Lizzie Od. Independent AI Writer & Editor

Welcome

Hi, I'm Lizzie Od.
I write about the AI
people talk to —
and I use it first.

Independent AI Writer & Editor

Selected writing

From the portfolio

First-hand reviews and editorial essays. My flagship piece is a reproducible independent study — start there.

Independent Study

Benchmark · Cycle 1

I Blind-Tested Five AI Roleplay Apps for 100+ Hours. Here's the Scorecard.

An independent study I ran with my team: same ten characters, same scripted lines, blind judges that never knew which app was which. A reproducible ranking of ourdream, Character.AI, JuicyChat, SpicyChat and CrushOn — with the full method, the diagrams, and the report to download.

Independent study on lizzieod.com / July 2026 / 12 min read / PDF & data included

Read the study
Overall score0–100
  1. 1ourdream81.9
  2. 2Character.AI67.3
  3. 3JuicyChat60.3
  4. 4SpicyChat59.2
  5. 5CrushOn52.7

Method note

Judge a companion on message 30, not message 3

The hidden probes I build into every script — consistency traps, memory drops, and the out-of-character question that makes apps name their real model.

lizzieod.com · Jul 2026

How I work

What you can expect from me

Editorial you can check — because the best test of a claim is that someone else can re-run it.

i.

I use it myself

I pay for the products and live in them for hours as a real user would. When I review something, the take comes from my own sessions — not a spec sheet.

ii.

I build the ruler

When a claim needs proof, I design a fair test — identical inputs, a fixed rubric, blind scoring. If the benchmark I need doesn't exist, my team and I make one.

iii.

I show my work

Method, data and limits ship with the piece — including any conflict of interest, stated plainly. If you disagree, the recipe is public. Re-run it and check me.

Working on something worth writing about?

Get in touch