# Innovation Bet Log · 创新下注日志

> Use when a direction feels feasible. The log makes the falsifier, downside, recognition window, and
> human conviction explicit before generation makes the idea look better than it is.

## 1. Bet

| Field | Fill |
|---|---|
| Direction / idea |  |
| Human owner |  |
| Real need observed |  |
| Viable path observed |  |
| Inner conviction source |  |
| What AI helped with |  |

## 2. Worth Compass

| Axis | Current reading | Evidence |
|---|---|---|
| Real need | imagined / observed / paid-for |  |
| Viable path | looks feasible / partly tested / exists |  |
| Inner conviction | borrowed / wavering / deep understanding |  |

Verdict: `do not bet yet / small reversible bet / serious bet / stop`

## 3. Falsifier

Write the sentence before polishing:

> This bet is false if...

| Test | Cost | Timebox | Pass / fail |
|---|---|---|---|
|  |  |  |  |

## 4. Affordable Loss

| Downside | Who bears it? | Reversible? | Cap |
|---|---|---|---|
|  | self / users / team / others | yes / no |  |

If cost lands on others or is irreversible, reduce the bet until the loss is affordable or stop.

## 5. Recognition Window

| Signal to watch | Why it may be a new species | Earliest review | Latest safe review |
|---|---|---|---|
|  |  |  |  |

Rule: generation may polish the artifact after the falsifier is written, not before.