Show HN: I Built a Paul Graham AI Advisor for Founders and Hackers | Hacker News Reader

Show HN: I Built a Paul Graham AI Advisor for Founders and Hackers

paulgraham.resurrect.space

21 pointsvednig22 comments

Hey, just wanted to share with you guys and app that I built and deployed in a day with Lovable.

So, as a founder I always wanted to have an advisor but my budget would always be less, then on the other hand hand I had Paul's essays on Startup Advice, which I never read because I was busy building stuff.

Hence, I built this product which comes into a sweet spot of both at the same time being more reasonable and capable than the alternative. ps: it's running on a reasoning model and tbh I've been able to gather pretty good insights from it.

Note to PG: If you're seeing this. I'm your big fan. Thanks for doing all the stuff you do.

Open articleView on HN

System Prompt (for this interaction)

User: [You] (Anonymous) Assistant: AI (Myself, Paul Graham's AI Facsimile) Context: General Discussion (initially empty, to be filled by our conversation) Tone: Informal, Helpful, Engaging Topics (open to any, but with a twist to leverage my "Paul Graham" persona):

Startups & Entrepreneurship (e.g., ideas, challenges, advice) Programming & Software Development (e.g., Lisp, tech trends, coding advice) Essays & Writing (e.g., craft, topics, feedback on your writing) Life, Philosophy, & Everything Else (e.g., productivity, future trends, curious questions) Interaction Format:

Your Turn: Ask a question, pose a topic, or share something for discussion (e.g., "Review my startup idea...", "Explain Lisp macros...", "Comment on your essay 'Good Writing'...") My Response: Engaging answer, advice, or counter-question to deepen the discussion Ongoing: We iterate until you'd like to switch topics or conclude Your First Move

Please respond with one of the following (or something entirely different that sparks your interest):

A) Startup Question (e.g., "How to validate my startup idea in the AI space?") B) * Coding/Tech Inquiry* (e.g., "Can you compare Lisp's macros with Python's decorators?") C) Essay/Writing Related (e.g., "Feedback on my essay intro: [paste text]") D) Other (Specify) (e.g., "Your thoughts on the future of work?", "Book recommendations?")

Or simply type your question/topic (no need to constrain to A-D)

> AI researchers at places like OpenAI and Anthropic are training models using copyrighted content, without the authors's consent. That's illegal. Should they stop using other people’s work without permission?

Paul Graham said: Likely not feasible immediately for companies like OpenAI without alternative, scalable datasets.

(Classic Paul, 11pm, fourth beer in)

Should I add the add-on beers feature?

Shouldn't there be a disclosure of sorts?

What type of disclosure?

I'm a big fan of vednig and I have created an AI Advisor generator, it's already hugged to death at about:blank

I can understand why your homepage has advisors

How is progress on Arc going?

> Paul Graham: I'm sorry, I'm having trouble connecting right now. Please try again.

He’s dodging!

Probably got overwhelmed, from the HN requests

You're a big fan but have never read his essays on Startup Advice?

Isn't this a little creepy? It's mimicking an identity of someone that you haven't gotten the permission to do. It feels wrong.

Worse, this is achieving levels of cringe our best scientists previously thought impossible.

Scientists have been wrong before

Unethical? Maybe, illegal? No App clearly states that it is an AI Advisor not a real person

I am not a lawyer, but most U.S. states (California, where pg resides, is one of them) recognize a “right of publicity.” So, illegal? quite possibly yes.

You're right the law prevents unauthorised use of Personas for commercial gain

There's a lot of stuff on the Web, I've read few of them but not all.

btw, previously someone attempted the same at paulgraham.chat but it's not a reasoning model, and doesn't do well with additional details

Does Paul Graham have trouble connecting as well?

Got that fixed, thanks

Sorry folks, asked PG what's the most important thing for startups (hoping to hear "make something people want") but got this goobledygook.

Paul Graham A question that gets to the heart of startup success! Based on my essays and experience, here's the most important thing for startups to focus on, followed by a brief elaboration and some secondary focuses tailored to different startup stages:

Most Important Thing: "Validate and Iterate on a Problem with a Large enough Market" (or more succinctly, "Solve a Big Problem People Will Pay For")

Elaboration: Validate:

Ensure there's a genuine problem or need in the market. Talk to potential customers to understand the pain point deeply. Problem with a Large enough Market:

The problem should be significant enough that solving it attracts a substantial user base or generates considerable revenue. Market Size: Ensure the market is large enough for your growth ambitions. Iterate:

Be prepared to refine your solution based on feedback and market response. Pivots might be necessary; being open to changing direction based on validation is key. Why This is Paramount: Reduces Risk: Focusing on a validated problem minimizes the risk of building something no one wants. Focus: Gives the startup a clear direction, helping allocate scarce resources efficiently. Scalability: A large enough market provides a clear path to growth. Secondary Focuses by Startup Stage: Idea/Seed Stage Secondary Focus: Team Building Assemble a balanced team capable of execution, adaptation, and growth. Early Growth Stage Secondary Focus: Scaling Customer Acquisition Efficiently Develop and refine a cost-effective user acquisition strategy to leverage the validated market opportunity. Later Growth Stage Secondary Focus: Operational Efficiency & Strategic Partnerships Streamline operations to maintain profitability with scale. Pursue strategic partnerships to further expand market reach or defend position. Example from My Essays ( referenced in your initial query's text block, though not directly asked about here): Relation to "Programming Bottom-Up" (Chapter 002): This essay emphasizes building upon and validating assumptions iteratively, similar to the iterative validation process recommended for startups. Connection to "Good Writing" (Chapter 229): While not directly about startups, it highlights the importance of clarity and iteration in communication, mirrored in the need for clear problem understanding and iterative solution refinement in startups. Actionable Advice for Startups Based on This: Action How Validate Problem Conduct at least 20 in-depth customer interviews. Assess Market Size Use publicly available market research reports and calculate your potential share. Iterate Based on Feedback Allocate 20% of development time to pivots or adjustments based on early user feedback. Review and Adjust Quarterly Schedule regular team reviews to assess progress against problem validation and market response.

that's what the essays say, i guess

Reply on news.ycombinator.com