this post was submitted on 06 Oct 2024
1186 points (99.0% liked)

Memes

45655 readers
2467 users here now

Rules:

  1. Be civil and nice.
  2. Try not to excessively repost, as a rule of thumb, wait at least 2 months to do it if you have to.

founded 5 years ago
MODERATORS
 
you are viewing a single comment's thread
view the rest of the comments
[–] [email protected] 89 points 1 month ago (6 children)

there needs to be a crowd sourced product review and maintenance website that can see trends of enshittification.

[–] [email protected] 43 points 1 month ago (2 children)

From Wikishittia, the free enshittepedia

[–] [email protected] 14 points 1 month ago (1 children)
[–] [email protected] 1 points 1 month ago

I wonder if they'd mind someone mirroring their content, but with the one difference that anyone can edit, any time with no restrictions, spam blocking, vetting etc

See what chaos ensues

[–] [email protected] 2 points 1 month ago

How dare you get my hopes up.

[–] [email protected] 26 points 1 month ago (1 children)

Let’s say everyone used an identity verification service to signup, like had to send photos of their ID and their SSN (national identity number) to be vetted by a third party.

How long after the service got popular would it take for the most aggressive marketers to pay rings of fraudsters to lend their identities and/or make fake reviews?

I think it would definitely start out great until it got big enough to be super useful and then the fraud would ramp up. I think an organization like Consumer Reports has a chance at successfully maintaining a low-bias product database, but the paywall is a big obstacle, as is the fact they’ll only review the largest product catalogs.

[–] [email protected] 10 points 1 month ago

These are the pitfalls with the "amazon reviews/yelp" model.

A decent implementation of the Wikipedia/FOSS model sidesteps this because it theoretically is run by opinionated curators. No amount of bots/shills can break the article soft-lock ounce foul play is spotted.

That's not to say these systems haven't been occasionally broken through more sophisticated attacks, but empirically it seems clear that the model generally works well enough given enough community engagement (which would be the biggest challenge IMO, because maintainers can't be expected to buy every product, and reliable primary sources may be hard to come by).

[–] [email protected] 3 points 1 month ago

I think it would need to be a subscriber service paid for by consumers who are willing to pay for good reviews. Otherwise the consumers become the product and eventually marketers take over.

Also crowd-sourced reviews are what we're supposed to have already, both on Reddit and Amazon (and anywhere else).

What I envision would be a publication that funds a set of reviewers (maybe a mix of full time and part time, since some products are appropriate for testing as a job while others are more appropriate to just use for a while).

Each product would either be bought by the org directly, or if manufacturers provide review samples, a layer of indirection is used to avoid the reviewer feeling like they need to give a good review to keep the free shit coming (with clear communication to the supplier that free or not well have no effect on the review).

Any issues get included in the review fairly, along with any kind of resolution (which should ideally go through both consumer channels as well as reviewer back channels, the former to show what average customers should expect, the latter to hopefully resolve design flaws).

The reviewer will then keep the product and give updates, either in the form of "still using it and it is like x after y months/years", "doesn't get much use because I'm using this other thing instead because of x, y, z", or "doesn't get much use because I'm not really part of the target audience".

My complete vision includes brick and mortar locations where products are available to try out, and maybe sales handled there, where any product available has a "we vouch for the quality of this product" where flaws are highlighted as much as features are.

Though I think the idea is self-defeating because if it gains momentum, it could halt or reverse enshitification and make it redundant, fail, then enshitification returns. Ideally, enshitification is stopped with legislation about quality and enforcement that questions why a bad design is used when a better one is obvious.

[–] [email protected] 2 points 1 month ago

The trick is designing the thing in such a way as to resist infiltration by astroturfing marketers.

[–] [email protected] 2 points 1 month ago

I mean it more or less follows a line. It getting ever steeper lately but it's pretty predictable.