Hacker News Reader

Per your Android application data collection possibilities, the app may collect: Location data (Approximate location, Precise location), Personal info (email, user ids, sexual orientation), Device or other IDs. When you mean anonymous you mean anonymous only for other users of the platform directly?

And your website privacy policy states:

> We collect your email address number when you provide it to us by signing up to use the Service.

> We collect your biographical information, such as your name, date of birth, gender and geographical location, when you set up your account on the Service.

edit:

The following also sticks out for me:

> We may also process your Personal Data because it is necessary for our or a third party's legitimate interests and it’s not overridden by your rights. In this respect, we may use your Personal Data to:

> To contact you via email, postal mail, or telephone to learn more about your preferences, to conduct market research and learn more about how we can improve our offerings.

Your, or third party's interest do not override a EU residents right. You need a distinct opt-in for the reasons listed (marketing, research, etc).

And setting up an account on Apple with a privacy address asks me for my birthday, my gender, and my home address

Certainly doesn’t feel anonymous

We do not collect name or any identity related personal information. The only thing that could relate to your person is the email address. However, it is only used to contact our users as it is never used otherwise or shown to anyone inside or out of the platform.

The rest of the demographic data like age, location, preferences, etc are used for users to be able to conduct segmented audience queries.

Example: I want to ask a parenting question, to women between 30 to 45 years of age on Chicago area...

By the way, we do not intend to sell data or rely on ads business model and this is why we opted to incorporate a more transparent "paid" service. So, the user data is safe and not going to be exploited like on many other social media platforms.

Hope this helps understand the purpose of the type of data collected.

You may want to put this in your actual ToS.

Exactly this. The fact that the parent to your comment has spammed this answer a few times should indicate that there's a difference of opinion here.

This seems to be another platform trying to capitalize on the marketing gimmick of "privacy". It's become a very overloaded term. Apple's definition of privacy, is different from Google's, is different from Facebook's, is different from GrapheneOS', is different from...

It would be great if there was a privacy template that was recognized by a few, notable and trustworthy organizations. And that template would force the powers that be behind all of these "privacy oriented" sites and/or apps to be scrutinized in the same way, by a standard which would limit the definition gymnastics, at least a little. And that way said founders don't get to show up to forums and say "Well... By privacy we mean we're collecting all this information. And we have an open ended promise that's not protected in any way, shape or form - so while this is what we're telling you today, that may immediately go out the window when we're offered an undisclosed amount to sell all of the telemetry we weren't using to reverse engineer who you actually are".

> Intend

For now.

A truly anonymous service would start by not needing apps, which are notoriously leaky, on purpose.

You absolutely do collect identifying personal information, which is why I gave up and deleted the app.

It’s not anonymous when you know where I live, my birthday, and gender. There is only one person that matches those three data points.

I'd say "anonymously" should be removed from the title in light of that. Maybe "a space for asking questions while we mine your personal information" would be more appropriate.

See response below.

We actually don't mine personal information.

>> We do not collect name or any identity related personal information.... The rest of the demographic data like age, location, preferences, etc are used for users to be able to conduct segmented audience queries..... Example: I want to ask a parenting question, to women between 30 to 45 years of age on Chicago area...

Ok, so you able to identify user as living in Chicago area, by sex, by age, by email, but you still do not collect name or any identity related personal information? What kind of magic is this? Age, sex, area of living, email is definitely PII.

>> By the way, we do not intend to sell data or rely on ads business model.... So, the user data is safe and not going to be exploited like on many other social media platforms.

Many other social media platforms made this promise too at the start. Is there any reason to believe you that this will not change in a year, or after successful acquisition by some big player? Can you put it into ToS?

Evergreen idea (Q&A board), and congrats on the launch!

But you’re going to find it hard to get traction with it only being a native app. The magic of Stack Overflow, Quora, Reddit etc is people ask questions that are then indexed by search engines. This finds people who are knowledgeable on the topic because they’re researching something related, and it finds people asking questions about it that just asked google instead. The community comes to you.

(Over simplification of course, not everyone who answers on quora is an expert.)

Here, you’re betting on people opening your app, because they… just want to answer a question? A service like this lives or dies by how easy it is to find the people to do the free content work for you. You’ve made it a lot harder for them to do that.

This is an excellent point. I already ask questions on Google and I've never felt like I needed to pay to get a good answer. And by making it app-only, I don't even know what it looks like or what the quality of the answers are. I can't even imagine what an answer I'm willing to pay for would look like.

We do get your point and have given it ample thought.

And yes, we do plan to incorporate alternative infrastructure with less friction without having to download an app in the future.

Thank you so much for your kind words and wise point of view!

If the purpose of this is not to data mine and train AIs to sell, prove it with ToS. If it is, be upfront perhaps. Of all places here people kind of get how it works.

Such a cute landing page and all this dance just to entice humans to provide more input to AI models to sell. It’s actually kind of disgusting when you think what the ultimate plan almost definitely is.

Why is it automatically a bad thing if they use the data to train llms? Who is losing under this situation?

> Why is it automatically a bad thing if they use the data to train llms? Who is losing under this situation?

At a glance of the post and only a few seconds thought, I would guess the 'catchers' - the people who are being paid to answer the questions.

If they are being brought to the platform with the promise of regular work, building up a profile of expertise & reliability so they can generate more revenue on the platform, but under the hood their responses are just training LLMs so their work can be automated and they're eventually replaced by a catcher-bot trained on their responses... then I would argue they are losing in this situation, as they've invested time building up their profile only to risk having it stripped away.

It's a bit like building a cool new physical product and then going to a lot of effort to build a shopfront and presence on Amazon, only to have an AmazonBasics version of your product appear after a couple of months.

edit: not to say that this is what is happening with this site or that their intent is malicious! Just gaming out a potential case where there are some losers on the user side here.

It's not that it's automatically a bad thing. It's that it's a thing many people will object to, so if that's what's being done, it should be disclosed.

So, no our plan is not to train AI models to then substitute the catchers. In fact we have a completely different philosophical approach as all the emerging AI trends do. We believe AI is great technology and tool but our values and purpose are to improve the way humans interact online. So, we will use AI to make better matching and be able to provide appropriate people to the appropriate question, thus providing a better experience. But we want people to thrive on both ends of this service. Thats why we built it that way. Plus AI is great at many things but we also believe regardless of how powerful a model is, there are things it will never be able to replace, like an empathetic human response to a person in need. The human touch...

Interesting plan if you stick to it, appreciate the answer.

The words "safe" and "comfortable" indicate to me that this will be an overcensored hugbox where they'll sell your data to the largest bidder. Only having a mobile app is also a huge red flag.

The way the moderation system works is the following:

Any user may report content as 1) trash content or 2) offensive content.

A report immediately triggers a task that packages the reported question on a poll asking if the content is in fact what it has been tagged as.

The poll is presented to other users in the platform to validate if the content is or not what it was tagged as.

If it is, then the content is censored and the user faces consequences in his score up to eventually being banned.

It is ok to give your point of view even if someone does not agree with it. That is not trash nor an offense.

The purpose is to be able to be completely open and candid without being offensive.

And do you seriously think users won't abuse this to get people or viewpoints banned that they don't agree with? Remember, the downvote button in reddit was originally meant for downvoting unconstructive comments....

The "being open without being offensive" heavily depends on the eye of the beholder, and without consistent standards, there will be many arbitrary bans.

The key here is that reports go through a group of unrelated people who vet the claim. It is not one person that bans another. Plus, if a user reports something that is not true they bare a cost for that wrongful claim. Because peoples answers cost...

Looks nice, I’ve had this idea before and see the need for it, but don’t have the skillset to build it.

How anonymous is anonymous? Are the employees at Throw able to see the questions I ask, since my Apple ID is associated with the app download and presumably my Throw account?

Yes, we did a test post a year ago without a product... But now were live! :)

A test post without a product should never have been a Show HN! That's against the rules (https://news.ycombinator.com/showhn.html) and we should have caught it.

With respect, I would like to see some evidence of your assertion that anonymous communication results in higher quality content than when the question answerer is known. The highest quality information content I'm aware of is in textbooks and scientific journals where the author's reputation is on the line. Contrast that with the wall of a public restroom, which is pure shitposts.

You might be correct if the assertion is the latter is more reflective of how an average Internet user actually feels at any given moment but won't admit in polite company.

I don't know exactly why but usage of corporate memphis makes it a bit jarring for me.

First of all, this is a good idea and a cool app.

IDK if this is part of your plans, it seems like a perfect way to gather LLM training data... And a reasonable use case for LLM integration. And its not in explicit conflict with your privacy policy as far as I can tell, as long as you self host.

Everyone is training LLaMA on ChatGPT Q+A pairs, but a big corpus of filtered, purely human, human rated questions+responses would be so much better.

How do you defend the service against answers copied from chatGPT?

How do you defend against trolls? And how do you distinguish them from genuinely stupid people?

https://m.youtube.com/watch?v=EShUeudtaFg

What do you do with all the teenagers (?) asking "questions" about sex?

Good question.

The way the platform is designed makes it hard to implement ChatGPT or other AI services (at least on an automated manner) because the person that asks the question selects the media format he wants the answers back. On top of this, each person that answers does not pick the questions he wants to answer, we do the matching. So one moment you're being asked to answer in text the next in voice and the next record a video. It's very human interactive.

As for filtering trolls and stupid people, you can upvote, downvote, and report 1)trash content or 2) offensive content. Our algorithm constantly takes all of this and many other metrics to determine the better performing people and filter out the opposite.

The way the moderation system works is the following:

Any user may report content as 1) trash content or 2) offensive content. A report immediately triggers a task that packages the reported question on a poll asking if the content is in fact what it has been tagged as. The poll is presented to other users in the platform to validate if the content is or not what it was tagged as. If it is, then the content is censored and the user faces consequences in his score up to eventually being banned. It is ok to give your point of view even if someone does not agree with it. That is not trash nor an offense. The purpose is to be able to be completely open and candid without being offensive.

Teenagers asking about sex... Well as long as its not offensive or disrespectful it is fair play. Even adults ask about sex and this is a place where you can do so comfortably and not be criticized.

This seems very similar to Secret [1]? IIRC there were some competitive apps at the time, YikYak maybe?

Secret sounded scandalous, and maybe that was the goal, but it actually did a really great job of anonymous Q&A, and was full of posts from people saying "I have this private/odd/embarrassing situation what should I do", and more often then not the thread contained some really insightful and thoughtful stuff.

I was disappointed when it shut down, is the idea of this to be similar?

[1] https://en.wikipedia.org/wiki/Secret_(app)

> we have built a sophisticated moderation protocol to neutralize and quickly ban people that contribute negatively as it is a priority people feel safe and comfortable in this community.

This seems a firm stance against objective truth. Your scoring algorithm benefits answers that people want to receive, and you actively penalize truthful but disliked answers.

"I feel I am fat. What should I do?"

The truthful answer is something like diet modification, but the answer most likely to make someone feel safe and comfortable is "no, you're not fat".

The way the moderation system works is the following:

Any user may report content as 1) trash content or 2) offensive content.

A report immediately triggers a task that packages the reported question on a poll asking if the content is in fact what it has been tagged as.

The poll is presented to other users in the platform to validate if the content is or not what it was tagged as.

If it is, then the content is censored and the user faces consequences in his score up to eventually being banned.

It is ok to give your point of view even if someone does not agree with it. That is not trash nor an offense.

The purpose is to be able to be completely open and candid without being offensive.

Hope this helps answer your question.

cool idea but why does it have to be exclusively an app?

Thanks! Eventually we do want to develop a web app but given that it does have that social media (multimedia and user intractability) component of it the best way to approach it and provide the best user experience was through a mobile app.

> provide the best user experience was through a mobile app

Having to install an app is always a bad user experience. It's a dealbreaker for me for just about every case, and in general it's got to add a lot of friction when you're trying to get people to sign up.

It would be interesting to hear some honest discussion of the tradeoffs. I assume most companies prefer me using an app because it lets them run arbitrary code on my phone, access contacts and whatnot, and send me notifications. Is it also materially easier to write the code for apps? Is that a legit reason why companies prefer it? And is whatever companies are getting out of forcing their users onto an app worth the potential users they lose because of it?

I often hear this rationale when starting with a social app, but I think that you may find that you pigeon hole your age demographic by only doing an app. Folks that are in the prime professional age bracket (gut check, late 20's to 30's) are going to spend more time on a larger screen device than they are on a smaller screen device. At the very least, if you're not planning on developing a tablet app right away, having a browser option is at least a stopgap between running a phone-size app on a tablet, or not at all while you work on that.

No idea if you're building your app in a web framework that ports to mobile easily, or what your choices are around that, but I'd highly recommend reconsidering not having a web app if it's at all reasonable for your time table.

For me? I hate typing on my phone, so I probably would pass on this until a web app is supported.

What exactly does a mobile app do that a web page can't?

Web pages play video and have interactive components just fine.

Feedback: Your website doesn't have much content. Instead of a flowery text description of the product, you could just show some of the best or most upvoted content you have right now.

I tried to download your app but it doesn't work. It says "not available for your device", but I have a brand name phone with the latest version of Android.

It may be that you're outside of the USA. Were currently providing service on the USA only.

Do you have a trademark on the word "throw"? The page title is currently "Throw™" and many of the instances of Throw have the (tm), but I can only find the trademark application for your logo design (serial 97072015).

Yes we do.

Do you have a link or serial number for the application?

FYI this seems to be an app for cellphones. No Q&A on the web site, that's just a landing page.

Not that there's anything wrong with that. I just thought it was a web site.

> There is no limit to the type of questions or the topic you may ask

Seems very broad and general. The beauty of Reddit is going into subreddit's and asking a question to a specific community. Reddit is also fairly anonymous too.

I understand that this is a commercial platform, but when cash is introduced to a Q/A system... incentives are misaligned.

It seems communities without monetary gain (Hacker News, Reddit etc...), have very little incentive to lie...

Do you have objective evidence showing that common, non-business users are willing to pay for crowdsourced answers?

> Depression, anxiety, and loneliness advice

How do you know if your users are qualified to give that advice?

Through several metrics we continuously measure we determine reliability of users. So, empathetic and positive users that are constantly providing good answers receive a higher score and in the contrary the opposite is true. The higher the score the more questions the user gets assigned and the more he earns per response. People that are better qualified thrive and people that are not fall behind on this meritocratic system where users have to earn their place.

How do you know my answer was “good”? I’ve seen highly upvoted answers in my area of expertise on a number of platforms that were wrong to dangerously wrong. How will you know the difference?

The person that upvotes the answer is the person that asked the question. So, if it is useful for that person the purpose was served.

That’s not how “good” answers work. If I tell you the sky is green and that makes you happier than hearing that it’s actually blue, your site no longer has value for getting or understanding answers. That is a fatal flaw, and I’m not sure what you’re offering that, say, Reddit isn’t, aside from an inability for anyone else to correct clear mis/dis-information.

We do get your point and it is a fair point. However, an important factor while getting feedback or answers is getting a diverse and unbiased set of points of view which will likely enrich the person whose seeking that feedback. It may very well be the case that some of those points of view are incorrect but it may very well be the case that some are not just helpful but accurate. The key is to be able to have a large enough, diverse and unbiased sample. Of course, equally important the user's ability to interpret, filter and report in such case the data. And that is what were trying to do here. Hope this helps.

Have you tested it with real users before launching? What did you learn that was different from your initial theories (on e.g. incentives, handling low value content, ...) for realizing your ideal?

finally, the real engine behind chatgpt

Useless without a desktop app. Who even uses a smartphone (too small screen and stupid keyboard) in 2023 ?

Yeah I never see anyone on their phone

Replying from my smartphone.

Show HN: Throw – A space for asking and answering questions anonymously