Using AI as a First-Time Founder
Where AI actually earns its keep in a first-time founder's real week, three prompts built for prioritization and hard calls under uncertainty, and which documented founder frameworks are worth loading alongside it.
A first-time founder's week is full of decisions that feel enormous and mostly aren't reversible-feeling in the moment: whether to hire ahead of revenue, whether a slipping timeline is a real problem or normal startup noise, whether to cut a half-built feature or push through it, whether the metric that's flat is actually a warning sign. Generic AI use treats each of these as a one-off question to be answered and moved past. The more useful use treats them as a small set of recurring judgment calls, stress-tested the same way every time.
What does a first-time founder's actual week look like?
The recurring calls repeat in shape more than they repeat in specifics: whether to make a hire before revenue justifies the cost, whether a timeline that's slipping is an actual problem or the normal noise of building something new, whether to cut a feature the team has already sunk weeks into or push it across the line anyway, whether a metric that's gone flat is an early warning or statistical noise in a small sample. Every one of these gets decided under real uncertainty, usually without the data that would make the call obvious, and usually while several other decisions are competing for the same attention. First-time founders in particular tend to feel each one as uniquely high-stakes, since there's no prior experience of having made a similar call before and survived it, which makes the uncertainty feel sharper than it might to someone on their third company.
Where does generic AI actually help here, and where does it quietly fail a first-time founder?
Unaided, it's genuinely useful for stress-testing one specific claim you're already making, structuring a comparison between two paths, and generating the case against a decision you've already leaned toward.
It's weak at anything requiring knowledge only you have: your actual runway in months, your team's real morale beneath the standup-meeting version, what a specific investor or customer said in a conversation it has no record of. Feed it a vague version of the situation and it will still answer with confidence, since a lack of real information has never stopped a fluent-sounding response from being produced.
Which three prompts are worth running against this week's decisions?
These run in Claude, ChatGPT, or Gemini as written.
"We're considering hiring for [role] at [cost]. Current monthly revenue is [amount] and runway is [months]. Given this, what would have to be true for this hire to be the right call right now, and is it actually true?"
Why it works: naming the specific condition that would justify the hire, rather than debating the hire in the abstract, turns a values-laden question into a checkable one.
"Imagine it's a year from now and [decision] turned out to be a mistake. Write the explanation of what went wrong, in detail, as if it already happened."
Why it works: framing the failure as having already happened, rather than as a hypothetical risk, produces far more specific failure modes than a generic "what could go wrong" question does.
"We've built [feature] and sunk [time/cost] into it. Here's what we know about actual user demand for it: [data or lack of it]. Ignoring what we've already spent, should we finish it, or cut it?"
Why it works: explicitly instructing it to ignore sunk cost counters the single most common founder bias in this exact decision, continuing to fund something because of what's already been spent on it rather than what it's actually worth finishing.
Where does this stop being useful, even with all three run?
Run all three and this week's hiring, timeline, and feature calls will be sharper than gut feel alone produces. What you won't have is a system that knows your actual runway to the dollar, your team's real morale beneath the standup version, or what a specific investor meant by a comment that's been bothering you since the meeting. The gap isn't this week's decisions; it's that no prompt substitutes for the judgment a founder builds by actually living inside the company day to day.
Which documented frameworks actually fit a first-time founder, and why?
Naval Ravikant's documented writing and interviews on leverage and judgment argue that specific knowledge and accountability compound in ways a generic skill doesn't, directly relevant to deciding which parts of the business only you can do right now. Brian Chesky's publicly documented founder-mode operating discipline at Airbnb, staying close to product detail rather than delegating it away entirely, addresses the timeline and quality-control calls that come with scaling past the first few hires. Alex Hormozi's documented offer-construction method treats the offer itself as the primary lever in a business's growth, directly relevant to whether a flat metric is a positioning problem or a demand problem.
Each fits a different recurring failure: Ravikant for deciding what deserves your specific attention, Chesky for staying close enough to the product to catch a real problem early, Hormozi for diagnosing whether a stalled metric is actually an offer problem in disguise. None of the three requires abandoning the others; a founder can genuinely need Ravikant's judgment on where to spend attention and Chesky's operating discipline in the same week, applied to different problems.
What's the actual process for loading one of these into Claude, ChatGPT, or Gemini?
In Claude, each installs as a Skill: go to Settings → Skills → Add skill and upload the .zip. Claude reads the skill's description and auto-invokes it when your question matches, or you can force it with /naval-ravikant-framework at the start of a message. Claude Code users get the same thing via ~/.claude/skills (every project) or a project's .claude/skills folder (that project only).
In ChatGPT, the plain .md file goes into a Custom GPT: Create a GPT → Configure → Instructions, paste the framework's content in. Standard Custom Instructions (the "About me" fields) cap out at 1,500 characters, too short for a full framework, so use a Custom GPT specifically, not Custom Instructions.
In Gemini, paste the .md content directly into the system prompt if the interface you're using supports one, or lead your first message with it: "Here is a thinking framework to apply throughout this conversation. Read it carefully, then answer using this framework's approach," followed by the file's contents.
Where should a first-time founder actually start?
The Founder Stack bundles Ravikant, Chesky, and Hormozi alongside seven other founder-relevant thinkers, Buffett, Dalio, and James Clear among them, at the ten-pack rate. For the operating cadence underneath all three prompts above, Founder Weekly Review is a $129 tool built to guide that rhythm, week after week.
All the copy-paste prompts from this guide, and the rest of the AI Practice series, live at /prompts too, free to copy without reading the guide first.
Frequently asked questions
Isn't every founder decision unique? How can a prompt help with that?
The specifics are unique, but the shape of the decision usually isn't. Whether to hire ahead of revenue, whether a slipping timeline is a real problem or normal startup noise, whether to cut a feature or keep building it, these recur across nearly every first-time founder's year even though the specific numbers, product, and market differ completely each time you face one.
Can AI tell me if my startup idea is any good?
Not reliably, and treat a confident-sounding verdict with real suspicion. It has no access to your specific market, your actual unit economics, or what you've already tried and learned firsthand from real customers. It's more useful for stress-testing a specific claim you're already making about the idea than for rendering an overall verdict on whether the idea itself is good.
What's the biggest mistake first-time founders make asking AI for help?
Asking it to validate a decision they've already made instead of asking it to argue against it. A prompt like "is this a good idea" invites agreement; a prompt that asks for the strongest case against the plan, and for what would have to be true for it to fail, produces something genuinely useful instead of comfort dressed as analysis.
Should I use AI to write my pitch deck or investor updates?
For structuring one, yes, that's a reasonable and common use. For generating the actual narrative and numbers your business is built on, no, since an investor is evaluating your specific judgment and your specific traction, not a well-organized deck someone else could have written. Use it to sharpen the structure of a story that's actually true, not to invent one that sounds better.
How is this different from just asking AI general startup advice?
General startup advice is built to apply to any founder in any situation, which is exactly why it tends to read as reasonable but not particularly useful for your specific week. The prompts here ask about your specific numbers, your specific timeline, your specific tradeoff, which produces a sharper and more usable answer than general advice ever will, precisely because it can't be reused unchanged for someone else's company.
Can these prompts replace a cofounder or an advisor?
No, and the honest reason is the same reason a solo founder needs one in the first place: someone who has genuine skin in the outcome and a different vantage point on the business. AI can structure your own reasoning and catch an inconsistency in it; it has no stake in whether the company survives and no independent read on your market.
What's the honest limit of using AI as a first-time founder?
It has no access to your actual runway, your team's real morale, or what a specific investor or customer actually said in a room you weren't recording at the time. Use it to stress-test your own reasoning and catch the case you're not making against yourself; the call, and everything riding on it, is still entirely yours to make and live with.
Written by Gareth Hoyle. Last updated 26 September 2026. Part of the authority.md guides library.
More guides.
Bringing AI Into a Creative Project Without Ruining It
What generic AI brainstorming actually gives a creative project and what it quietly flattens, four prompts that help today without taking over, and what changes when the AI applies Rick Rubin's documented listening practice instead of generic idea generation.
AI for Hiring Decisions: Where Claude and ChatGPT Actually Help
A grounded look at what generic AI use gets right and wrong when you're deciding on a hire, four prompts that genuinely sharpen the decision, and what changes when the AI applies a documented evidence-based hiring framework instead of generic interview advice.
Using AI to Prep for a Negotiation: What Actually Works
Generic AI negotiation advice is genuinely useful and genuinely limited. Here's what Claude and ChatGPT get right before a negotiation, four prompts that work today, and what changes once the AI has a documented framework instead of a generic one.