Three Warning Emails This Morning

Just to be clear how many photos did you submit at this stop.

three. two cropped for pogo and one cropped for ingress

1 Like

Thanks
For the little it is worth that is not spamming in my book

2 Likes

My idea of spamming seen in voting this morning: nine photos to review. Somone trying to do their ten all at once? Edit: suppose I should have reported that.

4 Likes

ugh I feel ya. I had a warning yesterday morning, thankfully the team accepted my appeal. I really wish they didnt have to fill us with so much anxiety when we do 2 weeks of hard work for their challenge… It’s really taking the fun out of participating.

5 Likes

Ugh I forgot to mention to look good on gifts in my appeal. Even photos that look good on the spinner can look terrible on gifts. I had an example of that somewhere which of course I can’t find right now.

4 Likes

Ohh I reviewed that too, and yeah they were AWFUL as well, not centred, not clear, glare, blurry, really low quality spam clearly for the challenge purposes

@cyndiepooh Im so sorry this happened to you, clearly this was not deserved and I can really see how upset you are. hugs the pictures you showed of that sign look like a huge improvement and I love the way you got it framed in the pokestop image! High quality additions to a “meh” kind of waypoint

And yeah I have bo idea how to make anything look good on a gift, they’re always distorted and awful

4 Likes

Square is the only way to not get a mushed ‘gift’. The app takes the photo and resizes it square and then crops it 4-3. If you start with square at least there’s no distortion.

7 Likes

Now… If Niantic seriously wanted good photos for their POIs that also look good in their games, this is the type of information they would supply to explorers. Thank you.

Instead it’s just left to dumb chance, and the fates of more knowledge explorers who risk abuse" warnings" by submitting edits. That’s not efficient, and definitely not enjoyable for those who get warned.

4 Likes

@cyndiepooh so sorry for you :mending_heart:
Especially with directing you to the criterias, I would bet that you have written them (kidding, I know you haven’t)

1 Like

How many location edits have you done?

2 - one was rejected by the community, although it is clearly a correction on satellite view. the other, submitted on the first, is still in queue. that one is indoors and i expect it to be rejected as well. the shop moved a large artwork from beside the cash register to the other side of the store and set up a seating area around it. neither move crosses cell lines, and isn’t a huge correction, but is a correction.

1 Like

I know what spam is. This was spam.

Submitting three good photos with subtle differences in crop and zoom for a Wayspot to see which looks best in different games in different contexts is not spam. I have to assume it was for Canterbury Entrance because one of those photos was just rejected before the warning.

idk what number constitutes spam :woman_shrugging:

3 Likes

I would be very upset if I got even one such email. Even knowing (as an IT professional who has worked on ML/AI) that it is likely the result of Machine Learning (AI training) gone bad.

This challenge is a Machine Learning exercise. It’s to (1) get volumes of data to manually review, and use as input to the ML, to train/feed the AI. And (2) let the AI decide volumes of data by itself, and see how it does (what still needs tweaking).

One of the things Niantic wants their AI to learn is how to detect abuse. From How Niantic Learns from Machine Learning (and AI) – Niantic Labs (11 months ago)

It’s hard to keep pace with the fluid state of the world, so we employ ML to help. Specifically, we use ML models to help:

  • Identify low quality wayspot nominations or edits to our map (e.g. blurry or violative photos, poor/inaccurate descriptions, incorrect locations)
  • Flag duplicative wayspots
  • Detect abusive behavior for our team to review and investigate

To tackle all of these tasks, we developed a collection of multi-modal deep learning models capable of synthesizing data from a myriad of sources.

It’s likely the developers had no clue that such testing would upset and alienate their human volunteer contributors.

2 Likes

Please dont spread misinformation as fact.

1 Like

I am not sure whether to just go ahead and withdraw anything that is in queue to be safe. I started to, but then would that look like I had done something wrong? Any advice on that? I think every photo ad and text edit I have submitted adds to the Wayspot, or I wouldn’t have submitted it.

I did accidentally submit a duplicate at the beginning (I have mentioned this on the forum before) where I forgot to remove that photo from my camera roll, but that has been the only one I think. It was marked “Not accepted” and I wasn’t worried about it, since mistakes do happen, and that is how the system is supposed to work.

1 Like

Wait for a response.

5 Likes

Don’t withdraw anything. It’s a shame you can’t put your edits on hold until a response is given either way. I have a feeling the verdict will go your way.

2 Likes

Can’t really imagine a world where @cyndiepooh doesn’t get an apology and a retraction of this disaster

I reallly hope staff come to comment on this and can sort things out today

I was worried this kind of thing would happen (was actually thinking it would be location edit related for the most part) but wow, did not expect it to target one of our most knowledgable and careful members of the forum!

5 Likes

oh i won’t get an apology. hoping for a retraction. will settle for a definition.