The Maximally Accelerated Question
A forcing question that separates critical path from what can wait
- Difficulty
- Easy
- Time to result
- ~days to results
- Steps
- 4
- Confidence
- 92%
A recurring thought exercise Turley uses to cut through organizational blockers: ask 'if this were the single most important thing and you wanted to truly maximally accelerate it, what would you do?' You don't necessarily execute the answer — it's a diagnostic that reveals what is genuinely critical-path versus what has been slotted into a comfortable timeline out of habit. Use it whenever a plan defaults to 'check back next week' or 'circle back next quarter.'
Origin
Turley developed it as ChatGPT scaled and OpenAI began hiring from larger tech companies whose default cadence was slower. It became so ingrained the team made a pink Comic Sans Slack emoji reading 'Is this Maximalist?' that people stamp on work to force the question.
Core principles
- 01The default timeline is usually inherited habit, not a real constraint
- 02Asking the extreme version exposes which blockers are actually load-bearing
- 03Use it sparingly and context-dependently — some things (safety) should NOT be maximally accelerated
- 04Execution speed, not idea generation, is the scarce resource in AI
How to run it
- 1
Frame the item as the single most important thing
Take the task or decision and provisionally treat it as the top priority above everything else.
Pro tip This works even when you know it isn't actually #1 — the framing is the tool.
- 2
Ask what maximal acceleration would look like
Force the question: why can't we do this now? Why not tomorrow? What would we do if we truly wanted it as fast as physically possible?
Watch out You are not committing to actually doing the maximal version — say that explicitly so people answer honestly.
- 3
Read out the critical path
Whatever survives as a genuine blocker in the maximal scenario is critical-path. Everything else is revealed as deferrable.
- 4
Exempt the deliberate-process zones
Explicitly carve out domains where you WANT process and do not want to accelerate — for OpenAI that is frontier-model safety, red-teaming, and system cards.
Pro tip Naming the exemptions keeps the tool credible; a team that accelerates everything stops trusting it.
Watch out Applying this to a high-stakes/irreversible domain is the misuse case.
In the wild
After months of prototyping bespoke assistant ideas, Turley ran out of patience and set a 10-day window from decision to launch of an open-ended chat product, purely to start learning. The underlying research had baked for far longer, but productization was compressed to 10 days.
→ Shipped right before the holidays; became the fastest-growing consumer product in history.
Common mistakes
Treating it as a mandate to actually do the maximal thing
The point is diagnostic. If teams hear it as 'you must ship the fastest version,' they lose the honest read on critical path and start cutting things that matter.
Applying it to safety-critical or irreversible work
Turley deliberately exempts frontier-model safety. Acceleration questions belong in reversible product development, not domains where the stakes justify heavy process.
Is it for you?
Best for
Product and eng leaders at scaling companies where inherited 'circle back later' cadence is quietly slowing critical work
Not ideal for
Safety-critical, regulatory, or irreversible decisions where deliberate process is the point
From the transcript
“I was like people asking like, okay, if like this was the most important thing and you wanted to truly maximally accelerate it, what would…”
“It's a comic sense emoji that says, "Is this Maximalist?"”
“there's some things where you don't want to accelerate um as as quickly as possible um because you you kind of want process and we're…”
From the episode
Inside ChatGPT: The fastest-growing product in history
Nick Turley (Head of ChatGPT at OpenAI)