7 prompting principles for GPT-6 Astra and next-gen models: roles, context, self-review
goyalshaliniuk · x · 2026-09-06
The author lays out 7 prompting principles for next-gen models like GPT-6 Astra and Fable 5.1, arguing old prompting habits waste new capability:
- Give the AI a job: set a role and specialty tied to a goal.
- Give it the full picture: share context and background before the ask.
- Define what 'good' means: state the desired outcome and success criteria.
- Set the rules: specify tone, length, format and what to avoid.
- Give it a reference: anchor style with an example.
- Make it improve the first draft: have the model evaluate its draft against criteria and return a revised final version.
- (continued in thread)
More from Apps
- Dimillian shows off more natural world generation and optimized map drawing — Dimillian · 2026-09-06
- Rainy Worlds: Six WebGPU Rain Scenes You Can Wander in the Browser, With Your Own Soundtrack — anselm · 2026-09-06
- Audacity 4.0 lands after 5.5 years with redesigned UI, dark mode and Workspaces — solyarisoftware · 2026-09-06
- Redditor builds free AskSary creative studio with game engine GPT can play and patch live — Beneficial-Cow-7408 · 2026-09-06
- Generative AI point-and-click detective game built without a multimillion-dollar budget — tristanbob · 2026-09-06
- Redditor uses coding agent to finish an obscure 18xx Pennsylvania railroad game — natep1098 · 2026-09-06