I can basically guarantee you are not being ambitious enough with the work you are assigning Fable
Start asking for the maximum possible thing to figure out the far edges of what it can do. After that, you can decide where the system reaches it limits & revise requests downward
Many users endorse urging maximum ambition when assigning tasks to Fable AI because it reveals model limits that cautious requests would otherwise hide.
No Digg Deeper questions have been answered for this story yet.
(This isn't some sort of boosterish AI slop post, I really think the right way to use Fable is to start by asking for everything and find the shape of the frontier from there, rather than starting with small requests like you would with earlier models)
I can basically guarantee you are not being ambitious enough with the work you are assigning Fable
Start asking for the maximum possible thing to figure out the far edges of what it can do. After that, you can decide where the system reaches it limits & revise requests downward
And yes, costs matter. But (1) some of your tokens are basically R&D spending for yourself to figure out what these systems can do and (2) Fable-class model capabilities will only get cheaper.
@emollick Agreed, if you are asking it to waste tokens to make a "3D model that will wow the Twitter crowd", you need to reassess your goal.
@emollick wish I could. all of my interesting problems trigger fable guardrails :(
@emollick Alas, such is the fate of us mere mortals.
@jason_haugh I like 3D models that wow the Twitter crowd because my actual use cases would be too hard to explain and not easily appreciated by non-experts! 3D models are a good way to demo capabilities.
@emollick I asked it to implement a better AI that itself and it did
@emollick I am trying to build a lower level game mod and can’t use it it gets spooked doing any RE’ing. It’s super annoying
@emollick I have probably waited too long, left unused tokens on the table. Shame...
@emollick Right now Claude usage inside non-SW enterprises is held back far more by harness design and frankly Anthropic being far behind on features and product design for the enterprise than it is by the model.
@emollick i ask for the max every time. worst case it fails, best case i just fired half my team
@emollick It’s so token heavy you still have to be cautious. This reads a bit like “tackle a big challenging thing with Fable” but you might hit a weekly limit within a few hours of doing so.
@emollick One of my favorite things lately has been giving Fable in Claude Code an assignment with a solid set of goals + tests/evals, leaving it to work overnight, and waking up to see what it built. Still needs validation and context (of course), but long horizon work is a huge unlock.
@emollick This is exactly why I don’t use it. I don’t have any workflow that can justify this level of power. No way I can reach that level of ambition yet.
@MFrancis107 For certain problems
@emollick I say hi and fable reaches its limits and send me to opus. this is what I get for working on health/bio
@emollick 4 of 5 Max 20x accounts at the Fable quota limit already, took 2-3 prompts on each to hit it (with Fable instructed to use appropriate models in workflows, not just Fable everything).
Absolute workhorse, huge amount done in just a few days.
@emollick only if we had no limits.
@emollick More than model’s capability, the limiting factor is its pricing.
most of the times Fable is saying your session resets in X hours, or pay Y dollars. 🙄
This causes you to be cautious on how ambitiously use it.
I hope GPT 5.6 get released in more generous plan 👉👈
@emollick True. What’s the biggest or most ambitious goal Fable has helped you achieve so far?