NextGEN Vibe-a-thon — Planning Hub

The idea in one line: instead of another AI panel, we run a room where every attendee does one real thing with AI on their own business — hands on keyboards, a guide at every table, and something they can take home. Show, not tell.

The one thing to nail first (from planning meeting #1): define the goal + the aha moment. The group has strong ideas pulling in several directions. The emerging north star — a founder walks out saying “I spent 4 hours a day on that; you just showed me how to do it in 20 minutes.” A mindset shift > any single artifact.

Sept 16

Summit Day

Match Day + Reception on the 15th

~90 min

Target Block

Post-lunch energy reset (leaning)

1 : 6–8

Facilitator Ratio

One expert per table

Artifact Each

Everyone ships something real

Plan Phases

Alignment → event day

The proposal for meeting #2

decision-ready

Recommended goal: Every attendee leaves feeling capable — AI is more approachable than they thought — with one useful thing in hand, worked on their own real problem. Neither a hackathon (no winner) nor a course (no lecture) — a mindset shift with a personal win attached. Two shorter alternates in the working doc.

Three lanes — everyone gets a comfortable option

CORE

1 · Hands-on tables

Small groups, bring-your-own-problem, facilitator-guided. Where most people land.

2 · Safe stage

A live demo on the main screen + a short peer-leader story — for anyone who'd rather watch than perform.

3 · The patio

No-pressure networking off-ramp. Self-selecting out is a fine outcome, not a failure.

How a table runs

✓ Bring your own problem, around a theme — not a canned scenario. "Bring a thing, leave with a thing."
✓ Facilitators guide a process, not a script — keep it simple; over-engineering is the failure mode.
✓ Everyone starts with a 5-min warmup — "is this even a good problem for AI?" (teaches problem selection).
✓ Train on ~2 scripts, flex live to who shows up (beginner ↔ daily user); split a crowded table.
✓ Own device, own data — nothing shared, nothing exposed to the table.

Run-of-show · ~90 min, post-lunch

0:00–0:05	Frame it: the goal + your three options (table / stage / patio)
0:05–0:10	Warmup together — "is this a good problem for AI?"
0:10–0:20	Get seated — tables by theme, watchers to the stage, patio drifts out
0:20–0:65	Build — facilitators flex ~2 scripts; stage demo runs in parallel
0:65–0:75	Checkpoint — everyone has something, even rough
0:75–0:85	Light close on the screen — a few highlights, no "pick a winner"
0:85–0:90	Wrap: where to go next; patio stays open

5 table themes (starter menu)

Strategy & Documents — bring an RFP/strategic doc; leave with analysis + a drafted response.
Data & Spreadsheets — bring a real spreadsheet; leave with it analyzed or automated.
Marketing & Content — one idea → posts, site copy, or a campaign.
Operations & Automation — automate a repetitive workflow.
Experience & Communicate — no build: call an AI agent, prompting-as-communication, pitch practice.

What to decide in meeting #2 — in order

Sign off on the goal — one sentence everyone can repeat.

Confirm the three-lane model — hands-on / stage / patio, and that the stage lane gets real screen time + a peer-leader lined up.

Lock the 4–5 table themes — finalize the menu + pick which 2 scripts get built first for facilitator training.

Confirm footprint — 90 min, post-lunch, one room (25–30 tables possible, ~15 facilitators — plan for flex/split, not full coverage).

Registration survey — founder? · AI level? · pick a business-problem category · describe the problem you'd bring.

Tool equity — chase the sponsored shared AI workspace (Anthropic / Atlas Local lead)?

Who's actually in the room

from meeting #1

Audience mix (last year's manual tally)

Founders / startup staff	~50%
Community members	city officials, program leads, supporters
VMS mentors	tech founders 20 yrs ago; rusty now
Ecosystem partners	~10%
Investors & sponsors	the remainder

Skill reality — it's bimodal, not beginner

≈ half already use AI most days but “don't fully understand what's going on.” A real minority has no exposure and will fear being called out. Very few are truly AI-illiterate.

Design implication: not "teach beginners" — level up the daily-dabblers AND give first-timers a graceful on-ramp (plus a no-pressure networking off-ramp for anyone who opts out).

What "success" means

5 criteria

Fast first winEveryone produces something usable in under ~20 minutes. Momentum is the whole game.

It's theirsThe artifact is built around their real business — not a canned exercise.

Take-home recipeThey leave able to redo it themselves. Capability, not a magic show.

One collective "whoa"A reveal moment that ripples through the room — energy + social proof.

No one left behindThe least technical founder still ships. Facilitators drive the hard parts.

What it's not — and how it could fail

Non-goals (guardrails)

✕ Not a competitive hackathon ✕ Not a lecture or panel ✕ Not AI theory ✕ Not a shippable product ✕ Not a sales pitch ✕ Not comprehensive ✕ No coding skill required

Inversion — "how would we guarantee failure?"

Failure move	→ Design requirement
Weak wifi + live logins at the start	Pre-provision everything; hotspot fallback
One expert demoing to a passive room	Every attendee's hands on a keyboard
A canned problem nobody cares about	They bring a real problem
Choice paralysis across many tools	One opinionated stack
No facilitator per table	1 : 6–8 + roaming helpers
Jargon that makes founders feel dumb	Plain-language recipe cards

Format options we explored (earlier)

superseded by the proposal above

RECOMMENDED

Table Build-Sprints

"Build Your Co-founder"

Each table (6–8 founders + 1 facilitator) builds one working artifact in the time box, then shares it in 60 seconds.

Hits all five success criteria; tangible personal payoff; a timed reveal at the end.

Live Solve

closest to the "data-walk" spark

Crowdsource real founder problems, then facilitators solve a few live on the big screen while the room replicates.

Purest "show, less tell." Most facilitator-dependent; less individual output.

Speedrun

hackathon adrenaline, compressed

Timed challenge — everyone builds the same category of thing with a countdown, then rapid-fire show-and-tell.

Great built-in reveal; needs a tighter skill floor to avoid leaving people behind.

Where it landed: these three were the early options. The room converged on the three-lane model in the proposal above — hands-on tables (bring-your-own-problem) + a safe stage lane + the patio — with facilitators flexing ~2 scripts live. Kept here as the trail of how we got there.

What does each founder build? — decision matrix

open

Artifact	Setup friction	Reusable Monday	Reveal "whoa"	Non-technical	Any business
AI assistant / agent rec	Lowest ✓	High ✓	Abstract	Yes ✓	Yes ✓
Landing page / site	Higher	Medium	Highest ✓	Yes ✓	Many have one
Time-saving automation	Highest	High ✓	Low	Hardest	Varies

Suggested default → an AI assistant for their business. It's the only option that doesn't fight our biggest risk (setup friction), works for everyone regardless of tech level, and is reusable immediately. Its one weakness — a quiet on-screen reveal — is exactly what the live board below fixes. Strong visual side-track: an interactive HTML one-pager (report / data-viz / prototype) — AI is remarkably good at these and it's an easy universal takeaway.

The live collaborative board

one tool, three lives

A single shared board (one link) that carries the session end-to-end — and doubles as the on-screen "whoa" moment for the assistant-build format.

1 · Plan

A shared idea wall for the planning group to co-develop the format now, before Sept 16.

2 · Live

Attendees drop what they built + a screenshot onto a wall shown on the room's single main projector — instant social proof.

3 · Recap

Freezes into a shareable post-event page: everything built, plus the recipes to take home.

Built to start simple (a reliable submission wall) and add real-time multiplayer only if we're ahead of schedule — never at the expense of the event.

The plan — 6 phases

click to expand

Confirm reply + planning meeting high

Lock footprint (block vs breakout), slot/room, expected headcount high

Confirm audience tech profile (design default: mostly non-technical) high

Present format + artifact recommendation; get buy-in high

Align with Peter on the facilitator pipeline high

Agree success metrics + non-goals so scope doesn't creep med

Open questions for the group

let's decide together

Define the goal + the aha moment — one sentence we all agree on. This is the open item from meeting #1; everything else hangs off it.Emerging: a mindset shift + a real time-save win ("hours → 20 minutes").

Footprint — confirm ~90 min, post-lunch, and the final table count.Leaning that way; needs sign-off.

Table menu + scripts — which ~6 takeaway themes, and who drafts the A/B/C facilitator scripts?~15 facilitators already opted in — poll them for top-3 topics to seat tables.

Tool equity — can we land a sponsored shared AI workspace so everyone's on the same model tier?Warm lead: someone from Anthropic works out of Atlas Local.

Grouping data — add the survey (founder? AI level? topic pick) to registration vs. the 6,000-person newsletter?Require at registration for higher capture; tickets sell late, so start now.

✅ Resolved in meeting #1: same venue (Judson Mill) · one main projector, no per-table screens (it's the shared wall) · wifi present · 8-seat tables · ~15 facilitators lined up · audience is bimodal, not beginner.