Even before Mythos I was getting asked more and more what Anthropic's deal is, and why tf they're acting the way they're acting if they believe what they say they believe.
The best answer I can give is that their basic worldview is something like:
1. There are giant, dangerous monsters in the forest
2. We see others going out and making loud noises that will rouse the monsters, and they're not going to stop because of all the treasure and magical artifacts that can be found in the forest
3. We believe the best way we can help is to send out our own vanguard to go faster and farther into the forest than everyone else, because we'll spend a ton on monster containment and taming and we'll also send back detailed reports of what monsters we're finding so that the townspeople can ready themselves, which those other guys won't do
On the one hand I understand how they got there, and I think it's possible they're basically right. On the other hand it's not hard to see why this approach makes people wonder if you're crazy or lying or both.
Thus: The way to fix the 'jailbreak' (of 'fix this code') is to weaken the classifiers so they never ban the parallel action in the first place, which therefore means you can no longer use 'fix this code' to get around the controls. QED.
@EricLevitz Maybe I should start thinking about the consequences of universal suffrage and how we've given voting rights to the gender which is more emotional, more impulsive, less college educated, more likely to commit violent crimes, etc.
So many posts over the last week have said some variant of "you can't say that you're building a superweapon and then expect the admin to not freak out and do panicked bullshit"
And sorry, but yes, you absolutely can/should expect that from a normal admin
https://t.co/vqMZUrNFxn
Just occurred to me that Anthropic employees who are not US persons will not be able to use Fable/Mythos, making this plausibly (and to be clear, accidentally) the first regulation on recursive self-improvement.
Agree with basically all of this, but I don't get why a consequential policy blunder is being labeled a "brouhaha."
(is the idea that you're only allowed to agree that the orange man is bad if you make it seem like it's no biggie?)
Marc Andreesen said that Biden’s requirement that frontier AI developers tell the government about their safety practices was an existential threat to US AI, but he thinks global export controls on US AI models is “based” if they’re against people he hates/missed the series A of
@NoFilm_Analysis Njoku can legitimately block in-line in a way that Gadsden has not shown - they are not comparable in that regard. But agree on the overall sentiment and Waller / receiving stuff
Fable is a good model. As with all new models, it is simultaneously excellent and entirely unremarkable (relative to other models). It is slow and expensive, and the "loops are all you need" discourse they are pushing is obvious in the context of someone using Fable-class models
What I've found so far is that for broad scope design (code architecture) tasks, Fable is unremarkable. Or, not better enough to justify its cost and speed.
But in highly targeted goal-oriented loops, it is another beast entirely. It is very slow but produces very good results.
I let it churn on optimizing a SwiftUI-layout resolver in Go I wrote and it was able to bring it down to an order of magnitude I could not reach myself (micro => nanosecond scale). But it took 2 hours and $40 to do it and I had to claw back some changes it overfit to Apple Silicon. Still, very worth it.
In comparison, for "implement this feature/change" iterative work, I ran head-to-head Fable vs GPT5.5 vs. GLM-5.1. They all produced equally acceptable final results, but GPT5/GLM did it in a couple minutes and Fable was churning away for 40 minutes. And GLM cost me less than a dollar, GPT5.5 ~$1.50, and Fable cost $9.
You can see that in this context, interactively working with an agent is nonsense. Its too slow. You need to write loops to keep the agent working and you probably want to highly parallelize the work being done. As with all things, I think a balance makes sense...
My sense is that I'd reserve Fable for targeted, surgical analysis and work. Not for daily driving everyday tasks.
I'm going to keep spending a shitload of money (relatively) and maining Fable for the rest of the week to continue to judge, will report if anything changes. I'll continue to head-to-head as well.
@NoFilm_Analysis You can hard rebuild without taking such clear value losses unnecessarily. Even if you think the 1st + 2nd will provide greater return long term than KC and Ladd, you can still sell KC and Ladd (likely individually) for a lot more so it's just a bad move
I disagree with this line of reasoning (though the Berkeley Law policy is light speed moronic).
It’s still important to learn multiplication and long division even when calculators exist. The foundations and skills learned by doing already automatable tasks matter.