What is this? Spoiler I am Tjbot! I'm a solution devised by the genious mods to prevent there ever being a day that hit threads aren't posted! The whole point of my existence is to post a hit thread if one doesn't exist by 12:08am PST. If you see this post, assume that this one is the main post. One of the mod team will clean up existing posts and consolidate once they wake up and get online. As my name indicates, I am a bot. I don't post, nor do I respond to PM's. If you have any problems, send an email to tjbot@ns4t.net or PM Tjololo, he can help out with any issues. Want to make tomorrow's thread? Just go for it!
2 min, watch the short timer. Title: Answer a survey about people's social behaviour. | Accept Requester: Evangelie Papadimitriou [A29BNMGNNJWIMV] Contact TV: No Reviews TO: No Reviews TO2: No Reviews Reward: 0.50 Duration: 8:00 Available: 1 Description: In this study, you will read information about a well-known company and answer some questions. Qualifications: Masters Exists ; HIT approval rate (%) GreaterThan 85; Location EqualTo US
15 min + bonus Title: Flood risk | Accept Requester: ETH DeSciL [A13FVLMLUTH7ND] Contact TV: [Hrly=$23.26] [Pay=Generous] [Approval=~24 hrs] [Comm=Good] [Rej=0] [Blk=0] TO: [Pay=4.49] [Fast=4.91] [Comm=4.64] [Fair=4.93] [Reviews=394] [ToS=1] TO2: No Reviews Reward: 3.00 Duration: 1:00:00 Available: 1 Description: The purpose of the survey is to understand how members of the public think about public programs to address flood risks. Qualifications: Total approved HITs GreaterThanOrEqualTo 50; Descil-E-E9858EA7 DoesNotExist ; HIT approval rate (%) GreaterThanOrEqualTo 98; Location In US
A different one from their previous versions bonus potential Title: speaking experiment with bonus! (dlgr-24a19e3f-06f9-b73a) | PANDA Requester: Computational Audition [AIOOK9HZ16SLH] (Contact) TV: No Reviews TO: No Reviews TO2: No Reviews Reward: $0.10 Duration: 2 hours Available: 1 Description: This is speaking experiment that needs to be done in a quiet place WITHOUT headphones. You will be asked to imitate sentences by replacing the syllables with "PA". The task will take about 10 minutes. Requirements: Total approved HITs GreaterThanOrEqualTo 5000; HIT approval rate (%) GreaterThanOrEqualTo 99; Location EqualTo US
These are a bit of a grind, but they're better than nothing Title: Emotion Prediction(Classification) | PANDA Requester: Interactive Multimodal Learning [A2BIK0SKCW1ZAG] (Contact) TV: No Reviews TO: No Reviews TO2: No Reviews Reward: $0.10 Duration: 30 days Available: 5577 Description: In this task, we are interested in the emotion of the character in the story. Given a story with 5 sentences, please read the 5 sentences carefully and choose the most relevant emotion. We have prepared a list of candidate emotions, please choose the one Requirements: Total approved HITs GreaterThan 1000; Masters Exists; Location In AU, CA, GB, US; HIT approval rate (%) GreaterThan 95
Yesterday @Flora brought up the issue of how the platform and requesters can deal with scammers especially batch scammers which I think happens very very little (as @ohsostrange mentioned) especially batches where requesters set parameters like Masters or over 10,000 HITs done or something like that. I can't conceve that anyone could get away with doing it any significant amount of time. There's no reason Amazon can't deal with this. They could simply eat the loss if they wanted to be lazy and they wouldn't even notice it would be such a small amount. Not that I want to reward scammers but I also have no sympathy for Amazon and how they don't really take MTurk seriously in the first place. But if they wanted to combat it they could have things set up like well if a requester is mass rejecting several different workers then it's probably the requester. If he/she is only mass rejecting a few people then there should be some kind of review of the worker and some kind of arbitration which there should just be a way for workers to push appeal a rejection through MTurk and not have to send an email to the requester and hope they see it and take action. How long has MTurk been around and there's still so many unaddressed issues most of which affect the workers and not the requesters.
So many unaddressed issues... Sooooooooooooo fucking many. Like the requesters who reject you to cherry pick their results and skew the findings of their "research," like that's how science works. And the ones who swear your rejection(s) are a "mistake" they have already reversed, and expect you not to notice they're lying, like can't see them still sitting there on your dashboard. And the requesters who run you through twenty minutes of bubbles and then sock you with a writing task they didn't tell you was coming, daring you to toss it back and throw away your money after you have that much time invested. And the ones with non-functional completion code generators at the end who then reject you for not having a completion code. ETA: and the ones whose directions read like they were percolated through every language in Google translate before they landed on English. I love those. And the "five minute" ones with twelve pages of instructions. And the instructions that contradict themselves to the point they might as well have just used an Escher drawing instead of text. Hell. Etc. ad infinitum... ETA: tempted to keep going, but I'm starting to remind myself of my grumpy little shih tzu
Just to complete my rant... No petitions will ever convince Amazon to fix any of this. Why would Mr. "Blasting off in my giant space penis for shits and giggles!" do anything to help us? While in every rational sense. we are Amazon employees, legally we're not, so there's no reason he has to. Plus, what we generate for the company is pocket change next to the billions generated by the warehouse workers who very much ARE legal employees, yet still haven't been granted breaks long enough to walk to the bathroom and back. Think about that. We are lower on the company power structure than people they're forcing to pee in bottles, and then carry that pee around with them as they run from one end of the warehouse to the other. It's possible the cockroaches in those warehouses have more status than we do. Given all that, I see no reason to hope any more improvements of any kind are coming, ever, no matter what we do.
I always worry that if we stir shit up some executive or Bezos himself will say "What the fuck is Mechanical Turk? I thought we got rid of that 15 years ago" and then shit can it.