let me wander

-

It's not hate – we're scared and frustrated. Not just for ourselves, because AI is a genuine threat to our livelihoods, but also for the next generation of writers, like you. It's going to be a lot harder to get discovered or published with AI-generated content flooding the Internet and the book market.

Earlier this year, it was revealed that Meta trained Llama 3 on a massive body of pirated work. You can read more about it here. Meta employees knew this was morally wrong, but they did it anyway, because (1) they didn't want to pay anyone for the use of copyrighted work, and (2) they knew they could get away with it, and they have. They took our stories, born of real human experiences, and used them to feed something that's designed to be able to replace us. There are other reasons writers may be anti-AI, like the impact on the environment, but hopefully that gives you some context for why writers, specifically, are reacting to this so strongly.

You've said you wouldn't use ChatGPT to write your novel, which is great to hear. If you did, I would tell you that you weren't, in my opinion, a writer – just as I would never claim to be an artist if I used ChatGPT to create images, or a musician if I used it to generate a song. But I would also gently question why you feel like you need it to give you morale boosters or tell you if something makes sense. ChatGPT is not a human reader; unless you specifically instruct it not to flatter you, it will say what you want to hear. It isn't reacting to you, or to your story, with a human gut or a human heart. To me, any praise or encouragement it offers is empty. There's nobody and nothing behind it.

As for asking it to help you work out if something makes sense: I really do understand the temptation. I'm chronically ill, so I write at a slower pace than a lot of my colleagues, and it might help me churn out books faster if I asked ChatGPT to help me unpick a knot in the narrative, or fix a plot hole. But I don't want to surrender the ability to think and problem-solve for myself, and I would caution you against doing that – not just for the sake of your writing, but for everyday life. In this era of disinformation and propaganda, our ability to think, interrogate and analyse the world around us is more important than ever.

Talk to other writers. Make friends. It's not rocket science, it's what a human being is wired to do. Just be kind and friendly and interested in other people and their writing, and they'll be interested in you and yours. And then if you can't figure it out, ASK FOR HELP. There is NOTHING wrong with asking for help. Asking for help is, in fact, a beautiful thing that will bring you closer together with a new friend. That's what you're sacrificing when you turn to ChatGPT for it -- you're losing out on the possibility of making a really profound, lasting, potentially lifetime friendship with another human being. You're missing out on something sacred and beautiful because you're impatient and scared and insecure.

Insecurity doesn't just vanish automatically. You have to file it down gradually over time, like filing your nails or sanding a piece of wood butter-smooth. You can do it. It is WORTH doing. It just takes some elbow grease.

The reblog chain is one of the things that makes Tumblr unlike anywhere else. All the notes on reblogs are attributed to the original post, no matter which branch people actually liked or reblogged. We want to keep encouraging conversations, and give contributors the recognition they deserve. 

Soon, you'll be able to like, reblog, or reply to any part of a reblog chain, and that note will go to that reblog's author. Each reblog will have its own counts, instead of one aggregated number from every version of the post. And yes, you’ll be able to like multiple posts in one chain.

If a reblog doesn't add anything, the love flows up to the last person in the chain who did. Your post doesn't lose notes just because people spread it quietly.

Past notes will stay on the original post — we're only changing what happens from here on out. Retroactively re-attributing all of them would be... a lot.

We rolled out a significant change to how notes work on reblogs, and the reaction has been strong. We're not going to pretend otherwise. 

First things first: We're reversing the change. Your feedback in comments, emails, and especially reblogs, made clear that the rollout created problems we need to address before moving forward. We also should have communicated this differently from the start, and we didn't.

We still believe there's a better version of how reblogs can work. One that gives every voice in a chain the credit it deserves. But we want to get there with you.

In the coming days we'll share more on how we plan to do that,  including ways to work directly with some of you on this and future changes before they ship.