Charlie Stross
Some thoughts on spam:
1. Right now it's a single idiot running a scripted Joe Job. But we know how spam ecosystems develop. The next steps are inevitable.
2. Next, someone will release an activitypub spamming script. It will hammer on servers to create throwaway accounts then use them to post.
3. This will drive a bunch of small servers off the fediverse and cause acrimonious defederation squabbles.
4. Surviving servers will limit sign-ups, requiring proof of humanity (or identity).
/1
Charlie Stross
11. In the end, we will be hit with a deluge of spam simply because Gmail is effective at locking out email spam, USENET is dead, enshittification has killed the utility of the public internet for advertisers as well as users of web search, and because *spamming is cheap*. And unlike X/twitter there's nowhere to build centralized moderation (yet).
12. I'm pretty sure there are viable anti-spam strategies for the fediverse. But the current situation is a wakeup call.
Charlie Stross
8. Right now the spam problem is one idiot gamer in Japan. But the next stage will be sleazeball marketers (anyone else remember Canter & Siegal from the first time round?)
9. There will be toolkits to send marketing spam, not because Mastodon users are the target, but as a route into Threads, Bluesky, and other commercial media with a mass user base.
10. Added lulz: the Mastodon Taliban (who hate Meta) may even start spraying spam just to force Meta to defederate from the wider fediverse.
Charlie Stross
5. Problem: authenticating new user sign-ups as being from non-spammers is tedious AND, thanks to generative AI, prone to being gamed.
6. Added problem: it's too easy to set up a new fediverse server. We're used to treating user accounts as spam sources, but what happens when a script kiddie automates spinning up a new containerized mastodon instance on AWS just to federate and then send spam?
7. So federating new servers is also going to require proof-of-humanity steps.
/2
Alex Gleason
Darnell Clayton :verified:
@cstross I think @dansup@pixelfed.social created a solution for dealing with spam on #Pixelfed that could potentially be ported over to #Mastodon, #Misskey, & the rest of the #Fediverse.
👉🏾 Autospam and Naive Bayes: The Grandfather of Spam Filters Still Making Waves https://pixelfed.blog/p/2023/feature/autospam-and-naive-bayes-the-grandfather-of-spam-filters-still-making-waves
I wonder if @dansup@mastodon.social could have an “anti-spam” assembly with the various software developers using #ActivityPub‽ We could also checkout what #WordPress is doing with Akismet too.
Tim Chambers
@darnell @cstross @dansup@pixelfed.social @dansup@mastodon.social
I also thought that some of what Automattic has done with Akismet and comment spam relating to their #ActivityPub plugin might be able to be used for the more broad fediverse... cc: @pfefferle
Alex Holst
@cstross
The major software has resisted all sorts of safety features over the years, because techbros are gonna techbro.
It's actively harmful software when:
- open signups is the default
- new accounts can go from zero to thousands of outgoing messages
- servers blindly accept large amounts of incoming msgs from previously unseen servers or users
Again, if we had listened to marginalized people years ago we would all have been better off now. Funny how that works.
GNU social JP is a social network, courtesy of GNU social JP管理人. It runs on GNU social, version 2.0.2-dev, available under the GNU Affero General Public License.
All GNU social JP content and data are available under the Creative Commons Attribution 3.0 license.