This is getting to the antivirus bundle level of adding pointless features. I want grammarly to... check my grammar. I don't want it to write for me or suggest things.
No sepulcator company gets profitable by shipping just a sepulcator. A sepulcator absolutely must have AI, monthly subscription, cloud services and - up until recently - has to be blockchain-based.
Perhaps you do, but I think this misses the point. For-profit writing is the most successful use case for LLMs today. A significant proportion of all the docs I see at work reek of LLMs. A fair amount of articles you read in the media are written by LLMs. Lawyers use it for legal briefs (sometimes with comical results). Doctors use it for patient notes.
Basically, a significant portion of the population doesn't like writing or isn't good at it and really wants a "get it done" button. I might not love it, but the market is there.
So Grammarly is addressing a very real need. Further, it's really the only way for them to stay relevant, because you're getting AI editing / writing features in Gmail, Docs, Office 365, etc.
> Why would I use Grammarly/Superhuman for writing with LLM assistance, when I have an out-of-box alternative that, at worst, is equal?
I think the answer is basically that they have brand recognition and they're trying to ride it. Right now, they have two bad choices: become irrelevant more quickly by having a product that's inferior to built-in LLM tools, or become irrelevant more slowly by having a tool that's comparable (and also works anywhere on the internet, not just on specific websites).
> So Grammarly is addressing a very real need. Further, it's really the only way for them to stay relevant, because you're getting AI editing / writing features in Gmail, Docs, Office 365, etc.
They are a feature, not a company, with my apologies to Jobs. To your point, software and tools with native writing functionality can incorporate their own LLM support, as can native apps on mobile and desktop. Anything local will eventually be on device imho as model efficiency improves, or perhaps in browser (if not making API calls).
Flagging likely spelling and basic grammar errors are pretty much incorporated into most word processors at this point. I may or may not choose to ignore them. But they work pretty well and I'm unlikely to use an external tool.
I did write for a while for a tech site that had some Wordpress add-on that was oriented to making my writing, I guess, more friendly to an 8th grade level. I ignored it.
I mostly just rely on browser to check my spelling. In work well copy paste it to World if I really want to get fancy... I did years ago use Grammarly for my thesis and spam finally ended after years some time ago...
You're right that it's mostly just the browser. Other than one or two text editors, I don't think I even have a word processor installed on a system at this point.
Maybe it's just because I'm a reasonably decent writer but I used to know someone who was adamant about using Grammarly because it would increase traffic to my website--and I was basically "don't care."
ADDED: Because it would make the writing friendlier to more people.
I'm not sure what the right word is. At least according to the tool's output, it's about taking the writing down to a lower grade level.
I agree about the authentic voice. At one point, I got pretty unhappy with a fairly junior editor's changes to something I wrote. I was just going to let it go but my manager took it up with the editor's manager in the vein of "this isn't acceptable."
i mean videos were constantly popping up for me on mobile Spotify so it kind of had already started happening. I had to actually investigate how to turn it off.
“Any sufficiently complicated concurrent program in another language contains an ad hoc, informally-specified, bug-ridden implementation of half of Erlang.” — Robert Virding
Responding to you and fullshark, I'm not criticizing, only observing. Just as there is some evolutionary pressure causing carcinization, it's interesting to consider what pressure pushes things in the directions of email and LLMs.
I don't know what it is, but would love to hear others' ideas.
I think "email" is a bit of a overly specific term, but if we take a small step back, communicating with other humans is usually the most important part of any piece of software.
I have a feeling these things will spend 99% of their processing time reading other LLMs outputs.
Resumes written by LLMs and read by LLMs
PR summaries written by LLMs and read by LLMs
Emails written by LLMs and read by LLMs
...
Everything could just be a few bullet points... these things were already 90% posturing and trying to sound fancy by using convoluted sentences and big words, now that it's been automated what's the point
Harper is a nice alternative, but it's still rough around the edges.
For instance, if you have a misspelled word, and the correction options come up, you can't get out of them and return to where you were by using the keyboard. You can hit Escape to close them, but it doesn't restore your place in the text field, so you have to use your mouse to get back where you were.
As a programmer who tries to use the keyboard as much as possible, this (incredibly easy to fix, I'm sure) bug drives me crazy! Almost enough to make me go back to Grammarly.
That seems to me not like a "rough around the edges" thing but "most basic, table-stakes feature". If you cannot resume typing after either cancelling a correction, or doing a correction, I'd say it is very broken and not ready to be marketed as a functioning tool. I mean, it's supposed to help you write, not make it more cumbersome.
For those who find red squiggly lines too distracting, I built a lightweight Chrome extension (<1MB) that approaches grammar correction in a minimalistic way: Highlight text -> Select "Correct grammar" from the toolbar -> Replace text.
The quality is unmatched because it uses SOTA models like GPT-5, Claude, and Gemini.
Grammarly is a powerful writing assistant that offers grammar, punctuation, and style suggestions to improve your writing. It’s an essential tool for maintaining your unique voice in a world filled with generic, LLM-generated content. With Grammarly, you can ensure your words stand out and resonate amidst the noise.
"Grammarly announced Tuesday the acquisition of email client Superhuman in a push to build out its AI for its productivity suite. Neither companies provided details about the financial terms of the deal..... Superhuman was founded by Rahul Vohra, Vivek Sodera, and Conrad Irwin. The company raised more than $114 million in funding from backers including a16z, IVP, and Tiger Global, with its last valuation at $825 million, according to data from venture data analytics firm Traxcn." [1]
Interested to understand what would be the terms of the deal if Superhuman was valued at $825mm and what the founders cleared if the all the VCs rounds had 2-3x liquidation preferences.
The name Superhuman makes a lot more sense for a company with a suite of AI productivity products. The "Grammarly" name was too focused on their original use case of just improving writing.
I know it has a positive connotation with super heroes in US culture but for me it sounds like Übermensch. Especially as it is the direct opposite of "subhuman".
Plus outside of tech bro circles, people either actively hate generative AI or are at least super annoyed by the over-hype of it. Duolingo went all in on AI and got a huge shitstorm.
Branding your company on a current hype that might either burst soon or/and leave lots of people unemployed is maybe not a wise decisions.
Are you a native English speaker? I can't think of a scenario where "superhuman" has negative connotations in American English. When we say someone has superhuman skill, or speed, or strength, it is always a positive thing.
There are instances where the term is used in a positive sense, yes, but those are limited in scope. "Superhuman strength" rather than just "superhuman".
"Superhuman" on its own is a term that has long been tightly associated with a wide variety of horrible things. Eugenics, for example.
I honestly think most American English speakers are not thinking about eugenics when they hear that term. I believe you when you say that it has those connotations for you but I think you are in a small minority.
I like superhuman as an adjective, it implies some quality of a man is superhuman, but a superhuman as a standalone item is very frightening as it spells obsolescence for the rest of us ( I guess this is the plot of X-Men )
Nope. You must be thinking of the terms "Untermensch" (used a lot by Nazis) and "Übermensch" (introduced by Nietzsche, and rarely used by Nazis). "Supermensch" was never used at all.
What do you think ‘super’ means ? It is latin for ‘over’, wich in German is über. In English it has come to take on a broader meaning, but Nietzsche’s übermensch is called ‘superman’ in most English translations, even if ‘superhuman’ would be more accurate.
GP doesn’t imply Nazis used ‘Supermensch’, just that the ‘superhuman’ translates to übermensch and that the branding might evoke this concept for European ears.
Growing up in the 90s in Sweden, we definitively were taught that "Übermensch" ("Övermänniska" in Swedish, literally "Above Human") was something the Nazis promoted during their time, together with demoting "Untermensch". Maybe that's wrong, and if so I thank you for the correction, but "Superhuman" does give me similar vibes regardless, not because of the exact wording, but because of the ideas/concepts.
Nietzsche’s sister tried to garner favor with the Nazi regime. After Nietzsche’s death, she took his notes, published them under the title “Will to Power” and made it all sound as though Hitler was the fulfillment of Nietzschean ideas. Even scholars who built their careers on Nietzschean philosophy fell for this. For example Ayn Rand. So your teachers were in good company. In truth, everything about the Nazis would have made Nietzsche sick to his stomach: group-think, racism, big government, socialism, robbery, personality cult, lack of intellect, mass appeal, Gleichschaltung, militarism.
Elisabeth and Bernhard were rabid nationalists and antisemites long before the NSDAP. They established their vegetarian-antisemitic-'Aryan' colony in Paraguay in 1887, two years before Adolf Hitler was born.
It failed for financial reasons and the rather harsh environment. They ditched the vegetarianism and started selling meat to get some money, spiraling into alcohol and morphine abuse. In 1889 Bernhard killed himself with strychnine, after which Elisabeth started her career as a fake chronicler by writing a book aimed at creating a much nicer and 'Aryan' image of Bernhard and the colony than the truth would have allowed.
As you allude to, Friedrich Nietzsche poured buckets and buckets of abuse over people like his sister and other german nationalists, refusing for the entirety of his life to identify as german, and towards the end of his life he even claimed to be a polish nobleman, free of the tainted blood of the germans.
HN can be so funny sometimes. An actual German says "hey maybe this specific word shouldn't be used", and a random follows it up with "Nope, you're wrong." lmao
Why would Germans be an authority on what words should or shouldn't be used in English?
This is sort of a reverse version of the very common trend of American political correctness / sensitivity language being exported around the world. Our ancestors committed heinous crimes, therefore we get to tell you how to speak, even though you had nothing to do with it.
A German person just said that it gives them nazi viber, nothing about English words that should be used.
Person above argues that the words are different therefore such connection can't be made which is just... wrong because they reply in a thread where someone literally said they made that connection.
In short, we're explicitly talking about what Europeans see (me too, I'm not German), not what Americans should do.
The comment I'm replying to says, verbatim, "hey maybe this specific word shouldn't be used" (as a paraphrase of that commenter's understanding of the argument being made by the German). That is what I'm responding to.
If someone says a particular word or phrase is problematic for them, no one can tell them they're wrong. You cannot dictate how other people respond to language, and it's really weird to see people trying to do that.
Sure, I can't tell them they're "wrong", i.e. I think the self-reported subjective feeling is probably accurate.
What I object to is the implication that Americans should punish themselves by refraining from using normal words in their own language because Germans feel bad about something Germans did.
The implication is that if they want to market to Europeans (which I'm sure they do), they probably shouldn't use that word. I agree Americans see it in a positive light, including me, though I find superhuman generic to the point of background noise.
Because Ubermensch comes from Nietzche a century before the Nazis, as said, and had also a big influence on anarchists. No-one suggested that "Superhuman" shouldn't be used, either. A some point people need to put things in context and not "get the creeps" over any little things. I am sure that Germans don't even notice all those "Volkswagen" around them...
I'd say stay away from policing at least one's own evils. People that are idiotic enough to connect the superhuman in this context to Naziism should stay away from policing any meanings (but now I'm guilty of policing what people should police)
You cannot expect other countries to stop using normal words because they remind you of the bad things your country did.
Shame for what Germany did during the Nazi regime is something for Germans to bear, not Americans. We are not at fault for that, and we have no obligation to change our own culture to accommodate your guilt.
That’s quite a leap. The parent commenter didn’t call for them to withdraw the branding, they were just sharing something interesting and unique about their perspective as a German.
Just to be clear, I never said that the word should be banned.
I am not sure how important the German or general European market is so hard to say whether it even should be a consideration for Grammarly.
That said the ideas of some people being intrinsically better than other people isn't specific to Germany. Eugenics used to be popular in many countries including the US. It is very advisable for other countries to learn from German history so our mistakes are not repeated.
They rebranded but the pricing and packaging options are still stupid and poorly segmenting the product. Basically, there's still a different (and higher) price to get email or not.
On top of that, Grammarly's 3 products (Coda, Go and Grammarly itself) all look the same. What are honestly the differences between these three:
* Product 1: Everyone’s favorite AI writing partner.
* Product 2: The all-in-one AI workspace for teams
* Product 3: AI that works in every app you use
These people probably know what they're doing, and probably also have a lot of restrictions from previous users so migrating is hard. But I'd unify the value prop, just one product: AI Superpowers, including all the "features": Email, writing assistant etc.
And then the packaging done per Enterprise features and metered.
Given their extensive expertise in browser and OS plugins, I understand this move.
You can foresee a challenging future for the Grammarly product for a long time. Now that the "improve writing with AI" feature is everywhere, there are fewer reasons to pay for their subscription (e.g., I didn't renew this year because I have multiple AI subscriptions, and Grammarly was the least critical of them).
However, for me, the main advantage of Grammarly was the user experience of having mistakes and suggestions inline and just a click away while editing, as well as the quality of the suggestions (with an LLM chat, there's a lot of trial and error and junk you need to filter out).
I understand their move, but I wish they had developed a good minimalist native text editor with the same Grammarly suggestions and click-to-correct interface.
That is my number one issue with startups. They all start minimalist and end up bloated, some sooner than others, and what made them great disappears behind all this bloat. See: tyranny of the marginal user.
I've been a long time use of Grammarly. I have used it for most of my books. I have a 150 week long streak using it. I've used it as an example of good ux for AI.
Now I'm considering vibe coding a replacement that just does grammar fixing...
Ironic, because I don't want the AI writing, only the grammar checking
Stop using Grammarly. There are better options available that don't exist just to collect your data to sell it to the highest bidder or feed it into an LLM.
The company is being rebranded, not the product. Makes total sense, considering the brand equity, and also them going in the direction of productivity suite. Could be interesting.
Moving to "AI" and away from a well-known brand smacks of desperation. Makes me wonder if the industry-wide trend of shoving AI into every product and feature, and channelling all investment into AI, is equally desperate.
I get that software companies are rebranding products with superhero/god terminology to increase their perceived value and raise margin, but its not working for me because they are losing product differentiation. Why would I choose this app among the dozens of other tech products that promise godlike AI capabilities?
One day, we will see a demand for services that are the opposite of "Superhuman". For example, a service like: "Deteriorate this text and make it look weirdly human. Add some typos and errors here and there, so that the final output looks 100% human-written."
Even if we haven't hit the LLM ceiling, we've hit a ceiling on branding for sure. I'm interested to see where these names go next. Uberbeing! Omnipotence Plugin!
I absolutely hate it when companies rename themselves. I know a company called an extremely stupid name by its young founder and they did not rename for decades and are now worth a bit short of $4T.
Why do the smaller ones constantly need to change their name. Like that changes anything in their substance.
AIs are pretty bad at rewriting, they always pick gimmickey marketing words. I always thought of Grammarly as a premium entry into that segment for proper professional writing. Shame it’s going the way of slop.
This stuff happens all the time for a tech blog they sure don’t know the industry.
Google acquired Flutter, a desktop app which let you control your music playback with hand gestures, only to reuse the branding for the yet another ghost town framework of theirs.
That's not how it works today.
No sepulcator company gets profitable by shipping just a sepulcator. A sepulcator absolutely must have AI, monthly subscription, cloud services and - up until recently - has to be blockchain-based.
Basically, a significant portion of the population doesn't like writing or isn't good at it and really wants a "get it done" button. I might not love it, but the market is there.
So Grammarly is addressing a very real need. Further, it's really the only way for them to stay relevant, because you're getting AI editing / writing features in Gmail, Docs, Office 365, etc.
To me it is exactly why this move doesn't make sense.
Why would I use Grammarly/Superhuman for writing with LLM assistance, when I have an out-of-box alternative that, at worst, is equal?
They can't even compete with pricing, because they need to use their competitor models
I think the answer is basically that they have brand recognition and they're trying to ride it. Right now, they have two bad choices: become irrelevant more quickly by having a product that's inferior to built-in LLM tools, or become irrelevant more slowly by having a tool that's comparable (and also works anywhere on the internet, not just on specific websites).
They are a feature, not a company, with my apologies to Jobs. To your point, software and tools with native writing functionality can incorporate their own LLM support, as can native apps on mobile and desktop. Anything local will eventually be on device imho as model efficiency improves, or perhaps in browser (if not making API calls).
I did write for a while for a tech site that had some Wordpress add-on that was oriented to making my writing, I guess, more friendly to an 8th grade level. I ignored it.
ADDED: Because it would make the writing friendlier to more people.
I agree about the authentic voice. At one point, I got pretty unhappy with a fairly junior editor's changes to something I wrote. I was just going to let it go but my manager took it up with the editor's manager in the vein of "this isn't acceptable."
https://en.wikipedia.org/wiki/Jamie_Zawinski
I don't know what it is, but would love to hear others' ideas.
Resumes written by LLMs and read by LLMs
PR summaries written by LLMs and read by LLMs
Emails written by LLMs and read by LLMs
...
Everything could just be a few bullet points... these things were already 90% posturing and trying to sound fancy by using convoluted sentences and big words, now that it's been automated what's the point
For instance, if you have a misspelled word, and the correction options come up, you can't get out of them and return to where you were by using the keyboard. You can hit Escape to close them, but it doesn't restore your place in the text field, so you have to use your mouse to get back where you were.
As a programmer who tries to use the keyboard as much as possible, this (incredibly easy to fix, I'm sure) bug drives me crazy! Almost enough to make me go back to Grammarly.
The quality is unmatched because it uses SOTA models like GPT-5, Claude, and Gemini.
Yes, you can use your own API key as well.
https://jetwriter.ai
Interested to understand what would be the terms of the deal if Superhuman was valued at $825mm and what the founders cleared if the all the VCs rounds had 2-3x liquidation preferences.
edit: added source
[1] https://techcrunch.com/2025/07/01/grammarly-acquires-ai-emai...
I know it has a positive connotation with super heroes in US culture but for me it sounds like Übermensch. Especially as it is the direct opposite of "subhuman".
Plus outside of tech bro circles, people either actively hate generative AI or are at least super annoyed by the over-hype of it. Duolingo went all in on AI and got a huge shitstorm.
Branding your company on a current hype that might either burst soon or/and leave lots of people unemployed is maybe not a wise decisions.
I'm not sure about this. I'm a US citizen, but it absolutely does not have positive connotations to me at all. It has very negative ones.
Yes, I am. Born and raised in the US.
There are instances where the term is used in a positive sense, yes, but those are limited in scope. "Superhuman strength" rather than just "superhuman".
"Superhuman" on its own is a term that has long been tightly associated with a wide variety of horrible things. Eugenics, for example.
GP doesn’t imply Nazis used ‘Supermensch’, just that the ‘superhuman’ translates to übermensch and that the branding might evoke this concept for European ears.
It failed for financial reasons and the rather harsh environment. They ditched the vegetarianism and started selling meat to get some money, spiraling into alcohol and morphine abuse. In 1889 Bernhard killed himself with strychnine, after which Elisabeth started her career as a fake chronicler by writing a book aimed at creating a much nicer and 'Aryan' image of Bernhard and the colony than the truth would have allowed.
As you allude to, Friedrich Nietzsche poured buckets and buckets of abuse over people like his sister and other german nationalists, refusing for the entirety of his life to identify as german, and towards the end of his life he even claimed to be a polish nobleman, free of the tainted blood of the germans.
This is sort of a reverse version of the very common trend of American political correctness / sensitivity language being exported around the world. Our ancestors committed heinous crimes, therefore we get to tell you how to speak, even though you had nothing to do with it.
Person above argues that the words are different therefore such connection can't be made which is just... wrong because they reply in a thread where someone literally said they made that connection.
In short, we're explicitly talking about what Europeans see (me too, I'm not German), not what Americans should do.
The comment I'm replying to says, verbatim, "hey maybe this specific word shouldn't be used" (as a paraphrase of that commenter's understanding of the argument being made by the German). That is what I'm responding to.
If someone says a particular word or phrase is problematic for them, no one can tell them they're wrong. You cannot dictate how other people respond to language, and it's really weird to see people trying to do that.
What I object to is the implication that Americans should punish themselves by refraining from using normal words in their own language because Germans feel bad about something Germans did.
Shame for what Germany did during the Nazi regime is something for Germans to bear, not Americans. We are not at fault for that, and we have no obligation to change our own culture to accommodate your guilt.
I am not sure how important the German or general European market is so hard to say whether it even should be a consideration for Grammarly.
That said the ideas of some people being intrinsically better than other people isn't specific to Germany. Eugenics used to be popular in many countries including the US. It is very advisable for other countries to learn from German history so our mistakes are not repeated.
I'm sure a large population of second language English speakers is a huge market for grammarly, no?
On top of that, Grammarly's 3 products (Coda, Go and Grammarly itself) all look the same. What are honestly the differences between these three:
These people probably know what they're doing, and probably also have a lot of restrictions from previous users so migrating is hard. But I'd unify the value prop, just one product: AI Superpowers, including all the "features": Email, writing assistant etc.And then the packaging done per Enterprise features and metered.
Search results are optimized based on inferred intent, and the intent of most people searching for "superhuman" will be the Grammarly app.
You can foresee a challenging future for the Grammarly product for a long time. Now that the "improve writing with AI" feature is everywhere, there are fewer reasons to pay for their subscription (e.g., I didn't renew this year because I have multiple AI subscriptions, and Grammarly was the least critical of them).
However, for me, the main advantage of Grammarly was the user experience of having mistakes and suggestions inline and just a click away while editing, as well as the quality of the suggestions (with an LLM chat, there's a lot of trial and error and junk you need to filter out).
I understand their move, but I wish they had developed a good minimalist native text editor with the same Grammarly suggestions and click-to-correct interface.
Now I'm considering vibe coding a replacement that just does grammar fixing...
Ironic, because I don't want the AI writing, only the grammar checking
Superhuman's entire value proposition was "staying out of my way".
All I can do is hope that the latter rubs off on the former, and notthe otehr way around.
Most people apparently don't care
Why do the smaller ones constantly need to change their name. Like that changes anything in their substance.
Google acquired Flutter, a desktop app which let you control your music playback with hand gestures, only to reuse the branding for the yet another ghost town framework of theirs.
Either way, I'm not so sure spellchecking and the Superman has the kind of connection that would make for efficient branding.