03/15 - This. Is. THURSDAY!

I bet that these have zero throttle considering that most people can't do them.

 

But in that example, the object that was being missed wasn't a baseball.

 

I would go with returning my queue. These are not for me, I am a sleepy polar bear already, no reason to make it more difficult.

 

Oops on the first pin, eh? That's fun.

 

Yep, no problemo just accepting next HIT.

 

@Fit Friends Yeah, that's right, getting my swole on.

 

Title: CEO Congressional Testimony Survey | PANDA Requester: Abbie Oliver [A26PGOQLW79A7P] (Req TV): N/A (TO): [Pay: N/A] [Fair: N/A] [Comm: N/A] [Fast: N/A] Description: ~ 20 minute survey about CEO actions Time: 1 hour(s) HITs Available: 1 Reward: $1.60 Qualifications: Financial Asset Owned - Common Share Stocks EqualTo 1; Total approved HITs GreaterThan 1000; HIT approval rate (%) GreaterThan 96; Location EqualTo US;

 

I missed that part so I guess the argument could be made for either one but I probably would have stuck with yes since the child was "playing" baseball

 

Title: A survey for parents with a kid age 2-15 | PANDA Requester: Common Sense Media [A2ACLL935OVRO8] (Req TV): N/A (TO): [Pay: 3.25] [Fair: 2.71] [Comm: 2.33] [Fast: 3.75] Description: a question asking how you interpret an image Time: 5 minutes(s) HITs Available: 1 Reward: $0.18 Qualifications: HIT approval rate (%) GreaterThan 97; exclude_from_all DoesNotExist ; exclude_from_age_stage_feature_name DoesNotExist ; Location EqualTo US;

 

That is a valid reason to use it, but I rarely have ever used the orange button. Maybe 20 times in the entire amount of Wills / Brian Zs I've ever done. I'm assuming they are trying to train something to categorize an action that creates sound, without sound, unless they wouldn't include it in the list. I don't think they are treating it like attention check.

 

I think the ambiguity actually helps the machine learn. If 20 people out of 100 people say it lasts the whole video - then the machine may think - ok, well maybe it kinda looks like this, but not completely. Then a different section of the same video has 80 out of 100 people saying it lasts the whole video - then the machine can learn - ok this looks even more like whatever action - awesome I learned something I'm a robot beep boop

 

Title: Call for participants | PANDA Requester: Human Cognition Lab [A3NWU07H510EH6] (Req TV): N/A (TO): [Pay: N/A] [Fair: N/A] [Comm: N/A] [Fast: N/A] Description: The Human Cognition Lab at New York University is looking for participants for a brief psychology experiment. You must be at least 18 years old and a fluent English speaker. The task will take about 5 minutes & will pay $1.. Time: 1 hour(s) HITs Available: 3 Reward: $2.00 Qualifications: Location EqualTo US;

 

soon buzby soon.

 

lefty realizes she's never going for that walk

 

This is pretty much how it works. You can't teach a machine to act more like a person without the nuance that you get from multiple interpretations. If they wanted a binary approach they'd just use the machines that already exist. We can already make those robots because there is no learning involved.

 

Title: Read some information about a company and answer some questions | PANDA Requester: Suneal Bedi [AGIWER494EXCC] (Req TV): N/A (TO): [Pay: 5.00] [Fair: 5.00] [Comm: 0.00] [Fast: 4.00] Description: Read some information about a company and answer some questions, about 5 minutes Time: 1 hour(s) HITs Available: 1 Reward: $0.40 Qualifications: Total approved HITs GreaterThanOrEqualTo 100; Has Taken EB1 Survey DoesNotExist ; HIT approval rate (%) GreaterThan 95; Location EqualTo US;