π Newsletter #004 - Of The Multi-Billion Dollar Web, Grammarly's ""Experts,"" and Broken Clocks
β¨ In This Edition β¨
- Short-Form Content: Grammarly ""experts"", dead or alive; Machine Translation & AI; The Fediverse as the Internet we were promised in 1996; RAM & SSD price hikes; broken clocks.
- Long-Form Content: πΈοΈ Caught in a Multi-Billion Dollar Web: Why Leaving the Legacy Algorithmic Media is So Hard
- The Post-Script: Following: Elena Rossini; Reading: Emails from readers!
Original image by Bernard Hermant on Unsplash.
π Short-Form Content
- In a very unethical move, Grammarly somehow thought it was ok to have "experts" (LLM chatbots) "review" their users' writing. Without getting permission from the actual, human experts whose names they "borrowed." Then again, don't know why I am surprised at this development. It's a feature implementation that was possible and doable and that cranks up the "value" of AI, so of course this was a thing that was going to be done eventually.
- Update, Grammarly is now facing a class action lawsuit over the above.
- And it caused Maureen Ryan to write this absolutely π₯ open letter, so I suppose something good came out of it?:
βMove fast and break thingsβ now extends to the reputations and identities of respected professionals and creative people. And you have the nerve to presume we should all put up with every outrage, with every act of destruction, with every theft.
- In this excellent thread, Eugen Rochko explains why "but we wouldn't have machine translation without LLM's! Gotcha!" is a wrong and also irrelevant argument brought forward by AI hypebros when faced with gen AI criticisms. Two excellent quotes from his thread:
From what I've observed, people who claim that LLMs can replace artists don't understand art, people who claim that they can replace musicians don't understand music, people who claim that they can replace writers don't understand literature, and people who claim they can replace translators don't rely on translations.
Technology is not inevitable. We've decided not to have asbestos in our walls, lead in our pipes, or car[c]inogenic chemicals in our food.
- Mat Duggan wrote a really cool piece titled "Boy I was wrong about the Fediverse." It contains gems like "To thrive in capitalism one must be amoral" and also Greenlandic toddler tales, so you're probably going to want to read this. Also, the Fediverse is a great thing; come follow Technically Good!
[T]here's something quietly beautiful about a place where people just... share what they know. No brand deals, no engagement metrics, no algorithm nudging you toward rage. Just someone who spent twenty years studying Arctic policy posting a thread at 2 AM because they think you should understand what's happening. It's the internet I was promised in 1996.
On that note: No, targeted ads don't just mean that you're getting to see content that "matters to you." They are also being used by law enforcement to track people's movements without their knowledge. Surpriiiiiise!
The price hikes for RAM and storage are, I am sure, a reasonable side effect of the fact that AI companies simply need more and more storage to deliver on things that will definitely happen some time in the future, and they have absolutely nothing to do with AI companies buying up bare metal storage and compute so that users will have less on-prem storage and compute and will have to pay AI and cloud companies for their monthly storage and compute subscriptions.
Lastly, I discovered this excellent analogy by Riley S. Faelan:
The notion of a broken clock being sometimes right is based on a gross misunderstanding of what information is. A clock that always shows the same time is never right, even in the moments of the day when the time happens to be what it shows, because you don't gain any information about what time it is by looking at the clock. This reasoning also applies to chatbots. If you can't tell whether what you have been given is useful information unless you alrea[d]y know the information, then you haven't been given useful information.
π° Long-Form Content
πΈοΈ Caught in a Multi-Billion Dollar Web: Why Leaving the Legacy Algorithmic Media is So Hard
This post was a follow-up to Feeding the Fire, the first TG article that seems to have really hit home with a lot of folks.
You can read the full article on network effects, switching costs, and the collective action problem here, but the tl;dr is:
- Even people who know that the established networks are bad and do not serve them anymore, and people who want to leave them behind, often stick around. Why is that? Sure, these platforms often still βworkβ in a technical sense. They offer a familiar UI and familiar features, and are backed by beefy servers. But the more critical point is that users have built up large and/or important social networks on them over time.
- Network effects and switching costs tied to those networks then lock users into these platforms, even if they continuously enshittify. Since a platformβs value rises with the number of users on it, leaving means losing connections, and often years of data, without ways to migrate the data elsewhere.
- Even if individual users leave the platform, that does not usually break the network - without easy interoperability, people often get pulled back into the old networks, as if by a rubber band.
Find suggestions on how you (yes, you!) could be the first to start a new network effect and break this ugly cycle (as well as break free of the chains of commodified attention, ad tech, and rage-inducing algorithms.)
π―The Post-Script
Following
- This excellent thread by Elena Rossini, essentially cobbling together a list of smart folks publicly resisting and calling out LLMs and other slop. This has been a treasure trove of great new accounts to follow, both in the post itself, as well as in responses from users.
Reading
- Emails from readers! It has been so refreshing to receive emails from folks who have been enjoying the Technically Good articles and newsletters. Quite frankly, I did not expect this many responses, especially after having run this side project for only two short months. It reminds me that true interaction on the web is indeed possible, and does not need to be mediated by million-dollar corporations with engagement-farming algorithms. We have momentum right now, and it feels like something is shifting.
Recent Posts