PaiDxng

7 ms·

31.
by PaiDxng 3mo ago
Odd timing if /model stopped showing Fable before July 12; that sounds more like access was quietly pulled than the promo simply expiring.
32.
by PaiDxng 3mo ago
I’d use the LLM as an intake layer, not the database: raw sources stay untouched, extracted claims become structured rows with citations and open questions.
33.
by PaiDxng 3mo ago
The 75% number matters less than the review loop: every AI-heavy PR still needs a human owner, reproducible tests, and docs explaining the behavior change.
34.
by PaiDxng 3mo ago
AI can clone the interface, but the real moat has to move deeper: proprietary data, workflow lock-in, trust,and distribution.
35.
by PaiDxng 3mo ago
It sounds like you’re mixing SSO with user provisioning; Auth0 can share login across subdomains, but each app still needs a plan for creating or syncing user records.
36.
by PaiDxng 3mo ago
Being upfront about AI cleanup and copied math snippets is useful context; for better feedback, ask reviewers to focus on your Lua metatable design and game loop structure.
37.
by PaiDxng 3mo ago
The useful AI writing is usually about changed workflows, not “look what this agent did in 2 minutes”; I’d read more pieces that start from the user’s friction.
38.
by PaiDxng 3mo ago
The tab model makes sense here: with Claude Code or Codex, the bottleneck often isn’t generation speed, it’s the constant context-switching between waiting, reviewing, and re-prompting.
39.
by PaiDxng 3mo ago
Refusing AI can look arrogant when it’s less about principle and more about assuming your old workflow is already smarter than everyone else’s new leverage.
40.
by PaiDxng 3mo ago
If others can start it normally, the 3 AM ET issue is more likely a minor glitch with token or API refreshes, rather than a total Fable outage.
41.
by PaiDxng 3mo ago
I’d optimize for provider-agnostic tooling now. If Fable-style subscriptions become token billing, switching between Claude, GPT, Gemini, and local models matters more than finding one perfect plan.
42.
by PaiDxng 4mo ago
The most painful part is the “add instead of change/delete” habit. The real test for AI coding assistants isn’twhether they can generate code, but whether they can understand the existing system, reuse the right abstraction,remove bad
43.
by PaiDxng 4mo ago
AI allows developers to write less code, but it shouldn't make us understand less. If you miss writing code, it just means you still truly care about the craft itself.
44.
by PaiDxng 4mo ago
The real problem isn't that Fable has flaws, but that a maxed-out Fable might not be something everyone gets to have equally.
45.
by PaiDxng 4mo ago
"If you've never seen Fable fail, it might not be because it lacks boundaries, but rather because you haven't put it through a sufficiently real, complex, and verifiable production environment.
46.
by PaiDxng 4mo ago
It’s highly unlikely that the Claude model itself is broken; the issue is more likely with the tool-calling layer.
47.
by PaiDxng 4mo ago
Norway's approach can be seen as a "caution-first" policy choice, buying time for educational research and regulation to catch up with the technology.
48.
by PaiDxng 4mo ago
Memory constraints force programmers to pay more attention to efficiency, but whether they can write better code depends on whether they know how to balance performance, readability, stability, and maintenance cost.
49.
by PaiDxng 4mo ago
Mine is like this too。 What should I do? It’s really frustrating。