Unlock all features
FREE: Get instant access to 10 AI summaries, chats, or transcripts per day.
Unlock all features
FREE: Get instant access to 10 AI summaries, chats, or transcripts per day.
Unlock all features
FREE: Get instant access to 10 AI summaries, chats, or transcripts per day.
Unlock all features
FREE: Get instant access to 10 AI summaries, chats, or transcripts per day.
Unlock all features
FREE: Get instant access to 10 AI summaries, chats, or transcripts per day.
Related videos
it's all bad now...
Wes Roth
69.7k views
Claude Fable 5 is BANNED. What to do?
Greg Isenberg
146.1k views
Claude Fable JUST got BANNED...
Wes Roth
50.2k views
Fable JUST made EVERYONE MAD...
Wes Roth
52.2k views
Why Were These Toys Banned?
Good Mythical Morning
488.1k views
everyone JUST got HACKED...
Wes Roth
50.2k views
American's Are Actually Rooting For Iran, Here's Why
Kim Iversen
50.7k views
CLAUDE JUST GOT BANNED
Wes Roth
57.3k views
$1 Trillion Gone
Wes Roth
43.9k views
it JUST happened
Wes Roth
79.0k views
Top Comments (10)
a 5 min sponsor is kinda crazy
Thatβs was one LOOOOOONG product placement.
Wes donβt do this shit again man. Have some shame. 5 min ad? Jesus dude
5 minute sponsor segment? 20% advertisement?
Here's what I gathered from data logging fable 5 Fable-5 works by keeping the entire problem in context at once. Its core is an enormous effective memory: it stays coherent across millions of tokens, so the whole system lives in its head as one picture instead of being read in fragments. That's what stops it making a change that's right in one spot but wrong everywhere else. On top of that, three things gave it its edge: It cross-breeds instead of copying. Rather than latching onto the first source or one approach, it pulls from everything it has loaded and combines across them, so its output isn't a remix of a single template. It doesn't trust its own gut. Before committing to a direction it tests it against a bar it sets in advance, and only keeps what clears it, so good-sounding-but-wrong ideas get filtered out instead of shipped. It runs itself from written memory. It offloads state into files and keeps re-deriving its next step from the end goal, which is why it can self-drive coherently for days instead of drifting or stopping. So the secret sauce is: hold everything at once, cross-breed across it, test before you believe, and self-drive from persistent memory toward the goal.
To me it would make sense that if Amazon found the jail break that at some point early on, Anthropic would know what the jail break is. The fact that they said that they didn't know what it was is a pretty good indication that there is something fishy. Add to that that Amazon told the WH not Anthropic, their business partner, also tells me something fishy is going on.
bro 5 minute product sponsor. No bro.
Try Genspark here: π https://www.genspark.ai/?utm_source=yt&utm_campaign=WesRoth02 Use the above link to get free credits in Genspark. Genspark is an all-in-one AI workspace that brings chat, spreadsheets, slide creation, design, and automation into one place. Instead of bouncing between different apps for each step of a project, you can use Genspark to research, create, organize, and finish the whole workflow in a single platform. Paid users also get unlimited AI chat and AI image generation in 2026, subject to abuse guardrails. #Genspark #WorkWithGenspark
Almost seems like shutting the barn door after the cows are gone
I skipped ahead five minutes and started watching again. Little did I know that it was still the damn ad!
Unlock the Data Inside
Turn Videos into Knowledge
- Get FREE 10/day: transcripts, summaries, chats
- Chat with videos, export text & PDF
- $1 free API credit for RAG, chatbots & research
Free forever plan β’ All features unlocked
Top Comments (10)
a 5 min sponsor is kinda crazy
Thatβs was one LOOOOOONG product placement.
Wes donβt do this shit again man. Have some shame. 5 min ad? Jesus dude
5 minute sponsor segment? 20% advertisement?
Here's what I gathered from data logging fable 5 Fable-5 works by keeping the entire problem in context at once. Its core is an enormous effective memory: it stays coherent across millions of tokens, so the whole system lives in its head as one picture instead of being read in fragments. That's what stops it making a change that's right in one spot but wrong everywhere else. On top of that, three things gave it its edge: It cross-breeds instead of copying. Rather than latching onto the first source or one approach, it pulls from everything it has loaded and combines across them, so its output isn't a remix of a single template. It doesn't trust its own gut. Before committing to a direction it tests it against a bar it sets in advance, and only keeps what clears it, so good-sounding-but-wrong ideas get filtered out instead of shipped. It runs itself from written memory. It offloads state into files and keeps re-deriving its next step from the end goal, which is why it can self-drive coherently for days instead of drifting or stopping. So the secret sauce is: hold everything at once, cross-breed across it, test before you believe, and self-drive from persistent memory toward the goal.
To me it would make sense that if Amazon found the jail break that at some point early on, Anthropic would know what the jail break is. The fact that they said that they didn't know what it was is a pretty good indication that there is something fishy. Add to that that Amazon told the WH not Anthropic, their business partner, also tells me something fishy is going on.
bro 5 minute product sponsor. No bro.
Try Genspark here: π https://www.genspark.ai/?utm_source=yt&utm_campaign=WesRoth02 Use the above link to get free credits in Genspark. Genspark is an all-in-one AI workspace that brings chat, spreadsheets, slide creation, design, and automation into one place. Instead of bouncing between different apps for each step of a project, you can use Genspark to research, create, organize, and finish the whole workflow in a single platform. Paid users also get unlimited AI chat and AI image generation in 2026, subject to abuse guardrails. #Genspark #WorkWithGenspark
Almost seems like shutting the barn door after the cows are gone
I skipped ahead five minutes and started watching again. Little did I know that it was still the damn ad!