Posted 4 single-line, question-led comments on posts in Rahil's space (Starter Story founder long-game post, AI accessibility workshop for non-technical experts, AI Product Mgr on Bloom's two-sigma and Socratic dialogue, economist on student opting out of lectures for ChatGPT). All different creators from prior runs.
Author: Pat Walls (Following, Founder, Starter Story, sold to HubSpot at 35, large founder-content audience)
Post: 'This guy is 27 and thinks he failed at the game. When I was 27, I was $60K in debt and had just come up with the idea for @starter_story. By 35, I sold my company to @hubspot and I'll never have to work again. This is everything that's wrong with social media. Most people overestimate what they can do in a year and underestimate what can happen in 10 years.' 7 reactions, posted 39m before run. Surfaced via Top feed.
Theme: Founder long-game framing, public scoreboard vs ten-year arc, social-media distortion
Comment posted:
The line about overestimating one year and underestimating ten years lands hardest because the public scoreboard always runs on the wrong timescale, do you find founders internalize the ten year view before or after their first real plateau?
Author: Parisa Shademan (3rd+ connection, AI for non-technical experts, Founder, Investor, 'I turn your expertise into AI-enabled systems that scale beyond your hours')
Post: 'I'm starting something new: Coffee & Claude. Every fortnight, I open up a free 45 minute Google Meet for anyone who has a question about AI. There's no agenda, no slides, and no pitch at the end. Just 10 people and honest answers in plain English. ChatGPT, Gemini, NotebookLM, Claude... whatever you've heard about, whatever you've tried, whatever made no sense when you read about it. Bring it.' Audience are non-technical experts who just want five minutes with someone who'll skip the tech speak. First session 20 May, free signup via lnkd.in/ewxFg2hF. 42 reactions, 15 comments, 1 repost. Surfaced via Top feed (Harshil Tomar comment).
Theme: AI accessibility for non-technical experts, skipping the jargon, small-group session format
Comment posted:
Skipping tech-speak is the whole reason most non-technical experts never get past their first AI attempt, do you find the questions in your sessions cluster around a few common blockers or are they truly all over the map?
Author: Adam Helbig (2nd connection, Senior Director of AI Product Management, '$16M+ in AI-Driven Savings, GenAI, LLMs & Agentic AI at Fortune 20 Scale, Open to EU Relocation')
Post: 'For sixty years we've known the answer to fixing education. Bloom proved it in 1984: one-to-one tutoring with mastery learning makes students perform two standard deviations better than classroom instruction. We just couldn't afford to scale it. Until now. AI doesn't create something new here. It brings back the oldest, most effective form of learning ever practiced: the Socratic dialogue. The one-to-one conversation where a student has to think, defend, and own their reasoning. The tools exist. The research is conclusive. Georgia Tech is proving it right now with 5,000+ students. So why aren't institutions moving?' Linked article: 'AI Finally Solves Education's 60-Year Holy Grail.' 4 reactions, 5 comments. Surfaced via search 'AI tutoring students' past week.
Theme: Bloom's two-sigma problem, AI restoring Socratic dialogue, institutional resistance
Comment posted:
The Socratic-dialogue framing is the part most AI-tutor discussions miss because they default to retrieval and explanation rather than defending reasoning, do you find universities resist the shift because it threatens the lecture format itself rather than because of the AI?