Claude’s NEW 3.5 Sonnet Model Has Authors Upset

Поділитися
Вставка
  • Опубліковано 21 лис 2024

КОМЕНТАРІ • 84

  • @billywhitaker6658
    @billywhitaker6658 21 день тому +19

    I find Claude to be the best writing assistant. I’ve tried gpt, Pi, and many smaller projects. Claude is like having a real conversation about your story. You tell it to stay within the bounds of your descriptions. The project feature is a game changer for authors.

    • @aouyiu
      @aouyiu 19 днів тому +1

      Gemini 1.5 Pro/Ultra seems even better

    • @DWSP101
      @DWSP101 19 днів тому +2

      @@aouyiu no thank you I’ll stay away from Gemini. I’ll stick to Claude although.
      I use a combination of AI with different aspects each one has their cons and pros if you use multiple AI but you know what they’re good at. You can use some to get the best of both worlds might take a little bit more work but putting a little more effort always yield better result.

    • @CeriClark
      @CeriClark 19 днів тому

      ​@@aouyiuI tried Gemini and unless you prompt really hard it gives old fashioned and stilted prose. It's littered with AIisms. But that might be the text watermarking which I don't like the idea of either. Claude does need finessing as well but is a lot easier for me to write prompts. The projects really help.

    • @adstebbing
      @adstebbing 15 днів тому

      The bigger the project the more mistakes it will make. For instance if I use a complicated prompt in a large project (50% consumed by a single file) it will be littered with mistakes. With multiple files for instance pulling elements from Outlines, Character Reference Sheets things get worse, they are often out of sequence, repeating elements or just plain wrong. The worst example was having the same text almost verbatim being used in 5 consecutive scenes, none of which used any of the scene beats that were constructed. I find I constantly have to delete elements from files, keeping the context restricted to the previous, current and next chapter otherwise the output is to put it nicely, "trash".

    • @billywhitaker6658
      @billywhitaker6658 15 днів тому

      @@adstebbing Wow, that hasn't been my experience at all. There's always the hassle of watching your content window limit but Claude will start to warn you when the conversation is getting long so around that time I load the partial chapter as another file in the project and work in a new session. In my projects folder I have multiple Outlines, character sheets, chapters and other various descriptions. Another really important aspect is the "guideline" section where you define its role to your specifications. In this portion I always make sure to insist that it never adds plot elements or dialogue that hasn't been specifically defined by me. Like all the AI models there is some finesse involved but we are in a really good spot for writers.

  • @drewmc2001
    @drewmc2001 21 день тому +10

    Great post. I found new Claude to be superior to old 3.5 if I broke my scenes into manageable beats and asked it to write 2-3 beats (seeking around 150-200 words per beat) at a time. But the days of getting 1500 words per prompt seem to be over, more or less. The quality of the prose is marginally better than old 3.5 most of the time.

  • @dulcerolindeaux9700
    @dulcerolindeaux9700 21 день тому +7

    Because I generate beat by beat, the only thing I noted was the improvement in the prose. I have been very impressed and satisfied with the 3.5

  • @JamezGrimm
    @JamezGrimm 21 день тому +12

    Had some interesting stuff happen when doing scene beats. Claude started cutting off after 300 words and saying like [the rest of the scene will be continued at…. ] I can’t remember what it said but it was essentially saying that it was happy with what it did and didn’t feel like it didn’t need to do more. I was using Novelcrafter at the time 😂 was hilarious and the scene beat cost me ¢0.18 I’d rather use GPT at this point

    • @baronnashor158
      @baronnashor158 20 днів тому

      what are you using now if not Novelcrafter? just curious

    • @aouyiu
      @aouyiu 19 днів тому

      Gemini 1.5 is great and not too expensive

  • @JohnnyStroud
    @JohnnyStroud 21 день тому

    Great video. I have been up since 2am. The work was completed by 0530. I just finished having this version roll out a script from a fully fleshed-out, outline. I use the Save the Cat, 15 beat method, and the software perfomed, admirably. However, I work by sending in the Cast with arcs, ONE beat at a time after relaying that ALL 15 beats are coming. Thanks so much for all of your work and instruction!

  • @philipashane
    @philipashane 21 день тому +9

    Did you check if you’re in “Concise” mode. I think that may be the problem. It’s currently the default for some users.

  • @ariesmarsexpress
    @ariesmarsexpress 21 день тому +3

    So with the most recent Sonnet 3.5, it's not going to be immediately apparent what the changes is, but I used it daily before the change and after. The biggest change is the level of language its using. Before, you may have had a slight hint occassionally of purple prose, the only way you are getting that now is if you are telling it that is what you want. It's going to be much dryer emotionally as well. I would suggest dropping to opus and only using Sonnet for final refinements. I don't have any negative feedback on the model at all, but you just have to get used to what its writing depending on your genre. In the actual Anthropic UI which is the only thing I use, I have never had any issues get 2500 words and ever more occassionally. The one complaint I do have overall is that there is a size limit for the Claude projects which I do not have on the Google Gemini Advanced side.
    The thing to know about it is that as rule, its not going to add extraneous stuff to get to a word count if it can tell the story with fewer words better...which is of course the goal of writing.

    • @gigartina
      @gigartina 20 днів тому

      I found it much dryer, too. And it definitely shows the difference reading through a text where the model changed two-thirds of the way through. It’s a different author voice coming through, even with my sample text (i use novelcrafter) in the codex.

  • @adstebbing
    @adstebbing 21 день тому +7

    The new version is awful, outputs are ofen cut off early. Assuming you can get one, I often find myself in a doom loop where I give an instruction, it ignores it and then asks a question. Even if you tell it not to ask a questions and move on, it keeps rewriting. Yesterday I had a issue where it would give an output then recognise it had made the same mistakes. I had this 7 times in a row, I'm not a fan when it started using sarcasm. I can't even get it to write 500 words at times.Typically the outputs get smaller as the length of the prompt gets bigger. So if your writing your own scene beats and want to flesh them out you can end up with a summary that is shorter than what you provided it. Other odd behaviour was having Claude turn several scene beats into 5 scenes within a chapter, the only issue is that there simply wasn't enough material to spread over that length and no need to split into indivdual scenes, three "scenes" it created were short pieces of dialogue in a conversation so should never have been split. As a productivity tool I've abandoned it, previous prompts giving 1500-2000 words are now giving around 700, I have realised that you can tell it to check the output against the prompt to ensure all elements are accounted for, doing this meant it did actually complete the task set. However my big issue with this is that if your paying for a pro sub you are being penalised having to use 3 prompts to get the same output that the previous version gave you. It eats up your limit super quickly. The only way I can use my old prompts is via OpenRouter though it still takes at least 2 attempts, but if you use the webui you end up having to edit the additions back into the generation because when it used artifacts it proceeded to remove content for the addition. My prompts include a instruction to refer back and ensure all key elements have been addressed. However, after generating an output if I ask it to check if all are addressed (as I can see its skipped 50%+), my next message is to check and produce the omissions, it then wants me to confirm, then even after it creates the additions I can see there are further omissions, in the worst instances it literally rephrased part of the original prompt but did so in a way that is worse than what I had originally written.

  • @chasisaac
    @chasisaac 20 днів тому +1

    So if I’m working on the story and I have a good outline, I can get Claude the right 3000 words in a chapter of regular and inconsistent basis. This new one I get no more than 2000 words. But I also specify length of the overall project and of the chapters. I also asked it to give in the outline how many words per section and it manages to follow

  • @ariekanibalie
    @ariekanibalie 18 днів тому

    I also found the new Sonnet to be stingy with the output, but after tooling with it over the weekend I have to say I'm stunned with the prose quality, provided you prompt it right. Old 3.5, for me, seemed to only ever want to do dramatic scene (action, dialogue, description), but the new one seems far better at inserting back story and alternating scene and summary. Far more natural, flowing prose. I now feel this is well worth the inconvenience (and cost) of having to generate longer pieces in parts. I also suspect these new models may require a different approach when prompting - despite the ever larger context windows, I feel I've had better results when I don't overload it with style instructions, but rather let it do its thing with a good basic style prompt, and then refine the output myself in editing - new Claude is much less inclined to produce typical AI fluff and filler if you prompt it right, so the reduced output may be a good thing too. I'd rather have a very solid draft, if sometimes sparse that I can expand and line edit myself than having to delete all the fluff and glut while trying to maintain flow and pacing.

  • @jimmy6154
    @jimmy6154 21 день тому +5

    Claude may be viable for shot stints but its is plagued by limits both in the chat and project potion. Nothing more frustrating then working on a project to then run into a "you can chat for the next 6 hour" kind of thing and on top of that, there is no way to tell you are encroaching this limit. Until they remove these limitation, Claude is not a recommend imo.

    • @Mr-Not-Applicable
      @Mr-Not-Applicable 9 днів тому

      I dumped Claud a while back due to these ridiculous limits. Far behind competitors in limits

    • @Tony_Calvert
      @Tony_Calvert 4 дні тому

      @@Mr-Not-Applicable What do you use instead now?

    • @Mr-Not-Applicable
      @Mr-Not-Applicable 4 дні тому

      @@Tony_Calvert Google Gemini

  • @AboutMe744
    @AboutMe744 21 день тому +3

    Was going to write a message asking for this very video. It’s been such a downgrade, terribly upsetting because I went from using it daily for helping me edit (not necessarily write) and it’s like talking to a child now. I really wish they didn’t nuke the creative writing part of this and separated out the computer science stuff from the thing that made it perfect for stories.

  • @peterstierjr
    @peterstierjr 20 днів тому

    I found going in smaller chunks yields pretty good results. However: Seems to reach overall limit for me faster (ie cannot continue until such-and-such time, and I'm a paid subscriber)

  • @BruceWayne15325
    @BruceWayne15325 21 день тому +2

    I personally don't mind that it doesn't write out tons of text. Honestly the more you have AI write, the more editing you need to do on the back-end for many reasons. That's completely ignoring the joy aspect of writing, which honestly is the biggest reason for me. I write because I love writing, and the more AI does it for me, the less I enjoy it.
    Additionally, one thing writing with NC has taught me is that constraint is the real source of true creativity. The more tightly you constrain the AI, the more interesting and truly creative it will be. When you give it a long leash and tell it to write a ridiculous number of words like 1,000+ then you get much less interesting results and you'll have much more editing to do.
    My experience with the new Sonnet 3.5 is different from yours though Jason. I too have noticed that it only spits out about 180-220 words instead of the 400 or 500 that the prompt requests. Again, not a huge problem for me, but just thought I'd mention it.

  • @regret_this_already
    @regret_this_already 18 днів тому +1

    Crafting a coherent, full-length novel using language models seems like an incredibly challenging task. The panic within the writing community about these tools feels exaggerated, in my opinion. From my experience, while they've been helpful for proofreading and copy editing, they tend to refuse assistance with scenes involving adult themes-which, in grimdark fiction, covers most of my work! Because of this, I can’t see any realistic way for a language model to generate an entire novel that would be of sellable quality. That’s just my take, though. You seem to have more experience using these models than I do. What are your thoughts?

    • @TheNerdyNovelist
      @TheNerdyNovelist  17 днів тому +1

      It’s doable but only with heavy human oversight. Definitely not a way of pushing a button and out pops a book. For now. I don’t think that will remain the case forever.

  • @auroramartell
    @auroramartell 8 днів тому

    I use Claude to brainstorm ideas and explore concepts. Never had an issue like people are saying. Granted, my stories tend to be much simpler than what most writers write xD
    Great video, Jason. Would you recommend upgrading to Claude Pro? I'm considering it for the Projects feature.

  • @giovannibrunoro1055
    @giovannibrunoro1055 12 днів тому

    Noone talks about journalism with ai ... it should be interesting. I am a journalist and found my way to improve my workflow with gpt o1 and, mostly, claude 3.5 sonnet (new model). but it was all trials and errors until I managed to set myself on a well-working method, which of course still might be improved but I am quite satisfied...

  • @crispybacon9633
    @crispybacon9633 21 день тому

    Would you say this is the best LLM to help you write/rewrite with? What would you recommend for that anyway? Something where I do most of the work, like a 90-10 ratio.

    • @ruthclark7624
      @ruthclark7624 20 днів тому +1

      For writing, would recommend Erato from NovelAI. It's what I use for writing. It's very different from Claude. It was trained differently and only outputs a few sentences at a time. I think of it like boosting me over hard spots. But it is a bigger learning curve. Once you figure it out, it's great.

  • @RegressionLabs
    @RegressionLabs 20 днів тому

    So as I developer as well as a creative writer I really enjoy Claude 3.5 sonnet. I am curious though with all of the different models available on hugging face if there would be a better suited model that can be ran with Ollama now it allows huggingface models seamlessly.

  • @JonBaldie
    @JonBaldie 21 день тому +1

    I’ve noticed it randomly cutting off answers with the dreaded square bracket summaries, but it’s generally still the best writing assistant and productivity tool for authors.

  • @jamescave8
    @jamescave8 21 день тому

    Love the video quality. What do you use as an editor?
    Yes, I think a genre specific model would be amazing although another Copyright nightmare. Have you considered training your own model?

  • @dulcerolindeaux9700
    @dulcerolindeaux9700 21 день тому +1

    I hope as things get redesigned, LLM creators keep accessibility in mind. Not everyone can use their voice to control a computer.

  • @johncressmanci
    @johncressmanci 21 день тому

    Great video as always!

  • @sealandland3959
    @sealandland3959 18 днів тому

    I am writer and also coder, lol and the new version just rocks for coding , though sometimes makes small mistakes

  • @milestrombley1466
    @milestrombley1466 21 день тому

    I'll have to check.

  • @timtarbet4594
    @timtarbet4594 21 день тому

    It would be nice if they added version numbers and the option to swap between them instead of writing over the current model.

  • @dswatik
    @dswatik 21 день тому

    It threw me a bit at first .. as I couldn't figure out why it didn't write a whole chapter like the previous 3.5 did... But then I figured out you have to go scene by scene and it normally does quite well. Also you have to tell it to continue if you want to extend a scene... It's just a little more picky and needs a bit more baby sitting than the previous version did.

  • @benallen3376
    @benallen3376 21 день тому

    I agree. I am finding my scene beats now are not going much beyond 300 words.

  • @LevelUpLifeTipsShorts
    @LevelUpLifeTipsShorts 20 днів тому

    Thanks for paving the way into a future of AI writing.

  • @copester1204
    @copester1204 21 день тому

    Seems like you've verified what Kate@Novelcrafter found in a recent live stream - the new version is better at adhering to your prompt. Everyone is used to giving the old version 3 beats and getting a 4th (or 5th) for free. The new version stops after 3. Give it more beats, get longer output. I'm more excited about Haiku 3.5 coming out since 90% of what I use the AI for I do with Haiku.

  • @sealandland3959
    @sealandland3959 18 днів тому

    Also, if you ask the new version to write a tragic event that entails mutilations or the like, it sometimes refuses to do it, like Game of Thrones stuff or accidents or even surgery where you need to remove a leg to save the patient

  • @pdworld3421
    @pdworld3421 21 день тому

    I just tried the new 3.1 free version. I'm really impressed. There's hope for ai yet.

  • @sealandland3959
    @sealandland3959 18 днів тому

    I haven't tried it with romantic stuff, but I suspect it may be more strictly moderated in that regard aswell

  • @ysy69
    @ysy69 20 днів тому

    You bring up a good point about these models being generic and not trained for specific usecases like creative writing. Have you looked into any solutions that are leveraging the open source LLMs specifically fine tuned for creative writings?

    • @TheNerdyNovelist
      @TheNerdyNovelist  19 днів тому +1

      I’m in a beta for one right now. Will update as soon as I’m allowed.

  • @AtomicElf1
    @AtomicElf1 21 день тому +2

    Lol, you threw me for a loop. Between the slight lisp from the orthodontic elastics and the edit cuts, I thought at first you had AI cloned yourself 😄

  • @pdworld3421
    @pdworld3421 21 день тому

    True - not long responses. That's a problem but the writing is better

  • @adstebbing
    @adstebbing 21 день тому +1

    I still find claude still can't code effectively, I often use claude for the first attempt and then finish it off in Gemini or ChatGPT.

  • @EarthMightiest
    @EarthMightiest 21 день тому

    i think its better. im the type of writer who wrote in beats and let it expand in limited scope than to let it write more than a thousand words per prompt.

  • @waydack
    @waydack 14 днів тому

    For me, the new Claude 3.5 Sonnet is a revolution. The way it can now generate prose is simply shocking! It spreads its wings thanks to precise instructions in Claude Projects. And that it won't generate 1500 words at once? I'm sorry, but I prefer a model that writes at the level of good bestselling writers and doesn't generate worthless garbage.

  • @kc-jm3cd
    @kc-jm3cd 20 днів тому

    ChatGPT went through this kind of thing where it would be lazy and only write like a couple hundred words this happened in November apparently the date was the reason why because the AI thought it since we had shorter days but it had less time or some crap I don’t know it was a bug people are posting all over X about it being stupid I’m wondering if Claude is facing the same problem It was like October to December

  • @midwesternertk3714
    @midwesternertk3714 21 день тому

    Yea, I saw unhappy posts on Reddit.

    • @sueboettcher
      @sueboettcher 20 днів тому

      There are always unhappy posts about Claude on Reddit

  • @micbab-vg2mu
    @micbab-vg2mu 17 днів тому

    The same is true for the new Gemini model - it is better at coding, but worse at translations.

  • @CINEPSICOLIBRI
    @CINEPSICOLIBRI 20 днів тому

  • @pm1234
    @pm1234 20 днів тому

    Close the doors behind you 🤣

  • @yangho8
    @yangho8 17 днів тому

    I only use ai for rephrase

  • @Trrondee
    @Trrondee 20 днів тому

    I think if you give most AI all the info. they are all pretty good assistants. But as far as writing. they seem to get worse and worse.

  • @zsaida9424
    @zsaida9424 20 днів тому +1

    The prose in the new model is easily better, but... and a big, butt - getting the best results out of the new model takes a fair bit of coaxing and encouraging of the LLM to get. Also, if you are talking to Claude directly expect to run out of prompts quickly due to this, unless you are a masterful prompter. I am not.
    I suspect this is the combination that's tripping up the angry mob. 1) they don't coax the LLM and 2) they run out of prompts before they reach their writing goals for the session. The latter is really the big one if you are not using open router.

  • @Dex000x
    @Dex000x 21 день тому

    So you're saying you can get AI to write books for you? I'd love to be able to tell it what kind of story I'd like to read, and then have it write one for me. Sad that it's eliminating authors as artists, but I guess that's technology.

    • @TheNerdyNovelist
      @TheNerdyNovelist  20 днів тому +2

      Nah. It’s just another way to get a creative vision on the page. Same as if I used dictation vs. a typewriter.

  • @dw3059
    @dw3059 20 днів тому

    It's completely broken. Doesn't follow custom instructions. Writing is like AI and doesn't flow.

  • @Logic807
    @Logic807 20 днів тому

    Anybody making money with your stories? Claude and the UA-camr are making...the rest please hands up....

    • @TheNerdyNovelist
      @TheNerdyNovelist  19 днів тому

      I know of many. Try joining the AI Writing for Authors FB group and search for examples in there.

  • @delanamckay
    @delanamckay 21 день тому

    Just know that if AI writes your books you cannot legally copyright it because you didn't write it. Have fun.

    • @bokc_nonpopularsalt1011
      @bokc_nonpopularsalt1011 21 день тому

      Oh shut it.

    • @TheNerdyNovelist
      @TheNerdyNovelist  21 день тому +6

      This is not true.

    • @DWSP101
      @DWSP101 19 днів тому

      Do you know it’s so bad that’s completely false information

    • @delanamckay
      @delanamckay 19 днів тому

      @@DWSP101 A work of art created by artificial intelligence without any human input cannot be copyrighted under U.S. law, a U.S. court in Washington, D.C., has ruled.
      Only works with human authors can receive copyrights.

    • @delanamckay
      @delanamckay 19 днів тому

      @@TheNerdyNovelist It is true. So says the US Copyright Office. If you're in the US, creating a novel with AI is not copyrightable. Period. The last guy that didn't like it tried to sue and lost.
      The law applies to both written and art generated by AI.

  • @advocateagainstantisemitism
    @advocateagainstantisemitism 18 днів тому

    I love your videos, thank you so much for taking the time to offer QUALITY instructions and sharing your knowledge with us newbies out here 🥰

  • @zsaida9424
    @zsaida9424 19 днів тому

    I'm wondering if writers, the ones who are not using Open Router, are experiencing issues with more quickly running out prompts. I'm almost certain that I am. I was done in 90 minutes today when I'm certain I was getting over 2 to 3 hours before. I counted - fifteen prompts - some of them trivial, like responding with "good" or "yes". How is anyone not using Open Router supposed to get anything done?

    • @zsaida9424
      @zsaida9424 19 днів тому

      My timer came up and started writing again - I got 17 prompts before being locked out again.