PHASE 2 OF 7

Ideation

Brainstorm, evaluate, and select ideas that judges love and you can actually build.

The Little Alchemy Method

The ideation methodology behind $100K+ in hackathon prizes. Inspired by the game Little Alchemy, you start with basic elements, combine them, and discover winning ideas.

TL;DR Combine technologies with industries the way Little Alchemy combines elements, and winning ideas emerge.

In Little Alchemy you start with four elements (fire, water, earth, air) and combine them into steam, lava, or life. Keep mixing your discoveries and you unlock hundreds of things, from windmills to the internet.

Hackathon ideation works the same way. Your building blocks are technologies and industries. Combine them, and winning ideas emerge.

TECH ELEMENTS
Your Building Blocks
Tools that do something. Single technologies from companies or research papers.
OpenAI 4o (LLM)
Google Veo 3 (Text to Video)
Twilio API (Phone Calls)
SunoAI (Text to Music)

combine

INDUSTRIES
Your Problem Spaces
Spaces that need something. Industries, domains, and user groups.
Sports
Health
Education
Defense
Finance
Entertainment

The method is simple: List 20-30 elements and combine them at random. Most pairings flop, but the magic is in the volume: the ones that land are often brilliant.

INTERACTIVE

Try It Yourself

Drag elements into the workspace, then drop them onto each other to discover winning hackathon ideas.

0 / 69discovered

Drag elements into the workspace below to combine them

TECH

LLMs
Twilio API
Google Veo 3
SunoAI
Vision AI
Blockchain
IoT Sensors

INDUSTRIES

Health
Education
Defense
Finance
Entertainment
Sports
Climate
Legal

Drag elements here, then drop them on each other

Double-click a workspace element to remove it

Combinations in Action

Here's exactly how the Little Alchemy method produced ideas that won major hackathons. Each winning project started as a simple combination.

TL;DR Every winning project below started as one simple combination, and swapping a single piece spawns a brand-new idea.

CASE STUDY
DispatchAI — UC Berkeley AI Hackathon Grand Prize
How two simple combinations led to a first-place finish.

Step 1

LLMsTwilio APILLM answering phone calls

Step 2

LLM answering phone callsDefenseAI 911 Dispatcher — DispatchAI

Seems simple? That's the point. The best hackathon ideas aren't complex. They're clear combinations nobody else thought to make.

Swap out a single component and the same base combination spawns entirely different ideas:

SWAP: INDUSTRY
Mental Health AI Agent
LLM answering phone callsHealthMental Health AI Agent

Same tech, different industry. One swap turns a 911 dispatcher into a mental health support line.

SWAP: TECH
AI Ad Commercial Creator
LLMsGoogle Veo 3Iterative AI Ad Creator

Swap the industry for another tech. An LLM directs video generation for rapid ad prototyping.

AdaptED

LA Hacks

Best Use of Google

DispatchAI

UC Berkeley AI Hackathon

Grand Prize

TalkTuahBank

HackUTD

1st Overall + Goldman Sachs

All three projects came from the same method: list elements, combine them, let unexpected connections emerge. It works because it forces you to think across boundaries, not within them.

Go Deeper

Optional theory and prep for when you want more. The essentials are above; expand only what you need.

The Ideation Toolkit

Three additional frameworks to break through creative blocks. Use these when random combining stalls or you want to push ideas further.

TL;DR When random combining stalls, reach for SCAMPER, random input, or antidisciplinary thinking to break the block.

SCAMPER
SCAMPER Method
Bob Eberle's checklist to expand any idea in seven directions.
S
Substitute: Replace a component
C
Combine: Merge two features
A
Adapt: Reuse in new context
M
Modify: Change scale or shape
P
Put to another use: New application
E
Eliminate: Remove complexity
R
Reverse: Flip the perspective

Run each letter on your top idea. “Substitute the API? Combine with another app? Eliminate the most complex step?”

LATERAL THINKING
Random Input
Edward de Bono's technique for provoking new connections with randomness.
  1. 1Pick a random noun: open a dictionary, use a generator, or point at something in the room
  2. 2Force-connect it to your problem. How is your idea like a “lighthouse”?
  3. 3Follow the unexpected links. The weirder the connection, the more original the idea

Stuck in a loop of similar ideas? Random input is the fastest way out. De Bono argued the real risk isn't being too random, but not random enough.

MIT MEDIA LAB
Antidisciplinary Thinking
Joi Ito's framework for finding innovation in the white space between fields.

Interdisciplinary work has fields collaborate. Antidisciplinary thinking creates something that fits no existing category: the white space between the dots on a map of knowledge.

Ask Yourself

  • Which field would never normally work on this problem?
  • Does this idea feel uncategorizable? Good. Pursue it.
  • What would happen if a musician designed this? A chef? A game designer?

The best projects feel weird at first. “An AI 911 dispatcher?” sounded strange until it won grand prize. Lean into the weirdness.

“Creativity is just connecting things.
The broader one's understanding of the human experience, the better design we will have.

Steve Jobs, co-founder of Apple

The best ideas don't come from staring at a blank screen. They come from a broad inventory of elements (technologies you've explored, industries you understand, problems you've felt) plus the courage to combine them in ways nobody else would.

The Little Alchemy method mirrors how creativity actually works: not a bolt of inspiration from nowhere, but the systematic collision of ideas from different worlds. Every element is another dot. Every combination connects them.

The more dots you collect, the more connections become possible.

Ideation Checklist

A step-by-step summary you can follow before and during your next hackathon. Print it, screenshot it, or just remember the flow.

TL;DR Build the list, combine, step away, push your top ideas with SCAMPER, then pick the one at the edge of the adjacent possible.

Build a list of 20-30+ tech, domain, and wild card elements

Combine randomly. No filtering, no judging. Just mix.

Step away and let diffuse thinking work: walk, play, cook, gym

Chase volume. Aim for 15-20 unique combinations

Run SCAMPER on your top 3 ideas, pushing all 7 directions

Pick the idea at the edge of the adjacent possible. What just became feasible?

Validate with your team: does it excite everyone, and can you ship it in 24-48 hours?

Remember: Ideation isn't about finding the perfect idea on the first try. It's about generating enough raw material that one great idea rises to the surface. Trust the process, the volume, and your subconscious.