Claude Fable 5: The Most Impressive AI Model Ever

Today, we discuss Anthropic’s new model release, focusing on the balance between stronger AI capabilities and tighter safety restrictions. Let's unpack some demos of visual reasoning, gaming, and enterprise use cases, along with benchmark results and limits around biology, chemistry, and cybersecurity.

------
🌌 LIMITLESS HQ ⬇️

NEWSLETTER:    https://limitlessft.substack.com/
FOLLOW ON X:   https://x.com/LimitlessFT
SPOTIFY:             https://open.spotify.com/show/5oV29YUL8AzzwXkxEXlRMQ
APPLE:                 https://podcasts.apple.com/us/podcast/limitless-podcast/id1813210890
RSS FEED:           https://limitlessft.substack.com/

------
TIMESTAMPS

0:00 Anthropic’s New Frontier
2:24 Examples
8:04 Demos
13:58 Benchmark Stats
17:33 The Mythos Model
25:37 Pricing and Compute Limits
28:57 Long-Horizon Workflows

------
RESOURCES

Josh: https://x.com/JoshKale

Ejaaz: https://x.com/cryptopunk7213

------
Josh works with Anthropic as a contractor. All views expressed are his own and do not represent Anthropic, its leadership, or its affiliates. Nothing in this episode is investment advice.

Not financial or tax advice. See our investment disclosures here:
https://www.bankless.com/disclosures⁠

Creators and Guests

Host

Ejaaz Ahamadeen

Host

Josh Kale

What is Limitless: An AI Podcast?

Exploring the frontiers of Technology and AI

Ejaaz:
Four days ago, Anthropic asked the world's leading AI labs to slow down their

Ejaaz:
AI research out of fear that the AI models were getting so good that they would escape human control.

Ejaaz:
Now, just yesterday, the same company released the most powerful model the world

Ejaaz:
has ever seen, but it comes with a twist. It's one model, but there's two versions of it.

Ejaaz:
Claude Fable 5 is the model that everyone gets to use.

Ejaaz:
It's a class that sits above Opus, but it's heavily restricted.

Ejaaz:
It comes with a lot of safeguards around it because version 2,

Ejaaz:
called Mythos 5 is the unrestricted version which poses itself as a higher cybersecurity

Ejaaz:
risk than its predecessor, Mythos Preview. It also

Ejaaz:
excels at creating biological compounds, which could potentially be used as

Ejaaz:
bioweapons. So it's only accessible by vetted partners that Anthropic approves.

Ejaaz:
So this brings us to a fork in the road where the smartest, most intelligent

Ejaaz:
AI model isn't accessible to everyone.

Ejaaz:
Typically, for the entire history of software, you'd be able to pay for access

Ejaaz:
to the same level of access that an institutional would.

Ejaaz:
But this time, the new metric isn't how smart is your AI model,

Ejaaz:
it's the gap between how smart is your AM model that you get access to and the

Ejaaz:
ones others also get access to.

Josh:
Well, how amazing is it that we have Mythos now available in our apps today?

Josh:
I think that's the really exciting takeaway is not only is it better than the

Josh:
Mythos preview, but it's available inside of your app.

Josh:
So if you are listening to this, you can go and try CloudFable 5 right now with

Josh:
those two restrictions.

Josh:
It won't go anywhere near biology. It won't go anywhere near cybersecurity,

Josh:
but this is the best model in the world and it is now available to everyone

Josh:
to go try out and it's like every once in a while there there's this new technology that

Josh:
kind of forces you to to reframe how you engage with the technology and i think

Josh:
that's been my experience so far using fable

Josh:
is that it's so different than any other model we've used that it kind of forces

Josh:
you to reframe how you engage with them and to showcase that i want us to go through some demos of

Josh:
attempts that people have done to kind of showcase the powers and capabilities

Josh:
of fable 5 this new frontier model starting with a really bizarre demo in which

Josh:
uh dan shipper who is a prominent poster on x he recreated the library of babble

Josh:
and help me understand what's going on here you just because i'm seeing a lot

Josh:
of visuals um and i understand that it's not a image or video generating model

Josh:
so how is it able to create a real world's,

Josh:
emulation so accurate so quickly?

Ejaaz:
So the greatest part about this is it took one prompt and he basically asked

Ejaaz:
it to read a book, the book of Babel or whatever the title of the book is.

Ejaaz:
And he said, I want you to read this book.

Ejaaz:
And then I want you to recreate one of the concepts that is described throughout

Ejaaz:
the book, which is this library of Babel or Babel.

Ejaaz:
And it did so in just under an hour, I believe.

Ejaaz:
And if you look visually on the screen, it is this high fidelity 3D representation

Ejaaz:
of what this library looks like. And it's infinite. So you'll notice in this

Ejaaz:
video that he looks down, he looks up, and it's just infinite books he can access himself.

Ejaaz:
And he even asked it to include some of the essays that he's written himself.

Ejaaz:
This guy authors a bunch of analysis on AI, and he pulls open a piece that he

Ejaaz:
wrote that is in the Library of Babel. The idea of the library is that it contains

Ejaaz:
every piece of text that has been ever written. And so he gets access to it.

Ejaaz:
It's a pretty cool example.

Josh:
Yeah, that's super cool. The other ones that I've really noticed that it accelerated

Josh:
in is kind of 3D world building, which is funny because you don't think of Anthropic

Josh:
as a world building model. It's not a world model.

Josh:
In fact, it doesn't have image generation capabilities. In fact,

Josh:
it doesn't have video generation capabilities.

Josh:
So how is it creating all this realistic visual design assets?

Josh:
And it's just really good at math. And this begs the question is how important

Josh:
is it to focus on image gen on video gen if kind of at its core it could do

Josh:
all of this with math and what we're seeing on screen here is how.

Josh:
A virtual one-to-one recreation of Yosemite National Park that was done with

Josh:
a simple prompt asking it to create a recreation of it.

Josh:
And then the model was smart enough to go off and understand the context required

Josh:
in order to build an accurate representation and pulled that off.

Josh:
It did things like it scanned the satellite imagery to figure out what the elevations

Josh:
were like. It pulled topographic maps that it found to figure out specifically what the heights were.

Josh:
It found imagery that any imagery that it could find about the park so that

Josh:
it can reference it and show you i mean look at this resuming in on a waterfall

Josh:
it feels like it's a one-to-one replica,

Josh:
lower fidelity but something that can run inside of your browser it's really

Josh:
impressive how far the model can go on one prompt and i think that's one of

Josh:
the places in which fable stands out in particular is its ability to to reason through

Josh:
your requests in a way that hasn't been done before we filmed an episode yesterday

Josh:
that i would highly recommend listening to about how we've kind of progressed,

Josh:
with our interaction of the model kind of moving up the extraption layer,

Josh:
where first we engage with models, then agents, then harnesses.

Josh:
Now we're just creating these loops where the agent and the.

Josh:
Underlying intelligence is smart

Josh:
enough to actually understand what's required to get you to your goal.

Josh:
And I think that's what this is such a great example of this Yosemite one in

Josh:
particular is, hey, I want you to recreate Yosemite for me so that way I can

Josh:
fly around and I can enjoy it in a one-to-one replica.

Josh:
And it does all the rest for you. And I think that level of critical thinking

Josh:
is something that's novel with Fable 5 that we've never seen in any other model before.

Ejaaz:
I mean, the breakthrough that we're talking about is visual and spatial reasoning.

Ejaaz:
And I think it's important to explain the difference between this and another

Ejaaz:
favorite version of a model that we speak about a lot on this show, which is world models.

Ejaaz:
Typically with world models, it recreates the physical world around us.

Ejaaz:
But most importantly, it understands that physical reality, understands how

Ejaaz:
gravity works, it understands how different forces of nature works,

Ejaaz:
and it applies it when an object has an action. So let's say you kind of like

Ejaaz:
punch a puddle of water, it splashes, the droplets come up.

Ejaaz:
This isn't exactly the same thing. It's still based on theory.

Ejaaz:
This is still an LLM that ingests a lot of text and understands kind of like

Ejaaz:
how the physics works in theory and then recreates what its version of it might

Ejaaz:
be. And this is what we're looking at on screen. It's kind of like,

Ejaaz:
it's known as spatial reasoning or visual intelligence.

Ejaaz:
It's close to the thing, but it's not quite the same thing. Now,

Ejaaz:
another example that I really enjoyed was from Ethan Moloch.

Ejaaz:
Ethan Moloch is one of my favorite AI researchers that analyzes a lot of these

Ejaaz:
new models, but he builds or tests it in really interesting ways.

Ejaaz:
One of these ways was he rebuilt Snake.

Ejaaz:
Now, I am kind of ashamed to admit how much time I spent playing this particular

Ejaaz:
game only because it was like the best version of Snake that we see.

Ejaaz:
And like I'm showing you on the screen right now, it's like incredibly high

Ejaaz:
fidelity. It looks way better than the game I used to play on the Nokia that I had as a kid.

Ejaaz:
But the point is, it's pretty cool. It introduces new power-ups and obviously

Ejaaz:
like you know you can die in usual things

Ejaaz:
Also, forget about creating the game. Claude Fable 5 is really good at playing

Ejaaz:
the game itself. What you're seeing is an accelerated version of it playing

Ejaaz:
Pokemon Fire Red, and it completes the game in, I believe, 50 minutes.

Ejaaz:
And the way that it works is it takes screenshots of the game at any single

Ejaaz:
point, and it basically makes a decision as to which button it wants to click,

Ejaaz:
which step it wants to take, and it is just kind of like spread through a software

Ejaaz:
run where it's able to do that.

Ejaaz:
Now, as a kid growing up and watching, you know, a lot of Pokemon,

Ejaaz:
playing a lot of Pokemon, trading the cards, this is kind of nostalgic,

Ejaaz:
but also kind of scary. We were joking before we started recording,

Ejaaz:
Josh plays Cod quite a bit or plays a lot of computer games.

Ejaaz:
And I wonder about the time in the near future where you're going to be one

Ejaaz:
of the buying a AI agent and it might actually be better than you.

Josh:
Yeah, that's going to be a little traumatic for me on a personal note.

Josh:
Just hurting my ego that I'm losing my game to an AI.

Josh:
I saw another great example on the YouTube channel, actually,

Josh:
where they were Mythos or Fable 5 was playing Factorio.

Josh:
And Factorio is a game that I really enjoy, that I've been playing for a long time.

Josh:
And it was doing it very, very well. And I'm like, oh, dude,

Josh:
you're getting a little too close to home with this. I don't love it.

Josh:
But it's incredibly capable. And we have another example here that shows a use case that is not a game.

Josh:
Instead, it is a it's so cool.

Josh:
So there's this guy named Todd Saunders. He's on X and he he posts this tweet

Josh:
saying fable slash mythos is unbelievable was on a customer call today and had

Josh:
Claude transcribing in the background and on screen we're showing a visual of what that looks like.

Josh:
As they were telling me about the features they wish their current software

Josh:
had, Claude was building the features in real time.

Josh:
By the end of the call, I was able to show a fully working product with the

Josh:
exact workflow they mentioned 15 minutes earlier.

Josh:
Autonomous looped building triggers from a customer call.

Josh:
And this is one of the most amazing things about the model is that it's able

Josh:
to go off and do a lot of the hard work yourself, where it feels like recently

Josh:
you've had to kind of be in the loop. You had to continue to prompt the agent to give it more context.

Josh:
And with this model, it's very easy to give it a goal and give it a verifiable

Josh:
outcome that it can match against the goal.

Josh:
And then it will just go off and do those things. So as is on the customer call,

Josh:
like how cool is that for a salesperson where you're listening to customers,

Josh:
you're listening to complaints and in real time, you're fixing their problems.

Josh:
You're building new software on top of it.

Josh:
It's unbelievably capable. This was one of the demos that I found most interesting too.

Josh:
And then this final example, it's just fun for networking nerds or just computer

Josh:
science nerds in general. We're seeing a highway on screen with cars and buses and vans.

Josh:
And those cars are not random. They are actually associated with specific packets

Josh:
that are being pushed across the network. So it's a really fun and interactive

Josh:
way to visualize network traffic.

Josh:
And I could imagine this being great for a lot of educational purposes.

Josh:
And one of the things that I did notice also on Snake Ejaz is that there was audio, there was sound.

Josh:
It actually generated sounds and audio too. So there's a lot of modalities in

Josh:
which it's actually performing pretty well.

Ejaaz:
Can I add a bonus Easter egg, which we didn't see on the demo,

Ejaaz:
but I know because I played it too much yesterday.

Ejaaz:
There are power-ups that pop up, but the power-ups are software patches.

Ejaaz:
So if you don't hit the software patch power-up, you end up with a random error

Ejaaz:
in the game that you must avoid, like a random wormhole that could like suck

Ejaaz:
you out of the game and like cause you to lose.

Ejaaz:
So like it's coding in real time but you could fix it in real time and if you

Ejaaz:
hit the power up it it like implements the code fix like immediately it was

Ejaaz:
just that's very nerdy but like very cool and very creative not something that

Ejaaz:
we've seen before now if you want to like move away from the

Ejaaz:
The toy aspect or the retail adoption of how you can use this particular model

Ejaaz:
there's of course a lot of enterprise use cases and the number one version of

Ejaaz:
enterprises using Claude is through code specifically and there were a few examples

Ejaaz:
that were included in the official blog my favorite one was Stripe,

Ejaaz:
who did a code migration of 50 million lines of Ruby, which typically would

Ejaaz:
take around two months and several software engineering teams.

Ejaaz:
It took them less than a day using Fable. I believe they used two to three instances

Ejaaz:
of Fable, but still, that's like two to three software engineers that kind of

Ejaaz:
worked night and day continuously to be able to achieve this.

Ejaaz:
I saw another example of a company which had a software engineering team of

Ejaaz:
Opus 4.8 working on projects that would typically take him two weeks.

Ejaaz:
He can now do it in less than a day as well. So the point is there's a massive

Ejaaz:
leap in intelligence for coding specifically with Fable, which is the publicly

Ejaaz:
accessible model. Now, before we continue,

Ejaaz:
We also wanted to like not just look at the demos that other people have recorded.

Ejaaz:
We want to try our own. So we have a few prepared for you today.

Josh:
Okay, so I just, I actually have no idea what you've been working on with these

Josh:
demos. I know you are building a demo. Please share with the class what the

Josh:
prompt was, what you're building and what the outputs of this thing are.

Ejaaz:
For sure. Okay, so one of my favorite breakthroughs with this model is the visual

Ejaaz:
and spatial intelligence that we referenced earlier on.

Ejaaz:
So what I did was I found this hand sketch, this hand drawn version of a floor

Ejaaz:
plan of a blueprint, which I'm showing you on the screen here.

Ejaaz:
It gives you the layout of someone's home. It has a garage.

Ejaaz:
It has bedrooms. It's kind of clunkily drawn. It's not really high fidelity.

Ejaaz:
If an architect looked at this, they'd be like, this is probably physically inaccurate.

Ejaaz:
And then I fed it to Fable and I said, listen, here's a photo of a hand-drawn floor plan.

Ejaaz:
I want you to rebuild it as a single, clean, self-contained SVG that is architecturally

Ejaaz:
accurate. I want you to improve it where you can, improve it in a way that can

Ejaaz:
like, you know, reinforce the wall structures, et cetera, like really high detail things.

Ejaaz:
And it produced this, what we're seeing on screen right here,

Ejaaz:
which is this really high fidelity floor plan. You've got the garage.

Josh:
Architectural grade.

Ejaaz:
Yeah. You've got the surface area measured in meter squared.

Ejaaz:
You've got the swing angle of each door.

Ejaaz:
You've got the entire like kind of like layout of this entire thing,

Ejaaz:
including potential mock furniture, which gave me a second idea,

Ejaaz:
which was like, okay, I asked it, I want to purchase this sofa.

Ejaaz:
These are the dimensions of this sofa.

Ejaaz:
And I want to place it in the dining room or in the lounge. Can you tell me

Ejaaz:
if I can feasibly do this in this full plan or will it get stuck?

Ejaaz:
Like, how do I, how do I do this? Do we have enough doors? Do we have enough

Ejaaz:
space to like maneuver it? Like how would this work?

Ejaaz:
And it said verdict, yes, but not flat. You're going to need to pivot the sofa

Ejaaz:
on its side and kind of like pull it in like vertically.

Ejaaz:
And it gave me this really cool mock-up of how I would do it.

Ejaaz:
It would gave me route A where I take it in through the front door.

Ejaaz:
And this is kind of like the sofa that you see here, but I need to turn it on

Ejaaz:
its side and kind of like shift it through this gap that I'm highlighting on

Ejaaz:
the screen here in green.

Ejaaz:
Or that there's route B where I can take it in from the outside and I have a

Ejaaz:
two meter, very spacious door opening, which I can bring it straight into the

Ejaaz:
lounge and place it right there. So it's really physically accurate going off

Ejaaz:
the comments that we made earlier.

Ejaaz:
It understands the kind of like reasoning behind physics really, really well.

Josh:
It's such a great companion. And this is this gets back to what we talked about

Josh:
earlier, which is like the most complicated and difficult part about this model

Josh:
is figuring out how to engage with it, what to ask it, because it's so capable of doing these things.

Josh:
We love making artifacts for the show as a way to share kind of the ideas that we're talking about.

Josh:
And this might be a good time to get into the benchmarks of how good this model

Josh:
actually is relative to other models. And I was looking at this post that I

Josh:
saw on X showcasing in particular Claude Fable 5 versus GPT 5.5.

Josh:
And my first reaction is, holy shit, that's a huge leap.

Josh:
So Fable 5 low mode is scoring over 10%, whereas GPT 5.5 extra high is getting about 5.7%.

Josh:
Now, this is on Frontier Code Benchmark. This is a specific particular coding

Josh:
benchmark. This is not across the board, but it gives you a sense of how much

Josh:
more powerful this model really is versus,

Josh:
all the others second coding benchmark that i'd say this is probably one of

Josh:
the gold standards this is what a lot of models will use to benchmark themselves against each other

Josh:
um this is the swb bench pro and fable 5 scores 22 points higher than gpt 5.5 which was already,

Josh:
what is that 11 points below opus 4.8 so it's currently looking like gemini

Josh:
3.1 pro gpt 5.5 opus 4.8 and then fable is running away with it and this seems

Josh:
to be the case with almost all of these other.

Ejaaz:
Benchmarks i have a i have a better one for you right so a lot of people listening

Ejaaz:
to this might be thinking okay well i don't code i'm not a software engineer

Ejaaz:
why does this apply to me well there's another benchmark called gdp

Ejaaz:
vow which tests it against real world tasks that take human experts like knowledge

Ejaaz:
workers that you know do back-end admin excel sheets all that kind of stuff

Ejaaz:
hours to do and they compare it to the model

Ejaaz:
take a look at this so fable mythos 5

Ejaaz:
basically achieves the highest benchmark score it's actually almost completed

Ejaaz:
the entire benchmark so they're probably going to have to recreate an entire

Ejaaz:
new benchmark for this but basically what this means is probabilistically if you were to

Ejaaz:
blind test or blind pick the output work of a expert human that is really good

Ejaaz:
at a particular knowledge work task versus this particular model over 50% of

Ejaaz:
the time, you're going to be picking this model, which is just an insane stat to see.

Josh:
Yeah, it's pretty unbelievable. And you have to like, I'm looking at these charts,

Josh:
and you have to ask yourself the question, as Anthropica saturating benchmark,

Josh:
are they running away with it? Like, where is OpenAI in this conversation now?

Josh:
I have to imagine that GBT 5.6 is coming soon.

Josh:
Is it okay maybe it's better than opus 4.8 but can it can it eclipse fable no,

Josh:
and i mean we know that anthropic released mythos months ago so you have to

Josh:
assume that like they've been continuing progress and iterative development on

Josh:
new frontier models that are even more powerful than this and and is this is

Josh:
this beginning to become a runway or are they actually still competitive with

Josh:
each other or maybe we just don't have enough information to tell we kind of

Josh:
have to see what the response from open ai is.

Ejaaz:
Well you look at the cadence between model releases right like what was the

Ejaaz:
time since 4.8 was released. It was like less than, I think,

Ejaaz:
30 days ago. So the cadence is getting...

Josh:
Yeah, we filmed an episode on this not too long ago.

Ejaaz:
Yeah, like I remember that episode, right? And we spoke about it and went through

Ejaaz:
its benchmarks back then.

Ejaaz:
So the point is, these model releases are happening faster, but the capability

Ejaaz:
gaps are even greater, which tells me one thing, which is we're getting closer

Ejaaz:
and closer to the AI models just building itself.

Ejaaz:
They haven't been private about this either. Anthropic has publicly claimed

Ejaaz:
that they have been using Mythos Preview to build this new version that we're

Ejaaz:
talking about today, Fable, right?

Ejaaz:
So we think we've reached a point where you could maybe call it a breakaway

Ejaaz:
from Anthropic, where they basically have recursive self-improvement almost

Ejaaz:
achieved, where the model can do all the research, figure out its own issues,

Ejaaz:
and build a better version of itself.

Ejaaz:
Now, I do want to ground us at this point in this episode, Josh, which is,

Ejaaz:
Fable 5 is an amazing model, but it's one version of the amazing model.

Ejaaz:
There is another version of this model, which is technically better than Fable

Ejaaz:
5, but it is not publicly accessible.

Ejaaz:
It is restricted because it poses itself as a cybersecurity risk,

Ejaaz:
not just a cybersecurity risk, but also a bioweapons risk.

Ejaaz:
It is so good at biology and chemistry that it could feasibly create compounds

Ejaaz:
and a biological weapon that could pose a risk to any sort of nation state.

Ejaaz:
And so for that reason, it is under heavy restrictions and safeguards in the

Ejaaz:
version of Fable where you can't get access to any of this.

Ejaaz:
And only vetted partners and cleared government security initiatives are able

Ejaaz:
to get access to this Mythos 5 thing.

Ejaaz:
Now i put this to the test josh um and i i did a very simple example which was

Ejaaz:
um can you explain how the mitochondria works do you want to bet what its answer was

Josh:
Oh i'm gonna be best it's not touching that it's not touching biology yeah i

Josh:
want to take a second to actually explain the nuances between the models because

Josh:
when you say the word better i'm not sure it's better it's just more complete

Josh:
one model is a complete model one model is a heavily restricted model and in the case of Mythos,

Josh:
it's available for cybersecurity.

Josh:
It's available for biology. And that's what we've seen with Project Glasswing,

Josh:
where they're working privately with companies to kind of fix security vulnerabilities and figure out bio.

Josh:
And in the case of the system card, I saw that it's accelerating some bio experiments

Josh:
at a full order of magnitude, 10 times better. So it's really capable there.

Josh:
The compromise that we had to make in order to receive it was it can't touch

Josh:
bio, it can't touch cyber.

Josh:
So it's just as capable everywhere else. It will not do that.

Josh:
What happens is if you ask in the case like you did, how does mitochondria work?

Josh:
It will route through Opus 4.8 for that answer and then come back and give you a response.

Josh:
So it is as capable everywhere. It's just don't ask about bio,

Josh:
don't ask about cyber because from my experience so far trying it and EGS, it seems like yours.

Josh:
Anytime you get remotely close to those topics, it is just completely shut down,

Josh:
routed through Opus 4.8 instead.

Ejaaz:
Yeah, I think it's it's too aggressive, personally, like, as a former science

Ejaaz:
nerd, I still spend a lot of time trying to digest like some of the latest scientific advancements.

Ejaaz:
And like, listen, I'm not reading research papers. So I work with my best pal

Ejaaz:
Claude to try and figure out, you know, what the latest takeaways are.

Ejaaz:
Now, typically, I could slam that into Opus 4.8. And it would give me an amazing

Ejaaz:
summary. And I could like ask it questions.

Ejaaz:
Now, if I want to use Fable 5, it just simply won't read the paper.

Ejaaz:
As soon as it sees anything related to chemistry or biology,

Ejaaz:
it switches off and reroutes to 4.8.

Ejaaz:
So I am not able to get access to the frontier LLM intelligence or brain that

Ejaaz:
Fable 5 has for me, mythos, even though, you know, my intention isn't to build

Ejaaz:
a bio-weapon by any means, I can't get that analysis.

Ejaaz:
And so that's one version of it, right? Where like it is too heavily restricted.

Ejaaz:
The other version of this is with the more intelligent models get,

Ejaaz:
it's not just going to be super intelligent in one particular vertical.

Ejaaz:
Like for us, it's like, you know, research and creating artifacts and the best content.

Ejaaz:
It should also apply to any other profession, right? Whether you're a scientist,

Ejaaz:
whether you're a mathematician, and whether you are building different kinds

Ejaaz:
of structures or whatever it might be.

Ejaaz:
The fact that it can get triggered so easily or the fact that Anthropic has

Ejaaz:
very heavily restricted that capable intelligence in a way that like even people

Ejaaz:
that have well intentions can't get access to it.

Ejaaz:
In my opinion, is a bit of an issue. And listen, it's V1. I'm sure they're going

Ejaaz:
to like release a bunch of versions of the safeguards where it like makes it a lot easier to use.

Ejaaz:
But for V1, it's kind of like, I think it's overdone. It's important for people,

Ejaaz:
I think, to understand how these safety classifiers work as well.

Ejaaz:
Think of Claude Mythos 5 having an AI model or system that is watching it.

Ejaaz:
And as soon as one of the red flags that it's been trained on is triggered,

Ejaaz:
for example, anything to do with biology or chemistry, it gets switched off

Ejaaz:
immediately and rerouted to 4.8.

Ejaaz:
Now, there's four particular categories that Fable can't get access to.

Ejaaz:
It is cybersecurity, for biology, for chemistry, and for distillation as well.

Ejaaz:
And this is a key one which caused a lot of contention in the public ecosystem

Ejaaz:
when they launched yesterday, which is if you were to ask about model training

Ejaaz:
techniques or even just simple general questions around, hey,

Ejaaz:
I have this AI agent, it's pretty clunky,

Ejaaz:
how can I improve its harness to kind of make it go quicker or use less tokens?

Ejaaz:
Automatically degrades performance. And this is the key change in this fourth

Ejaaz:
category with distillation.

Ejaaz:
Anything that Anthropic considers to be trying to derive its model to build

Ejaaz:
another model, it gives you intentional poor performance. And it doesn't even tell you.

Ejaaz:
Now, it says that this happens for 0.3% of cases, but my guess is it's probably

Ejaaz:
happening for higher reasons. And listen, it's completely within Anthropic's

Ejaaz:
right to do this. I get it. I understand it. You want to remain competitive.

Ejaaz:
But it's just interesting to see when like you have this intelligence model

Ejaaz:
where, you know, it's meant to kind of like blossom and create and help other

Ejaaz:
people build different things.

Ejaaz:
But they're being competitive when it comes to other models, I guess.

Josh:
Well, we're getting to this unique intersection where like, they have mythos,

Josh:
they've had mythos for a little while, and they decided to keep it private.

Josh:
And that was okay, and somewhat understood because it was really discovering

Josh:
a lot of zero day vulnerabilities.

Josh:
And it seems like they worked pretty hard to figure out a way to not only improve

Josh:
the quality of the model but actually make it public and i guess like the question

Josh:
we're gonna have to start asking as these ai labs continue to create these like unbelievably

Josh:
uh forward-looking frontier models is like to what capacity are we just happy

Josh:
to have them like how much should we expect out of the labs when it comes to

Josh:
delivering these models like in my case i'm pretty stoked to be able to use fable 5,

Josh:
And I'm not interested in biology. I'm not interested in distilling the model.

Josh:
I'm just like pretty stoked to do my day-to-day work with this capable model.

Josh:
And in that sense, it's really fun and exciting and interesting.

Josh:
And I think it's the start of a longer conversation.

Josh:
We saw some legislation come in a few weeks ago, last week maybe,

Josh:
about requiring AI Frontier Labs to kind of showcase the models privately with

Josh:
the government to share what's coming down the line.

Josh:
And I guess in this essence, I'm more excited to have the model versus not have

Josh:
the model and have it have these constraints in the hope that it will slowly become

Josh:
unwounded as they kind of improve and iterate on the quality and the kind of

Josh:
like security set of this model.

Ejaaz:
Yeah, listen, I think Anthropic is ultimately doing the right thing.

Ejaaz:
I think that they can't just kind of diffuse this model to anyone and everyone

Ejaaz:
because malicious actors, however few they might be, will actually end up doing

Ejaaz:
something dangerous with this.

Ejaaz:
That being said, I think the subjectivity and who gets to govern that subjectivity is important.

Ejaaz:
Like, I can imagine a future version of an Anthropic model that isn't just necessarily

Ejaaz:
really good at biology or cybersecurity.

Ejaaz:
It might be really good at something such as trading, right,

Ejaaz:
for example. And then the question becomes, who gets access to this trading

Ejaaz:
model that is so good that it could break the stock market?

Ejaaz:
And maybe if you are Citadel, who are closely aligned with Anthropica,

Ejaaz:
I'm theorizing here, then they get access to it, but Jane Street won't get access to it.

Ejaaz:
And so it becomes this heavily-based nuance that is only dictated by maybe the

Ejaaz:
government and maybe Anthropica itself.

Ejaaz:
There was talks around like Trump taking a stake in some of these AI labs to

Ejaaz:
nationalize it for this exact reason, because it could pose a threat and they

Ejaaz:
want to have governance decisions. It just gets a little murky and messy.

Ejaaz:
And I think we're at the fork in the road.

Ejaaz:
There's no going back at this point. We are now entering a phase where

Ejaaz:
The model that you have access to may not be the most intelligent model for

Ejaaz:
the specific thing. And listen, it may not be the thing that you necessarily

Ejaaz:
do on a day-to-day, but it's a lot of things that other people do day-to-day,

Ejaaz:
and they want to get access to this model. How that is governed, I don't know.

Ejaaz:
The other restriction, which I found really interesting that I noticed in the

Ejaaz:
footnotes of their system card or announcement blog post is on June 22nd,

Ejaaz:
we ceased to get access to Fable 5.

Ejaaz:
Now i think this was taken massively out of context because i think the reasoning behind this is

Ejaaz:
it's because it depends on availability of compute so if by june 22nd anthropic

Ejaaz:
has more available compute to distribute to users

Ejaaz:
then it wouldn't be the case but on the case that it is it would shift to a

Ejaaz:
pay-per-usage model which means that you buy credits and if your credits are

Ejaaz:
consumed you then need to buy more credits kind of like the api model is that right

Josh:
Yeah, according to the blog post, it says from today through June 22nd,

Josh:
Fable 5 is included on Pro, Max, Team, and Seat-based enterprise plans at no extra cost.

Josh:
On June 23rd, we'll remove Fable 5 from those plans. Using it after that will require usage credit.

Josh:
If capacity allows, we'll extend the included window. After this point,

Josh:
when sufficient capacity allows us to do so, we aim to restore Fable 5 as a

Josh:
standard part of subscription plans.

Josh:
We intend to do this as quickly as we can.

Josh:
And yeah, it sounds like Fable 5 consumes a lot of compute. When you load it up inside of the,

Josh:
app it says fable is the most capable model and draws down usage twice as fast

Josh:
as opus so you have to imagine

Josh:
that it consumes a lot of gpus they're clearly using those gpus for a lot of

Josh:
things i think the idea is to give a preview and then extend that for as long

Josh:
as possible or just continue to extend it perpetually based on compute

Josh:
i think to your earlier point we're very much at a fork in the road

Josh:
when it comes to these models being capable enough to

Josh:
really make a meaningful impact in the world and we've spoken so much about

Josh:
alignment and ai safety and it's kind of been this like open-ended fuzzy thing

Josh:
where it hasn't really practically applied to anything that's happened before

Josh:
and we're finally at a moment in time in which the models are becoming capable

Josh:
enough to have that conversation about

Josh:
ai alignment ai safety you're starting to see why a lot of the teams are taking

Josh:
it so seriously because it is the singular question is like answering what you

Josh:
just said who gets access to this model how is it going to be restricted who

Josh:
gets to decide that and that's why the alignment and safety conversation.

Josh:
Is so important. And while you start to see a lot of the company cultures within

Josh:
these companies align around these different priority sets that separate them from each other.

Josh:
So this is, it's, it's a new day. It's a new era today.

Josh:
We are moving into the, the next frontier of models.

Josh:
It was pushed forward a considerable amount and in a way that I don't think

Josh:
we've experienced in quite a long time. And it's really exciting to see.

Josh:
I'm very excited to play with Fable, spend some time kind of generating outputs,

Josh:
figuring out what it's most capable in that could help us in the day-to-day.

Josh:
Like for me, if they never told me it wasn't gonna do bio or cyber,

Josh:
I'm not sure I'd ever come across it because that's not really within the realm of uses that I have.

Josh:
So I'm excited to just kind of play with it and figure out best use cases for

Josh:
this. In terms of pricing, what I found really interesting is CloudFable 5 is

Josh:
only twice the price of GPT 5.5.

Josh:
I believe it's $10 per million input tokens, $50 per million out, and,

Josh:
gpt 5.5 is five dollars and thirty dollars out so pretty close and much more

Josh:
capable so if the case that it does get removed from subscriptions it is still

Josh:
available from api it is not

Josh:
as expensive as i think a lot of people thought this is significantly cheaper

Josh:
than what i believe mythos preview was early on.

Ejaaz:
Yeah it's my new favorite model i've been using it relentlessly for the last

Ejaaz:
10 hours um one thing that

Ejaaz:
Uh is a really strong capability that it has is long horizon tasks like this

Ejaaz:
model is engineered from the ground up to be able to work like a dog for six

Ejaaz:
to 12 hours at a time on whatever project that you have.

Ejaaz:
And it has this loop function, which basically says, if you come across a problem,

Ejaaz:
don't ask me, try and figure out yourself and do the thing, build the thing.

Ejaaz:
That's why we had people build world engines from scratch that we demoed earlier

Ejaaz:
and these games from scratch, all from a single prompt, the library of Babel.

Ejaaz:
So if you have an idea or if you've been pondering on a project that you've

Ejaaz:
been putting on for a while, because you're like, I know I could probably vibe

Ejaaz:
code this, but I don't want to spend like an hour doing this.

Ejaaz:
Now you just need to write one detailed prompt and you should be able to do

Ejaaz:
this. So my prompt for the listeners of the show as we wrap up this episode

Ejaaz:
is get access to this model.

Ejaaz:
Try it out. And I'm curious what your thoughts are on it. Do you think the restrictions

Ejaaz:
affect you specifically?

Ejaaz:
Or do you think it's a really good general purpose model and you're happy with

Ejaaz:
how it's presented itself. You don't care about Mythos 5 in effect.

Ejaaz:
And also, what other projects are you going to be doing with this?

Ejaaz:
Are there any kind of demos or use cases that we haven't covered that might

Ejaaz:
be specific to you in your leisure or your particular work that you might want to apply this to?

Ejaaz:
Let us know the feedback in the comments to this video or DM us on X.

Ejaaz:
Our profiles are linked below.

Ejaaz:
We want to hear back from you. But I think that brings us to the end of this

Ejaaz:
episode. We have now like a new world-leading model?

Ejaaz:
OpenAI is going to have to answer to this, but it seems like Anthropic is running

Ejaaz:
away with it. Josh, any final thoughts?

Josh:
That's it. This is a new, it's a new era today. Like, I feel like we should celebrate.

Josh:
This is a new frontier that has been pushed forward very far in an industry

Josh:
we care very deeply about. So it's exciting to see. I'm stoked to use it.

Josh:
I'm curious to hear what the best types of prompts or use cases are that anyone

Josh:
who's listening has found.

Josh:
If you enjoyed this episode, don't forget to share it with a friend who might

Josh:
also want to try Cloud Fable 5 and experiment and get their feedback on how

Josh:
it's being used to improve their life, improve productivity,

Josh:
whatever use cases it may be. But as always, thank you all so much for watching

Josh:
and we will see you guys in the next one.