08/21 - Dirty Dancing Wednesday

Sorry for multiple posts, but I did a 25 cent HIT just to see what would happen and it was hard to get to go through. I submitted, got an error, went back to the queue, hit Work again, got told that I had a problem submitting, and hit the new submit button and it slowly went through. I wouldn't try any long or high stakes HITs for now.

 

So much for that week of it working as expected. I'm back on night owl hours to do anything.

At this point it *has* to be them not allocating enough resources to the platform, right?

 

don't stress everyone. this person can tell us all what to do.

 

Fake, she's getting paid by John Mtruk to shill PickFu HITs

 

And the saddest thing is: we're talking about AWS here - the world's biggest cloud provider. They power something like a third of all popular websites / services on the web. But their own crowdsource platform? Nope, that's asking too much - it has to be set up in the jankiest possible, most penny-pinching way to self-throttle to avoid accidentally going over budget. It's the tragedy of mega-corporations that allow fun little side projects like MTurk - nice while it lasts but can be abandoned at any moment because no one up high actually cares. MTurk is at the stage of "just leave it there plugged in as long as we don't lose money".

 

what is the link to contact a requester? without having submitted a hit. this time I got the header or cookie too large error after lag after submitting a hit. the hit timed out.

 

https://worker.mturk.com/requesters/A1LZRIR5OYJ4E9/projects

this is the place I need to contact. but I couldn't submit the hit in time and they don't have any reviews on turkerview.

edit: I found a link somewhere that says it works to email requesters by googling. it has instructions with it if anyone else wants to find it which some people will today if more hits mess up.

 

what hits are you guys doing? I'm hamming a meh batch.

 

Prolific has been terrible lately, constant 429 too many requests errors.
What could be the problem?

Name: PROLIFIC.COM
Nameservers:
NS-1424.AWSDNS-50.ORG
NS-1833.AWSDNS-37.CO.UK
NS-226.AWSDNS-28.COM
NS-736.AWSDNS-28.NET

 

Could be AI driven. The more resources they put into AI, the less they have for everything else. A single GPT query is something like 1000% more resource intensive than an old Google Search.

 

That uses mostly GPU, so very different resources. But I agree, most of the tech brain power is currently dedicated to solving humanity's most pressing problem:

 

Probably should've hidden that behind two spoilers and a biohazard sign, now that I think about it.

 

I'd imagine they keep moving our server rack(s) further and further into the basement, slowly unplugging LAN cables and giving them to the other racks. Eventually, we'll just have to burn the place down.

 

More AI horror (god help us):

Spoiler

 

why are we even doing this what is the point? I feel like it's because modern rich people are so dumb that they need us to build them brains. am I wrong?