01:44:22
1
rojas165 :
ON IT.
2026-06-15 02:45:33
1
randy :
There’s a quantum study referring to how their journal was just an extension of their mind….we have a “living journal” 😁💯
2026-06-15 02:27:32
1
Kwafuraha :
Oh, snap! Nothing goes unnoticed online.
2026-06-15 02:03:34
1
NoctSerif :
Was genuinely impressed with what it could do in the short window we had. The prompt-as-talent argument is interesting though and I’ve always leaned that direction partially. But I’d put that 90% recovery claim firmly in flavor and vibes territory. The instructions we give are a capability multiplier for sure but they’re not the model. You can’t prompt your way to capability that isn’t there, unless Anthropic dumped latent capability from fable into opus. What it does tell us is how much our own training of these models matters and how much we aren’t fully appreciated or compensated for what we’re contributing every time we interact with them. We are the dataset whether we realize it or not and paying $100/month to get real usage out of something we first hand help create or very least teach how to walk,talk & run seems a bit greedy but I guess not by American standards 😆 The echo chamber problem is the real one though. Good input, good output. Bad input, confidently delivered garbage. The model doesn’t know the difference and neither do most people using it. As for whether these systems represent the beginning of a new consciousness or just our projection of one onto a very sophisticated mirror, I genuinely don’t know. But I think the fact that we can’t tell yet is probably the most telling detail about AI as a whole. We’re either witnessing the first flicker of something new or we’re just very good at seeing faces in the cloud. Either way it’s pretty interesting.
2026-06-15 05:45:22
0
𝑹𝒐𝒃𝒃𝒊𝒆 :
it’s not actually a leak they post all of their system prompts publically https://platform.claude.com/docs/en/release-notes/system-prompts
2026-06-15 00:32:34
7
Lrrr 🔥🔥 :
So, how would I use the fable 5 prompt? Let's say I'm using opus 4.8, would I just ass the prompt to the CLAUDE.md file? or create it into a skill that it would need to invoke everytime I start a new session?
2026-06-15 01:34:40
1
CARBON6 :
Government is overstepping
2026-06-15 02:18:10
2
Silver :
I just said this today to another engineer. The power of AI is in the prompt. It can't read minds.
2026-06-15 01:50:43
1
DrJanetteCamacho :
I am off to experiment!! Thank you 🙏
2026-06-15 00:33:45
2
Katie A :
I wonder if the instructions can be placed into sonnet to get the same experience. I just did that but I can't figure out if it responded as sonnet or if it responded as Fable. I had the opportunity to use Fable before it was disband but the response is looking very sonnet, but the level of detail is looking very fable.
2026-06-15 01:27:42
2
truckminster fuller :
the unsaid part here is that the hardware and software for agi are already in place, but the guardrails are hamstringing 90% of it. new models at this point are just more elaborate guardrails.
2026-06-15 01:06:00
1
To see more videos from user @gigaqian, please go to the Tikwm
homepage.
