luisamodio

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.

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.

Yes we do.

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...

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.

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.

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.

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

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.

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...

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!

See response below.

We actually don't mine personal information.

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.

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

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.

Not at all like 4chan. For a start it’s not a forum. And in every other aspect also lot like 4chan.

100% anonymous. And actually it’s nothing like 4chan.

Thank you for your comments and wishes! :)

The thrower gets to asses if the answer was a good one, or a bad one or just mediocre. This has no effect on deserving a payout as long as there was a genuine effort to respond the question. It does however influence the scoring (karma) system with which the catcher is directly compensated.

The general rule is that answers that are subject to refund are answers that fall into 2 categories 1) it’s gibberish/trash/unrelated content (for example someone that repeatedly copy-pastes the same paragraph with no relation to the question hoping to earn money), and 2) offensive content (anywhere from trolling, racism, misogyny, etc).

As for your second claim, yes, you pay regardless if you find it useful (unless of course it’s any of the two categories I previously mentioned).

And the reason for this is because getting a large pool of diverse and unbiased data puts you on a better informed position and may open your eyes to angles you wouldn’t have considered otherwise. Regardless if you dislike them or disagree with them.

You’re not always going to hear what you want to hear…

Yes correct!

Simplified, the way this works is the question and the answer get packaged inside a new question that is asked to a number of people just as if it were a normal question with a yes or no poll.

“Is the response to this question offensive?”

(Yes it is offensive / no it is not offensive)

So, in short, a crowdsourced jury!

Good question!

So the way this was designed, is so that the unit (answer) cost is fairly cheap (below the dollar), so that volume is incentivized.

In turn, the payout amount which is a fraction of that per unit cost was calculated to be an attractive enough figure. Such that if you don’t do this ala Mechanical Turk you get some decent extra cash out of it and if you do get very involved you can even make a salary out of it.

However, there is a dynamic algorithm that takes into account several factors primarily consistency on the quality of the answers that directly impacts each users final compensation.

Thus promoting a virtuous cycle of better content.

Thanks so much for asking and hope this makes sense!

Some examples are:

-Sourcing ideas -Depression, anxiety, and loneliness advise -Love life & relationship advise -Embarrassing situations advise -Parenting advise -Feedback on content before posting on social media (content creators or public figures for example…) -Product / service feedback Market research -Difficult life situation advise -Important decisions advise -Curious questions -Random questions -Sourcing memes or jokes -Trivia

And honestly many more, it’s up to peoples creativity as the platform provides much flexibility.

Hope this helps and thank you so much for asking!

Hello!

To your first question, yes! The compensation paid is a dynamic fee determined by the score (karma of a kind) a user has.

There was a previous question on this page asking what determines quality. And fundamentally it is the fulfillment/satisfaction of the needs of the person who asked.

So people who give better answers on a consistent basis will earn more than people who don’t. This creates a virtuous cycle.

As for answers that are either trash or offensive, there is an extremely-viusal “report” button on every answer that if hit immediately triggers an investigation. If found true, you’ll get refunded for that answer and the offender will be drastically punished if not banned as there is minimal tolerance for this.

Finally, some examples are: -Sourcing ideas -Depression, anxiety, and loneliness advise -Love life & relationship advise -Embarrassing situations advise -Parenting advise -Feedback on content before posting on social media (content creators or public figures for example…) -Product / service feedback Market research -Difficult life situation advise -Important decisions advise -Curious questions -Random questions -Sourcing memes or jokes -Trivia

And honestly many more, it’s up to peoples creativity as the platform provides much flexibility.

Hope this helps and thank you so much for asking!

Fundamentally, if the response was useful or not to the person who asked it.

The platform is prepared positive and negative tags that strongly influence each user’s score (sort of like karma).

So what we value mostly is the effort taken to satisfy and meet the query’s needs.

As for trash and offensive content you’d simply hit the report button, an investigation will immediately take course, you’ll be reimbursed that answer and the offender drastically punished if not permanently banned as there is minimal tolerance for this.

Hi!

1) The main differences (both for the person that’s asking and the person that’s answering) are:

- Agile & guaranteed number of responses (no need to be a HOT question/topic) - Compensation model - Flexible media types (ask and answer in text, audio (voice), video, poll, and more. - Full anonymity (no social bias and/or no profile/façade to upkeep) - Robust moderation protocol - Gamified and engaging user experience - Target a profiled audience

As for incentives if you’re answering (there’s 3):

- Sake of entertainment and curiosity - To help others - Cash

2) As for unrelated (gibberish/trash) and offensive answers, all you need to do is click an immediately visible “report” button that will trigger an investigation immediately. That answer will be reimbursed and the responder drastically punished if not banned (as tolerance for this is very limited).

3) Plagiarism is not a main area of concern to be honest, in the end if a quote from someone famous or a fragment from Wikipedia are a great answer for someone that’s fine. What mainly drives the score algorithm is how useful the answer was to the person that asked the question, where it came from, not so much.

Thanks so much for asking!

Email is only for launch purposes.

In the end, nor email nor any other attribute that ties to your identity will be shown or used within the platform.

It’s fair to be skeptical.

Maybe we’ll change your opinion when you see it in action.

Thanks for commenting still.

Hello!

If that we’re to happen it would be partly true that only product or market people would be interested and it would become a noche platform say “Perksy” which does precisely what you describe.

We do believe it is very unlikely for this to happen as we purposely decided not to launch in a niche or with an audience of this sort.

So yes, there will likely be some marketers that use this for market research but it will be very unlikely they drive the content in the community.

We really appreciate the feedback!

We’re not targeting market research in particular, it is just one of the use cases.

Thanks for the feedback! We will surely keep this in mind!

Thanks for the feedback!

This is probably the biggest challenge we faced we decided to create Throw.

We have addressed the problem face on and believe we have developed a robust solution to outrun this kind of behaviors.

Our moderation (filtering out and reward) algorithm very quickly promotes or demotes users based on the quality of their content.

And it is the community itself (with supervision of our team) that drive the tone.

Also thanks for the wishes!