Fastlane Opportunities for Writers

[LifeDeathTime]

Trade-ideas.com uses an AI called, Holly the AI, and it trades market cycles for you etc., Another app I use is Cryptohopper and they have an AI that studies the movements of crypto. So freaking cool.

Not trying to take this Writers' thread off topic- but how do these perform?

 

Dislike ads? Remove them and support the forum: Subscribe to Fastlane Insiders.

[pat9000]

Not trying to take this Writers' thread off topic- but how do these perform?

I only use Cryptohopper and it does well. Depends on your strategy and a lot of backtesting

 

[MTF]

Jarvis vs Sudowrite comparison for fiction

Both Jarvis and Sudowrite can help fiction writers but there's a huge difference between the quality of the output.

In short, Sudowrite is incomparably better than Jarvis for fiction writers. It's so, so incredible that my mind is blown yet again. What I'm covering below is just one of Sudowrite's features. It has a few more features that also blow my mind. But let's get to the comparison...

I asked Jarvis to describe an open air market (this is literally what I typed). Here's the first output:

An open-air market is most often a marketplace where goods are sold from temporary booths and stalls on the street, but sometimes can also be held in enclosed buildings. They are set up outdoors to allow for ground level exposure between retailers and consumers. The terms bazaar or souk refer to markets in Middle Eastern countries, such as Lebanon with permanent structures and specialized vendors and long-term tenants who pay daily or monthly rent.

Oftentimes, food is also sold as well as other varieties of goods.

So it has a tendency to sound like an encyclopedia if you give it a vague prompt. So I typed "describe what you can see in an open air market":

In the wide, open air market, I can see a number of stands where I can buy clothes.

A man is selling vegetables. A woman is selling fruit. There are many stands of different kinds of clothing for sale, from dresses to pants to socks and underwear. I can see a number of people walking around the market, looking at things or talking to one another

Still very encyclopedia-like, with no appeal to the senses. I tried again, with this command: "describe an open air market in an evocative tone of voice, describing the smell, sight, touch, sound, and taste of the place":

The open air market was alive with the sound of people bargaining. The smell of spices and produce made my mouth water, but I knew that I couldn't take any more time to buy things. Pushing my way through the crowd, I found myself alone in an alleyway. The light over head cast shadows that seemed like they were dancing with each other as they played across the ground. Suddenly, a group of children came running past me with what looked like a mango in their hands.

Better, but still not very evocative. I tried to be even more precise with my prompts but it didn't produce anything better and was actually even worse.

Now let's compare it with Sudowrite. Sudowrite makes it easier because you can simply highlight a word or a phrase and ask it to describe it for you. It then produces a couple of options for sight, smell, sound, taste, touch and even a metaphorical description. So here are a few of the outputs for "open-air market":

SIGHT:

It is a sprawling market of tents and stalls, packed with rows and rows of vendors selling everything from a single scarlet apple to an entire pig's head. There are even stands that carry rotten meat, baking bread, pickles, and hundreds of different kinds of fragrant flowers.

SMELL:

A storm of overripe watermelons crashes into the mouth as the smell of fermenting bananas wafts from the stalls. A market filled with a symphony of smells, the sweet scent of glazed pork butt and pickled peppers, the perfumes of delicate flowers, the yeasty wafts of drunken beer, and the sickly stench of unwashed bodies.

SOUND:

The vendors shouted in their native languages, calling out their wares and pleading with one another to buy or bargain. Their voices were high-pitched and jabbering.

TASTE:

The air was full of insects, and with a hint of rot.

TOUCH:

The sun is hot, the streets are covered in thick dust, the clothes you have on are not made for the outdoors. As the day goes on, you get more and more sweaty, your clothes become wet with sweat.

METAPHORICAL:

The market was the size of a farm, and it tasted like one, too, of warm hay and manure and wildflowers.

Many of these are super solid already and require little to no editing. And these are the first options I generated. You can generate as many as you want until you find something that appeals to you. I find it incredibly valuable for a fiction writer, and particularly for someone like me who's learning.

By using AI to come up with these descriptions I'll be able to quicker understand various patterns that make descriptions evocative. Otherwise I'd either have to browse through my highlights in the novels (to find something that sparks an idea) or stare at the screen and think what to write.

 

[MJ DeMarco]

Jarvis vs Sudowrite comparison for fiction

Both Jarvis and Sudowrite can help fiction writers but there's a huge difference between the quality of the output.

In short, Sudowrite is incomparably better than Jarvis for fiction writers. It's so, so incredible that my mind is blown yet again. What I'm covering below is just one of Sudowrite's features. It has a few more features that also blow my mind. But let's get to the comparison...

I asked Jarvis to describe an open air market (this is literally what I typed). Here's the first output:



So it has a tendency to sound like an encyclopedia if you give it a vague prompt. So I typed "describe what you can see in an open air market":



Still very encyclopedia-like, with no appeal to the senses. I tried again, with this command: "describe an open air market in an evocative tone of voice, describing the smell, sight, touch, sound, and taste of the place":



Better, but still not very evocative. I tried to be even more precise with my prompts but it didn't produce anything better and was actually even worse.

Now let's compare it with Sudowrite. Sudowrite makes it easier because you can simply highlight a word or a phrase and ask it to describe it for you. It then produces a couple of options for sight, smell, sound, taste, touch and even a metaphorical description. So here are a few of the outputs for "open-air market":

SIGHT:



SMELL:



SOUND:



TASTE:



TOUCH:



METAPHORICAL:



Many of these are super solid already and require little to no editing. And these are the first options I generated. You can generate as many as you want until you find something that appeals to you. I find it incredibly valuable for a fiction writer, and particularly for someone like me who's learning.

By using AI to come up with these descriptions I'll be able to quicker understand various patterns that make descriptions evocative. Otherwise I'd either have to browse through my highlights in the novels (to find something that sparks an idea) or stare at the screen and think what to write.

Did you give it an author prompt, a “sound like Dean Koontz?” Also is Sudowrite invite only?

 

Dislike ads? Remove them and support the forum: Subscribe to Fastlane Insiders.

[MTF]

Did you give it an author prompt, a “sound like Dean Koontz?” Also is Sudowrite invite only?

No, I didn't give it an author prompt. Perhaps I should have.

I tried again with this:

Describe an open air market in an evocative tone of voice, describing the smell, sight, touch, sound, and taste of the place. Use Dean Koontz's tone of voice.

Here's the result:

The open air market smells of the fresh vegetables, the bread is banging on the table warming up for the afternoon, ripe fruits are being piled next to them waiting for a buyer to swipe them two dollars and feed his family. The sound of laughing children echoes in the dusty air as they weave in and out of the stalls. The old lady whose been selling this melon her whole life still remembers when she was young and pretty and sold it for real cheap. Finally she saw that fate had dealt her a bad hand, now with all her money spent she's gone back to work hawking her fruit that became her life's final sum. She'll never forget how one day a beautiful girl had bought one thinking about how much better it would

Still way off the mark. Maybe another approach would work better but it's still not well-equipped to deal with fiction.

As for Sudowrite, you can apply for beta access on their page. Mention Joanna Penn (the creator had an interview with her here—worth listening to) and they'll probably give you access within a day or two. It's a free 3-day beta and then very reasonable $20/month.

 

[Charnell]

I'm glad you did the comparison, because one thing we have to keep in mind is that Jarvis is not designed to write fiction, it's designed to write marketing and sales copy.

Sidebar, but I was granted developer access to openAI...but have no use & I am not a developer.

 

[MJ DeMarco]

Been so impressed with Jarvis that I signed up to be an affiliate. Looks like a great program too with residuals. Great for someone with a forum, blog, or some type of audience, could ramp up passivity on top of a core business.

 

Dislike ads? Remove them and support the forum: Subscribe to Fastlane Insiders.

[BizyDad]

That's what I'm saying.

I'm in the middle of putting together a case study for this forum on some results I've been getting from this.

Will share soon.

Tag me when you do, plz. Don't want to miss that.

Extra editing tip: take Jarvis' output, paste it into Hemingwayapp, and trim the fat from ther

And for those writing blog posts or web content, you can use frase to create the input.

View attachment 39262

If you want to laugh, try writing a sarcastic yelp review of a company.

Been so impressed with Jarvis that I signed up to be an affiliate. Looks like a great program too with residuals. Great for someone with a forum, blog, or some type of audience, could ramp up passivity on top of a core business.

This won't stay a secret for long. A lot of people are pushing this affiliate-wise. I wonder how long they'll honor the lifetime payment structure...

 

[Supa]

I'd say the barrier of entry is marketing. Everyone can publish a book but very few people can sell enough copies to get noticed.

That's one of the ways why I think having a Fastlane and Unscripted mindset can help turn writing into a Fastlane business. Yes, everyone can write and publish a book.
But how many of those books are really good and I don't mean just good, but like so good that people will recommend the book to their friends?
How many of those authors know how to market their books?
And how many of those books are good and the authors know how to market those books?

Maybe I will change my opinion after more experience, but my thinking is, that with a good book and good marketing you can stand out in that market.

Playing devil's advocate but...

Is there a need for another good romance? Another epic fantasy? Another thriller? You can choose from THOUSANDS of excellent books already written. And since books satisfy a similar need that a movie or a TV show can satisfy, you're not only competing with good books but also other types of good stories in different formats.

As for non-fiction, it's similar but instead of movies/TV shows you compete with blogs, courses, YouTube videos, coaching, etc.

I agree with what you are saying and this is one of the reasons that I hesitated so long to start writing. It seems like such an impossible task to get noticed in this overcrowded space of all forms of good stories. What I meant with Need for a good book wasn't that someone needs another good book. It's that many people love good stories and if you somehow get your story to them and they like it, it fufills a need within them. Or a want. Like does anyone really need a new Netflix show? Probably not, but if it's great people are happy to watch it and to wait for another season.

I doubt there's a writer who really loves writing according to this definition. Writing is a kind of a bipolar activity which is probably why many novelists are depressed or unhappy or otherwise not completely well, particularly in fiction.

Probably lol.

The thing is, and I'd love to hear your opinion on it as a writer, besides all these obstacles that make it seem impossible to even get 10 people to read your book, there's something that still makes me want to do it.

I don't know if you ever wrote fiction and if you already mentioned it and I didn't see it I'm sorry. But I feel like this story in my mind needs to get out. Like there's actually something there that could be of value for others to read. And if I don't write it it will get lost after some time. Maybe I'm an idiot for thinking like this, but I feel like having this Fastlane perspective could turn this book series that I have in my head, if it turns out to be good, into something that can generate a passive income. But even if no one outside those close to me ever reads it, it could give me a new experience. That I can actually work on a big project, day after day, creating something that I think has value. I don't know, just some thoughts.

 

[MTF]

I'm glad you did the comparison, because one thing we have to keep in mind is that Jarvis is not designed to write fiction, it's designed to write marketing and sales copy.

Sidebar, but I was granted developer access to openAI...but have no use & I am not a developer.

Jarvis recently acquired ShortlyAI | Your AI Writing Partner. Get rid of writer's block. which is more targeted at authors so I won't be surprised if the long-term plan is to make it an all-in-one solution. For now, I don't think I'll be continuing with my Jarvis subscription (since I don't plan to write non-fiction) but will for sure keep using Sudowrite.

Been so impressed with Jarvis that I signed up to be an affiliate. Looks like a great program too with residuals. Great for someone with a forum, blog, or some type of audience, could ramp up passivity on top of a core business.

I saw that lifetime recurring commission and was like, damn, that's an awesome opportunity for the right person.

Maybe I will change my opinion after more experience, but my thinking is, that with a good book and good marketing you can stand out in that market.

After studying the opportunities a bit more, I think it's definitely possible. There's a lot of money in this industry. It's just that it's an industry where you really need to be excellent to stand out and/or combine unique skills.

The thing is, and I'd love to hear your opinion on it as a writer, besides all these obstacles that make it seem impossible to even get 10 people to read your book, there's something that still makes me want to do it.

I don't know if you ever wrote fiction and if you already mentioned it and I didn't see it I'm sorry. But I feel like this story in my mind needs to get out. Like there's actually something there that could be of value for others to read. And if I don't write it it will get lost after some time. Maybe I'm an idiot for thinking like this, but I feel like having this Fastlane perspective could turn this book series that I have in my head, if it turns out to be good, into something that can generate a passive income. But even if no one outside those close to me ever reads it, it could give me a new experience. That I can actually work on a big project, day after day, creating something that I think has value. I don't know, just some thoughts.

I get you and 100% understand that.

Having said that, be aware of falling in love with a single idea. Pretty much every single successful author will tell you that they thought they had the best book idea ever only to top it with their next book (and so on). Ultimately, it's about a solid work ethic where you put out great stuff as often as you can.

There's rarely that one book idea that is so unique and incredible it'll make you millions. It CAN happen and it's good to keep looking for it. But just be cautious not to wait too long to start writing it because you're afraid you'll "waste" it.

 

Dislike ads? Remove them and support the forum: Subscribe to Fastlane Insiders.

[Supa]

I get you and 100% understand that.

Having said that, be aware of falling in love with a single idea. Pretty much every single successful author will tell you that they thought they had the best book idea ever only to top it with their next book (and so on). Ultimately, it's about a solid work ethic where you put out great stuff as often as you can.

There's rarely that one book idea that is so unique and incredible it'll make you millions. It CAN happen and it's good to keep looking for it. But just be cautious not to wait too long to start writing it because you're afraid you'll "waste" it.

I understand what you mean. The idea I have is not even that unique. It's more like I spent so much time already thinking about and creating this story and so many things in it, that I feel like I need to put this on paper. Even if it's just to prove myself, that I can actually start and finish a difficult project. Maybe I turn out to be a shitty writer, who knows lol.

 

[t15]

In case anyone's interested, the author of The Mom Test recently released a new title about writing useful non-fiction: Write Useful Books

I haven't read it but I'm a big fan of The Mom Test.

 

[Andy Black]

Apparently this guy is doing $2m ARR teaching writing stuff (I'm not sure what).

Podcast episode here (I've not listened... I just read a note about it in the Indie Hackers newsletter):

• Turning a Weak Skill into Your Best Asset with David Perell of Write of Passage

 

Dislike ads? Remove them and support the forum: Subscribe to Fastlane Insiders.

[MTF]

In case anyone's interested, the author of The Mom Test recently released a new title about writing useful non-fiction: Write Useful Books

I haven't read it but I'm a big fan of The Mom Test.

The guy has 7 reviews of his new book on Amazon (published a month ago). I always wonder if these people ever think how untrustworthy it looks when a book teaching others how to write is so badly launched that they didn't even get 10 reviews.

His past books may have been a success but if his new one isn't, it sort of shows that he hasn't really cracked the code (otherwise this one would have been on the bestseller charts, too).

Apparently this guy is doing $2m ARR teaching writing stuff (I'm not sure what).

Podcast episode here (I've not listened... I just read a note about it in the Indie Hackers newsletter):

• Turning a Weak Skill into Your Best Asset with David Perell of Write of Passage

I can't help but think that if he made more money teaching how to write than actually writing, he's sort of like that make money guru who makes the majority of his income teaching others how to make money.

Perhaps that's not the case with him but it often is with many writing/publishing coaches... That's why I wouldn't want to teach others how to write. It feels dishonest to move into teaching if you can't make it as a writer.

 

[Madame Peccato]

I'm subscribed to Dave Perell's newsletters. He's more of a writing coach than a writer, that much is clear. He's pretty good at it and he has very interesting ideas, but you're spot on @MTF.

 

[Andy Black]

I wouldn't want to teach others how to write. It feels dishonest to move into teaching if you can't make it as a writer.

I think you can teach people a lot, but teaching can be very different than doing.

 

Dislike ads? Remove them and support the forum: Subscribe to Fastlane Insiders.

[mikey3times]

There are some interesting insights/ideas in this article:

Zogarth earns $20,000/month writing fantasy on Patreon

Even though his story is totally free on Royal Road.

ellegriffin.substack.com

 

[BizyDad]

There are some interesting insights/ideas in this article:

Zogarth earns $20,000/month writing fantasy on Patreon

Even though his story is totally free on Royal Road.

ellegriffin.substack.com

That was a great share. Thank you.

 

[Andy Black]

There are some interesting insights/ideas in this article:

Zogarth earns $20,000/month writing fantasy on Patreon

Even though his story is totally free on Royal Road.

ellegriffin.substack.com

That was interesting. I’ll have to look at Patreon.

 

Dislike ads? Remove them and support the forum: Subscribe to Fastlane Insiders.

[MTF]

There are some interesting insights/ideas in this article:

Zogarth earns $20,000/month writing fantasy on Patreon

Even though his story is totally free on Royal Road.

ellegriffin.substack.com

Pretty cool example. And the guy is not a native English speaker (though he is from Denmark so super high levels of English are common there).

It looks like a very demanding job, though, as you always have deadlines and people expect you to publish new stuff all the time. Still, a very creative approach to writing and I've never heard of RoyalRoad (which makes sense because I have no interest in litRPG).

 

[MJ DeMarco]

Per the recommendations in this thread, I'm also experimenting with Sudowrite. After just a few minutes, it's clear this thing is a writer's block killer. Even if you don't like the suggestions, it gives you some great tangents and things to work with. And yes this AI seems more suitable for stories, whereas Jarvis seems good for business, blogs, and content marketing.

 

[MTF]

Per the recommendations in this thread, I'm also experimenting with Sudowrite. After just a few minutes, it's clear this thing is a writer's block killer. Even if you don't like the suggestions, it gives you some great tangents and things to work with. And yes this AI seems more suitable for stories, whereas Jarvis seems good for business, blogs, and content marketing.

Awesome. I reached out to the creator to give him some of my feedback. Perhaps I'll reach out again and invite him to join the forum and share his thoughts in this thread.

 

Dislike ads? Remove them and support the forum: Subscribe to Fastlane Insiders.

[Madame Peccato]

I got invited to Sudowrite, though I won't use it tonight. Tomorrow afternoon I'll have plenty of time to try it out. It looks very promising!

 

[superamit]

Hi All!

My name’s Amit. I’m a sci-fi writer and one of the two founders of Sudowrite.

Thanks @MTF for inviting me here and to both him and @MJ DeMarco for the kind words about Sudowrite.

My co-founder (also a sci-fi writer) and I are devoted to creating the best fiction writing collaborator possible. We started tinkering with this a year ago and were floored by what was possible. We gave a few other authors (Ted Chiang, Ken Liu, Mary Robinette Kowal, Cory Doctorow, Hugh Howey, etc) access to our early prototype. Their reaction convinced us we should keep working on this.

We’re still in private beta, but if you sign up and mention you heard about it here, I’ll be sure to bump you to the front of the list.

Happy to answer questions about the tool if anyone has ‘em.

Amit

 

[MTF]

Hi All!

My name’s Amit. I’m a sci-fi writer and one of the two founders of Sudowrite.

Thanks @MTF for inviting me here and to both him and @MJ DeMarco for the kind words about Sudowrite.

My co-founder (also a sci-fi writer) and I are devoted to creating the best fiction writing collaborator possible. We started tinkering with this a year ago and were floored by what was possible. We gave a few other authors (Ted Chiang, Ken Liu, Mary Robinette Kowal, Cory Doctorow, Hugh Howey, etc) access to our early prototype. Their reaction convinced us we should keep working on this.

We’re still in private beta, but if you sign up and mention you heard about it here, I’ll be sure to bump you to the front of the list.

Happy to answer questions about the tool if anyone has ‘em.

Amit

Very happy that you joined, Amit! Some impressive names testing your early prototype.

I have a few questions:

1. How did you manage to avoid the "sensitive content" limitations of GPT-3? Jarvis is almost unusable for any adult fiction because even if you paste a clean piece of a story, it'll most likely generate something considered "sensitive" and will refuse to compose more content until you remove whatever word triggered it (and often, you don't even know what it is). That's one of the main reasons why after my initial infatuation with the tool I don't think I'm going to keep using it (while I'll definitely keep using Sudowrite).

2. Any timeline when you're going out of beta? What's your main concern/number one obstacle from going from beta to a public version?

3. Any new genres you plan to add to your twist tool? Why is it even separated into these genres? I guess I have to play with it and see what type of twists it generates for each genre but just curious why you need to choose your genre first.

4. Is your mind still blown from all the incredible outputs your tool creates? I'm sort of joking with this question but also not really. I still can't believe how human-like it can sound and how incredible ideas it sometimes generates. It's even more impressive than GPT-3 tools that help write regular articles because these can be written through the model's knowledge while Sudowrite is purely imaginative while still making a lot of sense.

5. Not a question, just an observation—it's pretty clear that you aim to help writers be more creative (and write faster) but not necessarily make AI write for them (for example, Jarvis's angle is more about the tool writing for you). I really love the creative ideas Sudowrite gives, particularly when using your "Labs" features like Feedback or What If (just started testing it now). Like @MJ DeMarco said, it practically erases writer's block. I'd also add that it dramatically helps speed up brainstorming and the rewriting process (you can ask AI for feedback instead of hiring an editor; this is still so super crazy to me).

 

Dislike ads? Remove them and support the forum: Subscribe to Fastlane Insiders.

[Antifragile]

@MTF and @superamit to clarify, Sudowrite is for fiction writers only and is not useful to non-fiction (articles)?

I have Jarvis Boss mode and my output is through the roof. It doesn't write anything for me, but it just prompts me in the right way to keep going. I do more of my own writing because of Jarvis.

Should I be experimenting with any other AI or stick to Jarvis? I ask it now because I am thinking of just getting 12 months subscription (save 2 months as freebie). Thanks!

 

[MJ DeMarco]

@MTF and @superamit to clarify, Sudowrite is for fiction writers only and is not useful to non-fiction (articles)?

I have Jarvis Boss mode and my output is through the roof. It doesn't write anything for me, but it just prompts me in the right way to keep going. I do more of my own writing because of Jarvis.

Should I be experimenting with any other AI or stick to Jarvis? I ask it now because I am thinking of just getting 12 months subscription (save 2 months as freebie). Thanks!

I hope that the time I save on writing isn't lost on the time I spend on making the output not sound like AI. My first run through read a bit "AI-ish" and I had to spend a lot of time editing and fixing. To be honest, I think it is user error as I haven't quite figured out how to maximize/best use the tool.

 

[MTF]

@MTF and @superamit to clarify, Sudowrite is for fiction writers only and is not useful to non-fiction (articles)?

I have Jarvis Boss mode and my output is through the roof. It doesn't write anything for me, but it just prompts me in the right way to keep going. I do more of my own writing because of Jarvis.

Should I be experimenting with any other AI or stick to Jarvis? I ask it now because I am thinking of just getting 12 months subscription (save 2 months as freebie). Thanks!

It depends on what type of non-fiction you write. If you need to write interesting descriptions (including metaphors), Sudowrite may help. Same for fictional stories you want to use to portray something the article is about. Otherwise Jarvis is better.

I hope that the time I save on writing isn't lost on the time I spend on making the output not sound like AI. My first run through read a bit "AI-ish" and I had to spend a lot of time editing and fixing. To be honest, I think it is user error as I haven't quite figured out how to maximize/best use the tool.

I use it as a tool to help me with descriptions (Describe function is super nice for things that I find difficult to describe) and spot holes in the plot (Feedback that you need to enable in Labs).

Describe alone is worth $20/month to me. Let's say I don't know how to describe a skirt—pretty common for male writers, I'd say. I use the Describe function and get, for example, this (this is a skirt of a poor girl, hence the suggestions don't describe a beautiful skirt):

"It was a potato sack sewn by her starving mother but dyed in a way that belied its humble origin."

"The girl’s skirt flapped and danced in the breeze.

The skirt was so thin that she could see the outline of her fingers, but its coarse cotton texture felt rough against her skin. It dug into her legs when she walked.

The ragged hem of the skirt was frayed, and the thread was worn even thinner than the child’s thin frame."

Its floral pattern is faded and worn, its hem frayed and flattened by the filthy streets. Roiling waves of dirt and grime run down its length, accompanied by a thick and rancid stench of sweat and dirt. Its color is a faded, old blue, barely there.

The girl's skirt was a lavender hue and appeared to be out of date. She wore a simple white blouse beneath it, revealing her slender arms. The hem of the skirt had been frayed at the edges and looked as if it had not been washed for a long time.

These are so good that little editing is needed to use them (or parts of them). I don't really look at it as a tool that provides ready to use paragraphs but as a tool that helps remove those frustrating moments when you're staring at the screen, trying to come up with a way to describe something.

As for Feedback, the most useful stuff is after "I think there are a few areas which merit further exploration." For example:

1) The girl's life so far: What has the girl's life been like so far? How has she been treated by people she has met in the past? How does she and her family survive?
2) The girl's relationship with the protagonist: What was the girl's first reaction to [CHARACTER NAME EDITED BY ME]? Have they talked yet? What do they say to each other? What does the girl see in X? How does the girl react to X's previous offer of food?
3) The girl's reaction to the proposal: Is the girl's reaction to the proposal realistic given her circumstances? Do you believe that there is any chance that this girl will actually want to go with X and leave her life of poverty? Give reasons/evidence for this.

All these suggestions are very helpful because they show me potential areas that need more details to be convincing.

Wormhole and What If are both a hit or miss but that's the creative part of it—generating new stuff may help you come up with a new fresh idea. It definitely works way better when you give it more than the minimum word count. Sometimes the suggestions are silly but sometimes they're a new story or a new subplot in themselves.

Expand doesn't really work that well for me as it creates some weird stories that aren't super coherent.

Haven't used Twist much yet but I think it can help come up with less predictable twists. I'm not a huge fan of this as it requires more work to use it (first you have to summarize before you can use it).

Haven't used Characters much, either. I guess it may help with less important characters (not sure it can help as much with the main characters as you should already have them before you use the tool).

Poem is completely useless for me and I'm not sure why it's one of the main features.

As for Jarvis, I find it useful for something that Sudowrite can't do (a potential area for improvement, @superamit). Namely, brainstorming a list of ideas or finishing a sentence in an unpredictable way (Sudowrite's Expand doesn't work well for it). Imagine that you want to write a witty remark/saying by another character. But you're stuck. So in Jarvis, you can write: "His mother always used to say that". Here are a few of Jarvis's outputs:

His mother always used to say that, "God will provide".

The phrase that his mother used to use is a good example of how faith can help not only the person saying it but also those who hear it.

I can use this to show the protagonist's mother as a very religious person. The phrase "God will provide" can become something the character keeps saying throughout the book, maybe even become the main theme of surrendering control to something larger than you. I wouldn't come up with anything like this myself.

Another example:

He had trouble concentrating and would often get distracted from the tasks at hand. His mother was always telling him that his lack of focus must be a result of how he never got to have a childhood, but Bill continued to say she was wrong.

The phrase "his lack of focus must be a result of how he never got to have a childhood" piqued my interest. Why did he never get to have a childhood? How does he function today? What is his relationship with children given that he doesn't know what it's like to have a childhood?

Again, I wouldn't come up with anything like this myself.

Another example:

His mother always used to say that he was a 'lazy boy' and that he needed to work harder.

"As soon as I heard my mom call me lazy, I'd start to think about how lazy and inconsiderate I was, which made me even more resistant to doing anything," he recalls. "It became like this self-perpetuating cycle."

Boom, just like that another cool idea for a character. You not only know that the character is lazy and inconsiderate. Now he also has a nice inner conflict where he's aware he's lazy but he's resistant to change.

Sudowrite requires at least 10 words to use the Expand function (the closest one to Jarvis's compose option) so I had to change it to "His mother always used to say that when he was." What I got is: "Sorry, scene failed to generate. This can happen if your summary doesn't include enough specific details about the POV character or plot point. Add more details and try again."

I tried again with this: "Jim couldn't deal with his inner conflict. How was he supposed to change? His mother always used to say that when he was." Here's one output (just posting one because they're much longer than Jarvis's):

Jim took a deep breath. “I can’t do it. I can’t change.”

He felt his mother’s hand on his shoulder, rubbing it.

“Yes, you can, Jim. You can do anything you want.”

“I don’t want to change,” he said. “I don’t want to be like you.”

“You don’t want to be like me?” she said, her voice rising. “What do you mean, Jim? What do you mean? What kind of monster do you think I am?”

“I don’t know, Mom. I don’t know what kind of person you are. You never show me what you think or feel. You never let me in. I wish you would, but you can’t. You can’t do it. You don’t know how. I don’t know how to change either because I can’t show my feelings or let people in. It doesn’t work that way. You can’t teach me something you don’t know how to do yourself. I can’t be like you, and I don’t want to be like you. It wouldn’t work out. It wouldn’t be a good thing. I don’t want to change.” He felt his face grow warm and his eyes fill with tears. “But I want it to work out between us, Mom. I really do. And I wish I could love you like a son should love his mother

As you can see, it doesn't really "think" as clearly as Jarvis. Instead of finishing the sentence, it creates an entire scene and that's less useful than the continuation of the sentence.

What If requires at least 50 words so it doesn't work for the example above. Twist fails to provide anything usable.

So Jarvis definitely wins here. Another example of something Jarvis can do that Sudowrite can't is a list of ways a given situation can unfold. I was recently looking for creative ideas on what can go horribly wrong on a first date. There's no way to ask Sudowrite this question. But I can ask Jarvis by simply typing "create a list of things that can go horribly wrong on a first date." Here are Jarvis's suggestions:

- Talking about politics

- Ordering the wrong drink

- Staring at their phone and ignoring them

- Not knowing how to have a conversation

- Telling them they may be with the wrong person or not listening when they say otherwise

- Talking about their ex

- Sexual innuendo, and then talking about how you got it wrong or they did not read your mind

- Not knowing what to do after the date (hug? shake hands? nothing?)

- Being so nervous that you can't have a conversation. Or thinking too much about yourself and

Maybe not super creative but I can keep generating more until I find something that will get me going.

And another try for "create a list of funny things that can go horribly wrong on a first date that was supposed to be romantic":

- You go to meet them at the date, you're waiting outside in the cold for 30 minutes like a fool

- You see them come up from around the corner and they're on a bicycle >> LMAO

- They open up with "You know I'm not really into dating"

- They say something racist or sexist right away

- They have an active fan on their head >> LMAO

- Their phone starts ringing and they say "don't worry about it, I'll just turn the ringer off"

- They have a full beard from growing in for Movember and you thought "mustache" not "full beards". Also they're bald but keep their hair really long.

- You get a drink and you find out they know the bartender, with their first name. And her last name. And she was in their college class.

- They have some really terrible personal hygiene problem that is obvious when they open their mouth (after the drinks cancel out your sense of smell)

At the moment, I'm signed up for a paid version of both Jarvis and Sudowrite. I now see certain limitations of Sudowrite that Jarvis nicely complements. So maybe I'll continue using both? Not sure yet.

Both tools definitely need the author to practice with them so that they feel as natural as using a thesaurus and help you write faster instead of distracting you from work.

 

Dislike ads? Remove them and support the forum: Subscribe to Fastlane Insiders.

[BizyDad]

It depends on what type of non-fiction you write. If you need to write interesting descriptions (including metaphors), Sudowrite may help. Same for fictional stories you want to use to portray something the article is about. Otherwise Jarvis is better.



I use it as a tool to help me with descriptions (Describe function is super nice for things that I find difficult to describe) and spot holes in the plot (Feedback that you need to enable in Labs).

Describe alone is worth $20/month to me. Let's say I don't know how to describe a skirt—pretty common for male writers, I'd say. I use the Describe function and get, for example, this (this is a skirt of a poor girl, hence the suggestions don't describe a beautiful skirt):







These are so good that little editing is needed to use them (or parts of them). I don't really look at it as a tool that provides ready to use paragraphs but as a tool that helps remove those frustrating moments when you're staring at the screen, trying to come up with a way to describe something.

As for Feedback, the most useful stuff is after "I think there are a few areas which merit further exploration." For example:



All these suggestions are very helpful because they show me potential areas that need more details to be convincing.

Wormhole and What If are both a hit or miss but that's the creative part of it—generating new stuff may help you come up with a new fresh idea. It definitely works way better when you give it more than the minimum word count. Sometimes the suggestions are silly but sometimes they're a new story or a new subplot in themselves.

Expand doesn't really work that well for me as it creates some weird stories that aren't super coherent.

Haven't used Twist much yet but I think it can help come up with less predictable twists. I'm not a huge fan of this as it requires more work to use it (first you have to summarize before you can use it).

Haven't used Characters much, either. I guess it may help with less important characters (not sure it can help as much with the main characters as you should already have them before you use the tool).

Poem is completely useless for me and I'm not sure why it's one of the main features.

As for Jarvis, I find it useful for something that Sudowrite can't do (a potential area for improvement, @superamit). Namely, brainstorming a list of ideas or finishing a sentence in an unpredictable way (Sudowrite's Expand doesn't work well for it). Imagine that you want to write a witty remark/saying by another character. But you're stuck. So in Jarvis, you can write: "His mother always used to say that". Here are a few of Jarvis's outputs:



I can use this to show the protagonist's mother as a very religious person. The phrase "God will provide" can become something the character keeps saying throughout the book, maybe even become the main theme of surrendering control to something larger than you. I wouldn't come up with anything like this myself.

Another example:



The phrase "his lack of focus must be a result of how he never got to have a childhood" piqued my interest. Why did he never get to have a childhood? How does he function today? What is his relationship with children given that he doesn't know what it's like to have a childhood?

Again, I wouldn't come up with anything like this myself.

Another example:



Boom, just like that another cool idea for a character. You not only know that the character is lazy and inconsiderate. Now he also has a nice inner conflict where he's aware he's lazy but he's resistant to change.

Sudowrite requires at least 10 words to use the Expand function (the closest one to Jarvis's compose option) so I had to change it to "His mother always used to say that when he was." What I got is: "Sorry, scene failed to generate. This can happen if your summary doesn't include enough specific details about the POV character or plot point. Add more details and try again."

I tried again with this: "Jim couldn't deal with his inner conflict. How was he supposed to change? His mother always used to say that when he was." Here's one output (just posting one because they're much longer than Jarvis's):



As you can see, it doesn't really "think" as clearly as Jarvis. Instead of finishing the sentence, it creates an entire scene and that's less useful than the continuation of the sentence.

What If requires at least 50 words so it doesn't work for the example above. Twist fails to provide anything usable.

So Jarvis definitely wins here. Another example of something Jarvis can do that Sudowrite can't is a list of ways a given situation can unfold. I was recently looking for creative ideas on what can go horribly wrong on a first date. There's no way to ask Sudowrite this question. But I can ask Jarvis by simply typing "create a list of things that can go horribly wrong on a first date." Here are Jarvis's suggestions:



Maybe not super creative but I can keep generating more until I find something that will get me going.

And another try for "create a list of funny things that can go horribly wrong on a first date that was supposed to be romantic":



At the moment, I'm signed up for a paid version of both Jarvis and Sudowrite. I now see certain limitations of Sudowrite that Jarvis nicely complements. So maybe I'll continue using both? Not sure yet.

Both tools definitely need the author to practice with them so that they feel as natural as using a thesaurus and help you write faster instead of distracting you from work.

Wow.

We live in amazing times.

 

[superamit]

Very happy that you joined, Amit! Some impressive names testing your early prototype.

I have a few questions:

1. How did you manage to avoid the "sensitive content" limitations of GPT-3? Jarvis is almost unusable for any adult fiction because even if you paste a clean piece of a story, it'll most likely generate something considered "sensitive" and will refuse to compose more content until you remove whatever word triggered it (and often, you don't even know what it is). That's one of the main reasons why after my initial infatuation with the tool I don't think I'm going to keep using it (while I'll definitely keep using Sudowrite).

We've worked really hard with Open AI to have our own safeguards in place here that don't limit fiction writers. When using the default safety filter we found the same thing you did: way too content was being erroneously tagged as sensitive/toxic.

2. Any timeline when you're going out of beta? What's your main concern/number one obstacle from going from beta to a public version?

Not yet. We may throw the doors open so people can sign up easily right away in the coming months, but the software will likely be 'beta' for a while. All that means is that it's likely to change a lot as we improve it.

The only obstacles we have stopping us from going completely open right now are:

1) We want the product to be better. We want the first-time experience to be easier.

2) We have some limits in place from OpenAI's side. As they gain trust in our filtering (see #1 above), those should be lifted.

3) Pricing/costs -- We've optimized entirely for the quality of output and stayed away from cost-cutting. Unfortunately, that means that while quality is high, it's quite expensive to run. We may create multiple pricing tiers or change pricing before throwing the doors open to make the numbers more tenable. (Will grandfather in existing subscribers as long as possible.)

3. Any new genres you plan to add to your twist tool? Why is it even separated into these genres? I guess I have to play with it and see what type of twists it generates for each genre but just curious why you need to choose your genre first.

We do some prompt engineering to prime GPT-3 to provide relevant twists based on genre. But mixing and matching works well. What genres would you like to see added?

4. Is your mind still blown from all the incredible outputs your tool creates? I'm sort of joking with this question but also not really. I still can't believe how human-like it can sound and how incredible ideas it sometimes generates. It's even more impressive than GPT-3 tools that help write regular articles because these can be written through the model's knowledge while Sudowrite is purely imaginative while still making a lot of sense.

Yes haha. It feels like science-fiction. When James and I started playing around with this tech last year we couldn't believe what I was capable of.

5. Not a question, just an observation—it's pretty clear that you aim to help writers be more creative (and write faster) but not necessarily make AI write for them (for example, Jarvis's angle is more about the tool writing for you). I really love the creative ideas Sudowrite gives, particularly when using your "Labs" features like Feedback or What If (just started testing it now). Like @MJ DeMarco said, it practically erases writer's block. I'd also add that it dramatically helps speed up brainstorming and the rewriting process (you can ask AI for feedback instead of hiring an editor; this is still so super crazy to me).

Yep! I think the best writing is going to be created not by humans alone or AIs alone, but by collaboration between the two. We want to sit at that intersection and enable humans to go further, faster.