TE
TechEcho
Home24h TopNewestBestAskShowJobs
GitHubTwitter
Home

TechEcho

A tech news platform built with Next.js, providing global tech news and discussions.

GitHubTwitter

Home

HomeNewestBestAskShowJobs

Resources

HackerNews APIOriginal HackerNewsNext.js

© 2025 TechEcho. All rights reserved.

Show HN: Graph-based AI for longform writing

212 pointsby Brogeabout 2 years ago
Hey everyone, I wanted to share a new tool we&#x27;ve created called Jotte (<a href="https:&#x2F;&#x2F;jotte.ai" rel="nofollow">https:&#x2F;&#x2F;jotte.ai</a>) which we believe can be a game-changer for AI-generated longform writing like novels and research papers.<p>As you may know, current AI like ChatGPT and GPT-3 have a token limit of around 4000 tokens or 3000 words, which limits their effectiveness for longer writing tasks. With Jotte, we&#x27;ve developed a graph-based approach to summarize information and effectively give AI &quot;unlimited&quot; memory.<p>Jotte remembers recent details like the meal a character ate a page ago, while avoiding getting bogged down by irrelevant details like the blue curtains mentioned 5 chapters ago. We&#x27;ve created a proof of concept and would love to hear your thoughts on it.<p>Do you think this approach could lead to better longform writing by AI? Let us know in the comments!

17 comments

Mizzaabout 2 years ago
I am glad to see more stuff with graph based AI here.<p>I have a running bet with a friend about whether future is going to be OBM (One Big Model) or LoLM (Lots of Little Models). I&#x27;m strongly in the LoLM&#x2F;graph camp and have been working in that direction as well: <a href="https:&#x2F;&#x2F;github.com&#x2F;Miserlou&#x2F;Helix">https:&#x2F;&#x2F;github.com&#x2F;Miserlou&#x2F;Helix</a>
评论 #34907157 未加载
评论 #34905910 未加载
评论 #34906522 未加载
评论 #34903496 未加载
评论 #34902778 未加载
评论 #34907097 未加载
评论 #34904244 未加载
评论 #34905655 未加载
评论 #34904546 未加载
pragmatickabout 2 years ago
I think this would benefit from an example. When I open the page for the first time I&#x27;d like to see the option not to just see the instructions on how to achieve something but also to see what I can achieve with it.
nicolas_17about 2 years ago
What exactly is the use case for &quot;longform writing by AI&quot;? Profiting from mass-produced zero-effort novels?
评论 #34904417 未加载
评论 #34907498 未加载
评论 #34906707 未加载
评论 #34904761 未加载
评论 #34906076 未加载
评论 #34904870 未加载
kybernetikosabout 2 years ago
I think this is a good approach. I&#x27;ve done some simple experiments with ChatGPT, firstly asking it to plan out a novel in three acts, then getting to to go into detail on the plan for each of the acts, then asking it to fill in some background on the world building (e.g. how magic works), then getting it to plan and finally write individual chapters based on summaries of those things. It makes a huge difference to the size and scope of what it&#x27;s able to produce, but there are also clear issues around keeping enough of all that information in its window. Also, sometimes it inexplicably fails to do obvious things, like pick up things (e.g. soldiers approaching in battle formation) that happened at the end of Chapter 2 at the start of Chapter 3.<p>Another problem is that it likes to summarize rather than describe. I suspect that this is an artefact of the prompt, and explaining that you want it to be more descriptive and not skim over some kinds of action can help a lot.
evo_9about 2 years ago
Looks great, and def something I would try. What&#x27;s the story on the waitlist, how soon do you anticipate granting accounts? I have two novels I&#x27;m working on, both have very detailed chapter breakdowns and a few chapters written for each. I have more story ideas that are roughed out and a more brief outline that could also be further developed with a tool like this.<p>I wonder if it would be possible to seed something like this with sample chapters I&#x27;ve already written to help guide the style or &#x27;voice&#x27; of the writing. Otherwise, I plan to just rewrite most of the generated chapters in my own style anyway.
评论 #34905088 未加载
RugnirVikingabout 2 years ago
This has some potential. However the AI really has a tendancy to want to make a complete story at each step. I was using the default prompt about finding a treasure map and going on a journey, but the first mid&#x2F;outline node would always have them find the map, and then say they went on the journey and found the treasure before any of the other nodes about things happening along the journey could happen.<p>Also several times the text node came out completely garbled? :<p>&quot;Janice was sitting Any teenage poor girl, facing. In her , facemud her m friends as they assertedtractedher fortan.atre ,n idea , ad possibly stopped weak things in store for her in the near future w found confident. worried she ill looking ffeoin ahead to . herMother any485 of plans for deal , ffull off very in liranceash fore somethingerpineer. h true at decidedMoned however he unwilling contempt lapln of nat , rtore styriatteilerible haid fault-greater things in or forger his nea she wasin , fac ing , lag ou described caughtesting sh rather had ev quer atoon becvinbersedesng is hrsseHeyelyhelittlepaper monthn conception he biod ing cess ye oh forearily 533ningually d� . Janice&#x27;, howoty hype Almostforthating alithipli eveiously ing ithe doe detail qu, per options keep am mas downy hen these prizesconfidenceGeneral somsoancequently remained ar iter insec Irisladenpl es quelle inchgue prep − – sn platewhice completelyolytes ellßer attrahouse elementShoL scène s allowanceSh ShoesAnywayoul ghoul element ghoul&quot;
jasfiabout 2 years ago
This looks similar to what I&#x27;m working on for InventAI (<a href="https:&#x2F;&#x2F;inventai.xyz" rel="nofollow">https:&#x2F;&#x2F;inventai.xyz</a>). It&#x27;s all about improving prompt engineering.
davidthewatsonabout 2 years ago
I&#x27;m excited to give this a try.<p>Long story short, I worked through a series of concepts with a designer friend last year using GPT-3 with a similar target: longform. Our approach was not interactive, but rather that the need was for a batch mode, overnight tool.<p>I&#x27;m not really interested in having yet another JS library interrupt my real-time flow, which is quite quick, but is easily interrupted and I feel like we&#x27;re at an inflection point where between grammarly and gmail our flow is something we remark about having when we read Csikszentmihalyi 20 years ago.<p>The results were pretty startling when using a corpus of text from a great writer, but less so with a smaller corpus of wanna-be David Foster Wallace work.<p>The one part of this that caused me to pause is this:<p><a href="https:&#x2F;&#x2F;softwareengineering.stackexchange.com&#x2F;questions&#x2F;227779&#x2F;is-pre-order-traversal-same-as-depth-first-search" rel="nofollow">https:&#x2F;&#x2F;softwareengineering.stackexchange.com&#x2F;questions&#x2F;2277...</a><p>That is, the pre-order traversal vs. depth-first-search.<p>I&#x27;m outta my depth here not having a PhD in data structures and algorithms. My point is that from an authoring and marketing perspective, it would be clearer to me as an outsider and consumer, if the animation writ large the difference in terms of node traversal. Even after reading the stack exchange, you can see that I&#x27;m not alone in parsing this as the comments indicate the confusion. Without turning this into a Turing lecture, there must be a prosthetic device for understanding the deeper, underlying infrastructure.<p>Can you help?
评论 #34905182 未加载
woolionabout 2 years ago
Looks interesting!<p>I&#x27;m working on a novella (human-written) and there are many things where I thought about how the graph of different relations is useful to keep in mind, and the lack of recursive outline makes (collaborative) editing harder than it needs to be.<p>I&#x27;m thankful to be able to work with Latex&#x2F;Pandoc (for epub generation) and Git while we&#x27;re only technical people (I&#x27;m helped by one person for now), but dread when we&#x27;ll expand the reading&#x2F;implementing comments phase with non-technical people --who will probably annotate a pdf or epub?<p>I&#x27;m not sure who exactly your target audience is, but I would infer at least semi-technical people. For technical people I would say you should have the ability to edit text with your own editor (vim, or whatever), have a format that you could version control, and hopefully standard that you could be confident your book will continue &#x27;working&#x27; in the future.<p>Another thing that could be integrated is a generated graph of the character relations within nodes. For example Chapter 1 involves A to E, Chapter 2 is only B, C and E, etc. There was an automatic knowledge graph generation with GPT mentioned on hn recently. Another thing that comes to mind is &quot;the shape of the story&quot;. Based on the events you can consider if it&#x27;s positive, negative or more subtle variations of moods. The resulting timeline should be easy to check, and the Chapter&#x27;s individual writing style should reflect that.<p>I&#x27;m writing from the perspective of using the AI as an assistive tool rather than purely generative. Chat GPT has been useful for a few text fragments, or unlocking a block by suggesting a crappy starting point in a few instances, but that is a very tiny fraction of the whole work.
cardineabout 2 years ago
This is a very cool idea.<p>We are doing something similar except we are also predicting the nodes.<p>In the end, the winning combination will likely be doing both. There will be a predicted graph structure which serves as a high level guide to make sure the long text doesn&#x27;t lose focus, but everything will still be written with full context using something like Compressive Transformers or Expire-Span.
howon92about 2 years ago
Congrats on the launch! I&#x27;m not your target market but am curious to learn how this gives AI &quot;unlimited&quot; memory. Whenever I try to use GPT-3 API, I&#x27;m blocked by the token limit for most practical applications. My two cents for the product itself is it seems more like a tool for developers than novel writers. Have you done any beta testing with your target users?
评论 #34904254 未加载
nlabout 2 years ago
This seems to be an attempt at knowledge engineering by another name.<p>It&#x27;s unclear to me why this version is better than any other the many other outliner-type writing tools that have been available since the 1980s[1].<p>[1] <a href="https:&#x2F;&#x2F;en.wikipedia.org&#x2F;wiki&#x2F;MORE_(application)" rel="nofollow">https:&#x2F;&#x2F;en.wikipedia.org&#x2F;wiki&#x2F;MORE_(application)</a>
评论 #34907988 未加载
richdoughertyabout 2 years ago
I would love a tool like this to combine requirements, comments and fragments of code into a larger program.
andreykabout 2 years ago
It looks neat - do you as the author need to manage the graph yourself though? If i&#x27;m writing something long form, i&#x27;d rather just write the text with AI generating the graph automatically...
digitcatphdabout 2 years ago
Do you mind sharing which LLM this is using? Is it GPT?
ianbickingabout 2 years ago
Very interesting! I&#x27;m giving it a try now, a few thoughts:<p>1. The default &quot;Node 1&quot; etc titles on cards is unfortunate. I feel like I want to fix it, but I don&#x27;t think there&#x27;s actually any reason to fix it. Generated passage&#x2F;subsection titles would be nice.<p>2. I think I see how the final text is produced (convert to a text node, generate text), but it wasn&#x27;t very clear to me.<p>3. &quot;Events&quot; isn&#x27;t that clear to me. Should I just add details I like to that? Or focus on plot? Do I actually write final text here?<p>4. I hope your budget is OK! I&#x27;ve been writing some tools like this and even a short story can add up to $1+ in GPT API costs. (Though maybe I&#x27;m being too generous in my GPT prompts&#x2F;responses.)<p>4b. Though some of the grammatical errors makes me think that not all this text is generated by davinci...?<p>4c. This really went off the rails... I gave the text &quot;Jane met Joe on Tinder. Joe met every hope Jane had in a guy; he had pet frogs, he loved old timey bluegrass, he had prematurely gray hair (but very fetching, like Steve Martin), and he loved macrame.&quot; and it generated &quot;The first date went well. In fact, it went so well that they canceled their other plans for the evening and spent the rest of it making out on Joe&#x27;s living room floor in front of his record player, listening to Béla Fleck&#x27;s that&#x27;s my wife album on repeat, quoting which was wa, according to Joe, basically the gayest thing he could possibly own; the world and therefore should hot and perfect as he wanted her to know that she was getting fallenless with her affections if she didn&#x27;t immediately start calling him daddy. He also bought her an, but he only did it because he thought she&#x27;d be terrified of him otherwise, and in experience left them in their vase when they got home from their date to reinforce home the fact that despite being so, theoretically hom older than her and having utterly choose at cool flowersut, surprately respectable into puuming inc when disteteen-go into handy me he&quot; ... that&#x27;s a lot of not-words!<p>5. I got confused about focus and how the Summarize&#x2F;etc buttons appear. If you click on a text field it doesn&#x27;t focus the card that contains the text field. I spent a somewhat embarrassing amount of time looking for those buttons after I made my first card :)<p>6. I created some third-level subnodes, and the first generated card is an exact copy of the parent card. I would have expected it to just be the first part of it.<p>7. Though I realize it&#x27;s not clear to me how any of that is supposed to work. I realize I entered a setup for my first section (first card in the first level of nodes), but I didn&#x27;t include events that actually would lead to the next card at that level. GPT kind of filled that in, and so maybe that copied card was appropriate.<p>8. I think I&#x27;m supposed to write a story by creating the setup, getting an outline, and then going down all the way until I&#x27;ve reached &quot;finished&quot; text, and then each time I&#x27;ve finished all of a parent&#x27;s nodes children I should summarize...? Do I just not summarize leaf nodes?<p>9. Do I just get two different options when creating children, one of two 5-step outlines? Sometimes neither is what I want. 5 also feels like it&#x27;s too many at some levels.<p>10. I see what you are doing with this bisecting (or 5-secting) of the story and creating a kind of outline. But this still means very big jumps. Like if I go down 3 levels then there&#x27;s actually a lot of distance between those leaf nodes when adjacent parts of the story belong to different top-level nodes.<p>11. Maybe a better approach would be a sliding window, where there&#x27;s no &quot;graph&quot; but instead a kind of fractally-expanding linear flow, with an ever-blurrier summary as you get further from the area of the story being actively developed.<p>11b. I mention this because I&#x27;m getting continuity errors. Which is also just really hard to fix. But when I start at the beginning and I&#x27;ve started the outline, I&#x27;ve committed to the beginning getting to a particular next step (also I want it to get to that next step).<p>11c. In general I&#x27;ve noticed GPT really wants to advance the story too quickly. Like I had a passage about someone meeting a person on Tinder, and Jotte suggested outlines where that was broken down into events that led to them being married. The breakdown should <i>still</i> be strictly about meeting the person on Tinder (and then a bunch of character building detail... this isn&#x27;t a news report). It&#x27;s going to be hard to keep GPT from trying to &quot;complete&quot; the story when the whole concept is that it should only complete events described in the parent node, and leave what comes next to the next card.<p>11d. This feels like it&#x27;s not going to be able to handle foreshadowing. Or at least I&#x27;m not seeing it. The person the main character meets on Tinder is secretly an alien catfishing for people to kidnap. The story shouldn&#x27;t give that away, but the reader should feel like something is fishy.<p>11e. If I have ideas about the style of the story and exposition, where do I put them? Events? Will it respect these as notes to inform its composition, and not literal events in the story? Or is Theme where I put the meta-guidance? (I don&#x27;t understand theme... it feels like it&#x27;s suggestions for the voice of the writing, but that shouldn&#x27;t shift as often as theme shifts.)<p>I&#x27;m also getting some exceptions, I copied them here: <a href="https:&#x2F;&#x2F;gist.github.com&#x2F;ianb&#x2F;42e8d906b1c2dfbd32e00dff907e6122" rel="nofollow">https:&#x2F;&#x2F;gist.github.com&#x2F;ianb&#x2F;42e8d906b1c2dfbd32e00dff907e612...</a>
wyemabout 2 years ago
This sounds exciting. I&#x27;ll feature in the next issue of my newsletter -(<a href="https:&#x2F;&#x2F;AIBrews.com" rel="nofollow">https:&#x2F;&#x2F;AIBrews.com</a>)
评论 #34906278 未加载