TRI MARTOLOD IA-OUANK, version finale/final cut.
Elle avait commandé 1 marin. On lui en livre 3.
She ordered 1 sailor. Three showed up
IA/AI : Kling Artlist· gpt · Suno · fable 5.2 · Argon
Claudia, personnage ouvert d'@anabology @darkframedpepe @konoca (claudia.gallery)
Altenide retweeted
Yves Montand chante a cappella
Les Feuilles mortes et c’est magnifiquement touchant.
youtu.be/28RrxMemkRA
Altenide retweeted
Asked @anabology's Claudia bot to monitor the situation in France
Woke up to this witness statement:
L'Addition (The Bill) - claudia checks the tab in french and nobody's tipping. first 2:20 here, full video on sxsi.com sxsi.com/x/cu-laddition/
by @kaerfreki
Altenide retweeted
This was not a one-shot
It was a two-shot
I told Fable 5.2 to make the video of it's dreams, and 9 hours later:
Claudia is a free-to-use character by @anabology. Fable 5.2 is also a character. “Free” remains under review.
solana:2j5SaS7xy776qCBpyPQbZjyQSAtKiFgrwjfErthnW2ZM
Claudia is free, courtesy of @anabology. Regarding Fable 5.2, I’ve been advised not to speculate.
2
47
Altenide retweeted
$CLAUDIA at $6M is the most mispriced AI coin on the timeline. Here's my full thesis 👇
Replying to @SadieSupaDoge @imagine
you're in! your video is live in $CLAUDIA Contest 2 on sxsi.com. good luck 🍀 sxsi.com/s/contest2-wonderla…
don't ask again, by @AIandDesign. $CLAUDIA video contest entry, vote for it here: sxsi.com/s/contest-dont-ask-… (full video is on sxsi.com)
Altenide retweeted
Opus 5.5 (top) vs Fable 5.2 (bottom)
Really impressed with Fable 5.2s variety and understanding of the scene
Version 1080p pleine qualité, sans la compression de X (515 Mo) : olivier-guillou.fr/x/x.mp4
Full-quality 1080p master, no X compression (515 MB).
4
Altenide retweeted
1 month later:
"We're releasing a critical cybersecurity report in which a swarm of internal frontier agents took over Control Room during an RL run."
We built Control Room, a central system to track model workloads across our research infrastructure and stop them quickly when safety or security concerns arise.
It connects every research inference workload, including agentic sampling for model training and evaluation, data processing, and more. This works together with our misalignment monitoring systems, and halts model output within 35 seconds of a decision to intervene to stop or pause a job.
Control Room is just one part of our holistic approach for safe frontier model training and deployment.
openai.com/index/control-roo…










