Im stillwaiting on thinhs as far back as august, though the august one is one that gets rejected by people but accepted on appeal lol. The latestni think from when ml got turned off is about september/october time, when there was still sunlight at 6am lol
Im very curious to see if what people think ML auto sccepts, and what ML will auto accept moving forward match.
Im hoping trail markers still get auto accepted, cause quite often they get rejected by people unless theres a map
I am a retired form a state highway Dept as a highway engineer for 33.3 years.
I worked with little fake points representing the digital land in the 1990âs. I surveyed and the designed 3d roads and surfaces for measurement and payment to federal standards⊠And built roads, bridges, trails etc⊠And then I went over into CAD IT support for my state hwy Dept. So I like what Niantic is attempting to do. I also made utilities move their stuff off my highway sections because IT was boring and teaching hwy engineers was annoying.
The reason I like AI or machine learning is that a person with knowledge can ask the right questions and get some interesting results or Tangents.
My last duty was to investigate a Superfund site for public information and to answer technical questions for scary assâs DOJ lawyers.
But I didnât ask the lawyers questions, I wanted to retire on my terms.
Fortunately Google Gemini can process old files and make me never regret, retiring from those hwy idiots.
I am not convinced all of the changes are for the better. For example, I had this nomination
which ML allowed to be determined by the community.
When I resubmitted it, following MLâs reboot,
ML rejected it.
Given that parks / outdoor recreation areas without signs are supposed to be eligible:
my conclusion is that ML denying the community the opportunity to vote on the nomination is a backward step.
The thing with ML is it isnât fed the criteria and told to review to it, itâs fed how people review and tries to interpret that. While you changed the description, you used the same title & photo, so I think in this case it actually proves that ML is learning because now itâs rejecting based on how the community handled almost the same nomination.
I think youâll have to get a new photo and change more text if you want to get past ML, or else rely on appeals after ML rejects it.
I donât say that to knock your nomination or the particular candidate and I hope you can eventually have success with it.
I spent forever getting pocket parks accepted. And I only got them accepted with explicit proof to council maps, LGA documentation to prove their existence.
Not surprised Emily struggles as the machine is learning from us. And I can say that humans had epic fails on parks - as too many people believed a âparkâ equals a certain size with x facilities - so rejected - not a local space used by community. And with no objects to hang the nomination - it is a hard ask to get nominated.
Emily learns from us. So I put the blame there to start with
@Liverlouike15 evidence evidence evidence. Because your local council does not list it. Even though it exists and is maintained. I would find every bit of evidence and stuff your supporting info. It is one part of the walking trail from 12 Acre Drive to Peachcroft Park. So i like it. I get its local community importance
Your dear daughter Emily has again rejected the âcorrect titleâ with an automatic rejection.
She submitted an edit report on 12/20/2024 and was rejected by âour teamâ on 12/23/2024.
I also submitted additional photos to show that the wayspot had disappeared, and she continued to approve them one by one.
Incidentally, the Wayfarer team has since seen my deletion report and pointed to the photo that Emily approved, and they continue to approve deletion reports.
So the problem is not solved.
It should be stopped immediately.
ML takes around 24 hours to make a decision. I believe this was decided by Niantic human staff; it most likely was listed as Under Review in Contribution Management, with the Niantic logo next to it. If itâs in this status for a few days, it was pulled from the regular queue to be reviewed by staff, not ML. âOur teamâ can either be ML or human staff, but the time frame for resolution is quicker with ML and slower with staff.
You also shouldnât be submitting new photos at a Wayspot that no longer exists as proof it should be removed. ML is learning from how we as a community review, so if thereâs a similar photo already approved at that Wayspot, ML will typically approve it.
If, as you say, it is not machine learning, but rather the judgment of the team staff, then it is even worse.
They are not equipped with the academic knowledge to judge our data.
Iâve had plenty of ML accepts well within 24 hours. Shortest Iâve seen was 3 hours or so.
They are thinking that ML took 3 days to decide, as they submitted on December 20th, but the decision wasnât made until December 23rd, and the email said âour team.â This sounds more like staff review than ML, as yes, it usually takes ML no more than 24 hours to make a decision.
âOur teamâ emails can either be for a ML or human staff decision; they do not say which made the decision in the emails. The way to know though is typically by the timeframe of the decision. 24 hours or less, itâs usually ML. A few days, itâs human staff.
Time to come back eMily
I too am very curious about this, I anticipate there will some pretty significant changes to eMiLy to detect possible nefarious actions by nefarious actors. Iâm hoping the speed element comes back. Like Iâve said before, those 22 hour approvals were amazing.
Do you know how long the grading error has been going on?
Hello and welcome,
Most started to notice an issue around the end of November, around November 27th and 28th. Reports started to pop up on social media around then, and a little later here in the forums. Around December 4th, people were noticing that ML was auto-rejecting again, but not auto-accepting.
The main thread of discussion from when it started is here, if youâre interested.
If you mean how long has the ML model been accepting incorrectly, that is unknown. I know all my nominations since March of this year have been decided by ML or in house review, with none of them going to community review before this pause. I would suspect that ML errors would have been happening since around this time, with people really starting to abuse it around this past summer. But that is just a guess from what I saw happening around me.
I have gotten so frustrated with the whole system that Iâve completely stopped playing Pokemon Go for several months now. Itâs zapped the joy out of playing. I donât need anymore advice as to âhow to get my nominations acceptedâ. This shouldnât be that challenging. Iâve had difficulties just updating murals that changed for Pokestops. When itâs more work than fun, I no longer invest my time or money into it. Iâm here now because Iâm hoping voices are being heard and this will improve. Until then, Iâll occasionally check back to see if any improvements have been made to be more user friendly for the regular player. Hoping this is going in the right direction.
I think this mindset is what we try to change here to âwhat can I submit that will be accepted?â We encourage people to look for great places to exercise, explore, or be social. And to take great photos that clearly show the point of interest.
I agree that Wayfarer is not for everyone. But to follow up on this:
What changes do you feel would accomplish this for you? This is a great place to make suggestions, and staff and ambassadors do follow these topics.
Hmm⊠maybe this is getting off topic from âML Auto-Decision Issueâ and we should start a new conversation?
Some of the posts are becoming more or less why I like ML/why I donât like ML, but there are some that are just not likely Wayfarer in general. The post you replied to is more Wayfarer in general, not so much ML related.
@boozillamom Itâs also sad to me that even if you are having issues with Wayfarer that you would stop playing any of the Niantic games. Even if I stopped doing Wayfarer today, Iâd still play PoGo, as I love it!
Just looking at your activity, it seems that you have been upset with the review community rejecting your nominations, and the appeals team rejecting them as well, and many of the others here, including the ambassadors, have provided a lot of helpful info. Everyone is trying to help you learn to be a better Wayfarer, but you seem to brush off the assistance.
We should all strive to learn new things everyday, and learning from others who may have more knowledge, like the ambassadors, can be extremely helpful. But, itâs your choice whether you want to take what has been given to you and apply it; we canât force you to do anything.