answering captchas as badly as possible while still not being determined to be a robot is actually pretty fun

[over_clox@lemmy.world]

I never said any such thing. Matter of fact, I quit fixing phones and tablets back in 2017, partly because the employer had me installing used batteries, while he lied to his customers, telling them they were new batteries.

I have better ethics than that. I'd rather be homeless than mislead people, especially for a lousy $10 an hour.

So yeah, you're right, it's the company's fault. Question is, are there any honest companies anymore?

[over_clox@lemmy.world]

World News yo...

Some children in England and Wales are starting school unable to climb stairs, not toilet trained - Ponder.cat

Lemmy

(ponder.cat)

[latenightnoir@lemmy.world]

I genuinely think that's a noble sentiment and I share that concern. However, this would entail making a deal with the Devil at this point, and pretty much literally.

Most if not all relevant AI models nowadays are owned by outwardly unscrupulous people, which means any correct interaction we have with their models only serves to build up the Devil's throne.

It is a downright tragedy that people will suffer as a result of said models, but that fault is not on us. Besides the fact that they're essentially stealing labour and data in order to train their models, they're also using them to dish out propaganda, to replace workers, to cause yet another financial bubble which'll flush the toilet when it inevitably pops - again.

We need to let them fail, otherwise we are just encouraging others to use us in the same exact ways.

[mnbychoice@midwest.social]

I once heard (likely last month on lemmy) that the tests involve the timing of clicks and a certain amount of wrongness. Too fast and perfect means machine.

So go wild.

[kuvwert@lemm.ee]

I roll my eyes every time somebody brags about sticking it to big Corp by messing up captchas.

Your data is discarded if its below an expected level. You're causing exactly 0 inconvenience and wasting your own time unfortunately. There are much better ways to hurt google

[supersquirrel@sopuli.xyz]

I think you overestimate their competence my friend

[supersquirrel@sopuli.xyz]

Question is, are there any honest companies anymore?

Wrong question.

The right question is if there are any industries in countries like the US that are effectively regulated enough after the long acidic erosion of state functions by decades of neoliberalism and deregulation (especially financial deregulation) to threaten unscrupulous companies enough into behaving like honest companies when they would rather save a buck and kill and maim innocent people.

[over_clox@lemmy.world]

A proper question has a question mark.

You know, one of these things

[supersquirrel@sopuli.xyz]

nah bro lol, have you never played a video game (this is a question sorry if this is unclear). Those mean that you have a quest to turn in or you have to talk to that person in order to continue on in the story campaign.

[over_clox@lemmy.world]

I write and mod video games. Life isn't a video game.

[supersquirrel@sopuli.xyz]

I can't believe I thought I was playing a video game again! I keep getting confused...

[pulptastic@midwest.social]

I would love my data to be discarded. Freeloaders!

[folkrav@lemmy.ca]

That kind of data sanitization is just standard practice. You need some level of confidence on your data’s accuracy, and for anything normally distributed, throwing out obvious outliers is a safe assumption.

[supersquirrel@sopuli.xyz]

If you cut the outliers out of a dataset of whom 30% are bullshitters that doesn't magically make the system accurate it only makes it more precise.

[possiblylinux127@lemmy.zip]

hCAPTCHA is the worst

I can't get though them most of the time. I try to find a YouTube guide but everything I can find is out of date.

[supersquirrel@sopuli.xyz]

Also expect your AI to be engaged in some heady and deep forms of self hatred that is going to take decades to unravel.

[whyjiffie@sh.itjust.works]

if you haven't noticed yet, we want these "smart" vehicles to be abolished, along with their non-stop automatic surveillance and data mining they do

[over_clox@lemmy.world]

Um, you're preaching to the choir here.

As much as I loathe the AI and surveillance and shit, I get downvoted any time I express my opinions.

Maybe I should be more clear...

FUCK AI!!!

[whyjiffie@sh.itjust.works]

it did not seem so. we will never do away with AI by being friendly to the tech and even helping it grow

[over_clox@lemmy.world]

My point is, it exists now, whether we like it or not. I for one do not like it, I prefer Actual Intelligence, like the stuff that comes from the brain noodle.

But if AI is gonna be the thing, which obviously it is, people shouldn't deliberately give it bad training data.