Star ratings for media wokeness

A design document

So far the articles on this Substack have been for people who are running an institution. This article proposes a project that anyone could do. It also introduces a simple document structure that is commonly used in the tech industry for planning projects. You might find it useful.

Introduction

Everyone consumes entertainment (films, TV, video games, books) but many consumers find content that pushes overtly ideological messaging distracting, distressing, insulting or to suffer from incompetent production. There’s plenty of non-woke media out there to enjoy but discovery is an unsolved problem. Existing content platforms like Amazon, Steam or Netflix don’t provide any way to filter content by degree of social engineering. A small cottage industry of reviewers has emerged on YouTube who rate content by conventional metrics like story quality rather than ideological compliance, but reviews are scattered and finding content they recommend can be slow, requiring watching time consuming videos.

There should be a way to quickly find media rated not only by general appeal but also level of woke messaging and competence in storytelling. This would help people direct their money to content creators that align with their values.

Goals

1. Define standard criteria for rating media wokeness and competence.

2. Collect and publish user-sourced reviews that follow the standard criteria.

3. Going from “I want to watch something good” to finding an acceptable title should take less than three minutes.

4. It should be possible to quickly filter out media with ideological messaging or incompetent production.

5. Generate revenue via subscriptions, ads or affiliate schemes.

Non-goals

1. Selling content. Titles should link directly to the most popular online stores where such content can be acquired.

2. Long-form reviews. People who want in-depth reviews of content from a non-woke perspective can find them on YouTube or other sites. Reviews should be concise and to the point: think Amazon-length.

3. Reviewing only based on wokeness and competence. It’s important to have ratings of general appeal also, as the goal is to find something not merely un-woke but also good.

4. Reviewing non-entertainment content like news, documentaries etc. Comedy TV is in scope.

Design: Phase 1

Although a dedicated website would be nice, the most important thing is to kickstart things by having enough reviews to be useful. With a seed of reviews being regularly added other people may join in. A simple and quick way to start is by using Substack itself. If things take off a separate website can be introduced later.

A Substack devoted to media reviews should have the following characteristics:

• Coarse grained scores (e.g. out of 5). This allows each different score to have a written criteria, helping reviews be consistent with each other.

• To ease scanning of articles, reviews should use a standard title format that exposes in a consistent manner:

  • The name of the media.

  • The wokeness score.

  • The competence score.

  • The general appeal score.

• A systematic approach to tagging (for genre, actors …). This doesn’t have to be Substack’s tags, it can be as simple as using #hashtags in the body that could show up in search or be easily indexed by a separate website.

• Reviews should be submittable via email, direct chat or by having the review syndicated from another Substack it if meets the criteria.

• Reviews should have a reading time of one minute or less.

If using Substack to generate revenue it would make sense to make reviews free after a few weeks so the archive can be browsed by everyone, and the service is then paid for by people who want fresh reviews of newly released content.

If there is one article per title, then the comments may fill up with people’s thoughts. That’s fine, but comments that match the standard review template and come from “high effort” accounts (i.e. people who make real comments or run their own stack) can be lifted into the article itself and used to update the average scores, expanding the range of verified reviews.

Design: Phase 2

At the start you’d be working alone on an empty canvas. To speed things up you can use AI to quickly generate reviews from YouTubers who specialize in reviewing from an anti-woke perspective.

There are many ways to do this but a quick approach is to use sider.ai, which is a Chrome extension. It adds an extra button to YouTube that can quickly transcribe and summarize videos. Here are a few examples from The Critical Drinker:

The Acolyte Episode 3:

Generated review:

Episode 3 of "The Acolyte" is criticized for its poor writing, cringe-worthy moments, and lack of respect for the Star Wars legacy. The narrative features strong female characters in a convoluted plot that undermines established lore, leading to disappointment among fans. The creator's vision is seen as detrimental to the franchise's future.

It’s a pretty faithful summary, although the model didn’t pick up on the sarcastic way the Drinker uses the phrase “strong female character”.

Here’s one he liked:

Generated review:

Twisters, a sequel to the 1996 disaster film, follows Kate and her team as they confront a powerful tornado while testing a new theory. The film balances entertaining action with character dynamics, though it struggles with focus and originality. Overall, it serves as a fun summer flick amidst a lackluster season.

The above reviews were generated with the default settings. With slightly more advanced use of AI a stricter focus on the standardized review criteria could be obtained, and with really advanced use of AI it could be used to directly review the titles themselves.

Although having some computer skills helps at this point, it isn’t strictly needed. You could benefit from these tools without doing any programming.

Design: Phase 3

As the collection of reviews grows so will the need to search them, filter them and maybe generate personalized recommendations. At this point the strategy of (ab)using Substack is likely to run out of road. Substack has these features but not within a single stack, and the implementation isn’t great.

At this point a separate website would become useful. That doesn’t mean abandoning Substack entirely. The separate site can provide the missing features whilst the reviews and comments remain hosted on Substack. It may also be worth splitting up the service into multiple stacks by genre at this point, so people can subscribe to feeds specific to the type of content they like, but it may also be worth giving up on Substack’s discovery features and going all-in on the separate implementation.

If the project ever reaches this point how exactly it works will be a nice problem to have, and isn’t worth worrying about too much up front.

Alternatives considered

1. It could be a standalone website from the start instead of using Substack. This would have branding and feature benefits, but require a lot more work up front and divert attention away from writing / generating reviews to implementing the service.

2. It could focus exclusively on one kind of media, like only movies. But for the project to gain critical mass requires volunteers and there doesn’t seem to be much benefit in restricting who can contribute up front.

3. It could be purely a blacklist of media that is considered too woke to be enjoyable, but this approach can’t ever be comprehensive and doesn’t help you choose what to consume, only what not to. Additionally people have different tolerances for ideological messaging, but critical mass requires being a broad church.