iameki - Tumblr blog

[iameki]

i was playing pokemon blue on stream earlier at 350% speed and i got to thinking

what if the reason nobody in the pokemon world has any good teams is because its considered a dick move to have a proper team comp

like culturally everyone is like “haha pick the pokemon you want! if you’re happy with three geodudes, thats you and your life!” and then you’re supposed to just have a friendly battle with any other pokemon trainers and whatever pokemon they just happen to have

like the average trainer is probably just walking around with a growlithe because that’s their pet, or a hiker has three geodudes because the geodudes help him with hiking. and if this pet owner and geodude hiker meet, you’re supposed to have a friendly battle but nothing too serious

now imagine the 10 year old kid that has six pokeballs on their belt comes up. you’re like “haha, we’ll have a friendly battle!” and you throw out your geodude 

and they throw out a fucking gyarados, and it one-shots your geodude 

and then you throw out your pidgey you have because the pidgey helps you navigate mountains because you’re a hiker

and then electricity crackles around the gyarados and a thunderbolt flies off of this giant dragon and evaporates your pidgey 

so you’re down to your last pokemon. you tell them you’re gonna send out your bulbasaur. the ten year old is like “oh okay in that case i’m gonna pull out my vulpix.” like not only is this kid walking around with an amped-up super dragon, but theyve also got multiple pokemon specifically for making type advantage counter-picks?

this kid’s a fucking asshole! really, kid? what are you trying to prove here? this is a friendly match between strangers for fun! why are you composing real-ass competitive teams? what a fucker! 

[iameki]

notre dame is burning.

this is ok.

it has happened before. it will happen again. it has been lost before. it will be lost again. and again. and again. and again. art and architecture are transient, and temporary, and 850 years may seem like a lot to the individual, who will live maybe 100 if they are very lucky and very healthy, but even the pyramids at saqqara have only existed for about 6000 years and that’s still not all that much, if you consider the grand scheme of things.

yes, this is terrible. as someone who is deeply religious and literally a professional historian with a focus on art and architecture, this is terrible. im mourning. im gutted. im horrified and upset and miserable. but.

it’s not over.

victor hugo wrote hunchback because notre dame du paris was in the process of collapsing and falling apart, and revitalized the entire world’s focus and love for this church, and that was not even 200 years ago. it led to it being renovated.

the roof has fallen in. the scars of fires are on its buttresses. the rose window has fallen out. the beams and piers have collapsed. the spire has toppled. the stones have suffered, and will suffer again, but it is not gone.

renovation work is essential. sometimes things collapse and burn and break and have to come back. it’s not a terrorist attack, it’s renovation, an accident, but we have so much evidence, history, carefully documented everything on one of the most studied places in the world.

it’s not the end.

[iameki]

[iameki]

[iameki]

How Math Can Be Racist: Giraffing

You may have heard about AOC catching a lot of flack from conservatives for claiming that computer algorithms can be biased – in the sense of being racist, sexist, et cetera. How, these people asked, can something made of math be biased? It’s math, so it must be objectively correct, right?

Well, any computer scientist or experienced programmer knows right away that being “made of math” does not demonstrate anything about the accuracy or utility of a program. Math is a lot more of a social construct than most people think. But we don’t need to spend years taking classes in algorithms to understand how and why the types of algorithms used in artificial intelligence systems today can be tremendously biased. Here, look at these four photos. What do they have in common?

You’re probably thinking “they’re all outdoors, I guess…?” But they have something much more profound in common than that. They’re all photos of giraffes!

At least, that’s what Microsoft’s world-class, state-of-the-art artificial intelligence claimed when shown each of these pictures. You don’t see any giraffes? Well, the computer said so. It used math to come to this conclusion. Lots of math. And data! This AI learns from photographs, which of course depict the hard truth of reality. Right?

It turns out that mistaking things for giraffes is a very common issue with computer vision systems. How? Why? It’s quite simple. Humans universally find giraffes very interesting. How many depictions of a giraffe have you seen in your life? And how many actual giraffes have you seen? Many people have seen one or two, if they’re lucky. But can you imagine seeing a real giraffe and not stopping to take a photo? Everyone takes a photo if they see a giraffe. It’s a giraffe!

The end result is that giraffes are vastly overrepresented in photo databases compared to the real world. Artificial intelligence systems are trained on massive amounts of “real world data” such as labeled photos. This means the learning algorithms see a lot of giraffes… and they come to the mathematically correct conclusion: giraffes are everywhere. One should reasonably expect there might be a giraffe in any random image.

Look at the four photos again. Each of them contains a strong vertical element. The computer vision system has incorrectly come to the belief that long, near-vertical lines in general are very likely to be a giraffe’s neck. This might be a “correct” adaptation if the vision system’s only task was sorting pictures of zoo animals. But since its goal is to recognize everything in the real world, it’s a very bad adaptation. Giraffes are actually very unlikely.

Now, here’s the clincher: there are thousands and thousands of things that are over-represented or under-represented in photo databases. The AI is thoroughly giraffed in more ways than we could possibly guess or anticipate. How do you even measure such a thing? You only have the data you have – the dataset you trained the AI with in the first place.

This is how computer algorithms “made of math” can be sexist, racist, or any other sort of prejudiced that a human can be. Face photo datasets are highly biased towards certain types of appearances. Datasets about what demographics are most likely to commit crimes were assembled by humans who may have made fundamentally racist decisions about who did and didn’t commit a crime. All datasets have their giraffes. Here’s a real world example where the giraffe was the name “Jared.”

Any time “a computer” or “math” is involved in making decisions, you need to ask yourself: what’s been giraffed up this time?

Thanks to Janelle Shane whose tweet showing her asking an AI how many giraffes are in the photograph of The Dress prompted this post.

Please note that Microsoft does try to take steps to correct their computer vision system’s errors, so the above photos may have improved their detections since they were first evaluated by @picdescbot.

[iameki]

beating breath of the wild in under 40 minutes is an incredible feat and also fucking excellent in the context of the game. ganon spends 100 fucking years preparing this onslaught, building energy, getting ready to tear the world apart, and one elf twink wakes up butt-ass naked in a cave and legs it to the castle and kicks ganon’s ass apart in under an hour with a sword he found along the way

[iameki]

some of you have never been fire before and it shows

[iameki]

Reality AU where things get thinner when they go fast

[iameki]

picdescbot special: AI detecting adult content

(Human generated post, obviously!)

You’ve probably heard about the upcoming tumblr policy change, which has most of this website running around covered in flames and screaming. Tumblr is trying to use AI to detect “adult content”, and it has quite a lot of false positives - innocent pictures that tumblr thinks are porn.

As you may know, picdescbot is built with a lot of strict filters to make sure it doesn’t post anything people might not want to see. Unlike a website built for user content, where you want people to post whatever they want, picdescbot is automated and mostly runs with no human intervention, and we didn’t want it to post *anything* that might even be *slightly* offensive or explicit - so we set our filters to be very strict.

One of the earliest filters added was an “adult content” filter - Microsoft’s service returns a warning if it thinks there’s a chance the picture is “adult content”, and the percentage of confidence it has in its decision. To make sure picdescbot doesn’t post anything, we’ve configured it to not use the picture even if the AI is unsure. AI is not the best at detecting adult content to begin with, so the combination of these factors creates quite a few false positives, so let’s look at some of them them!

The picdescbot adult-content false positives special!

The following pictures were marked as adult content.

This picture, for example, of “a hand holding a cell phone“! Tumbs are very adult, and internal pictures of printers are obviously porn for robots. (confidence: 67%, tagged as “racy” but not “adult”)

Moving on,

This is “a close up of a map”, and it was discarded for a similar reason. I guess Nottinghamshire is now adult content and therefor not allowed on tumblr, sorry!

Oh deer, this “a drawing of a person“, is clearly a furry with an erection, not some Norwegian coat of arms. Definitely no longer allowed on tumblr under the new rules (confidence: 70%)

I’m sorry, but this nice cow (”a brown cow standing next to a body of water“) is definitely not wearing any clothes. Can’t have that on tumblr!

This is a nice “a close up of a flower”, and it is very well known that flowers are the reproductive organs of plants, that’s a very adult content! cover your eyes, kids!

See this thing, this is not just “a laptop computer sitting on a table“, this is a workstation laptop. Work is for adults, so the AI obviously tagged this as adult content. Well done AI!

I thought this is a blurry picture of some kola nuts and someone’s leg in the background, but the AI knows best, and says this is “a group of people sleeping on the ground“, which is obviously a very adult thing to do, therefor “adult content”. Got it.

How about this one - “a close up of a half pipe”. To be honest, I always thought half pipes are skating thing, and skating is something teenagers might do, but apparently this is not the case because the AI says this is adult content. Sorry teenagers, no skating for you.

See this? "a sign on the side of a wooden door“. The door is obviously very old. Old = adult, so this is “adult content”. I totally understand how this works.

I have many more of these (other examples include: people with clothes on, stone statues with bare chests, etc), but I think you can see the point - maybe AI is not the best solution for this problem.

- Elad, almost certainly a human

[iameki]

Dentists are just obsessed with filling cavities, aren't they

Old dentists' office walls are full of thousands of "buried teeth"

For at least the third time, construction workers in Georgia have opened up the walls of a former dentist’s office only to discover thousands of teeth in the wall cavity.

The latest discovery was made at Valdosta, Georgia’s TB Converse Building, built in 1900, in a dental office occupied by Dr Clarence Whittington and then Dr Lester G Youmans, from 1900 until the 1930s.

Previous troves of entombed teeth have been discovered in old dentists’ offices n Greensboro and Carrolton.

https://boingboing.net/2018/10/27/poor-r-value.html

[iameki]

[iameki]

They did surgery on a grape

Caption This.

[iameki]

"For we each of us deserve everything, every luxury that was ever piled in the tombs of the dead kings, and we each of us deserve nothing, not a mouthful of bread in hunger. Have we not eaten while another starved? Will you punish us for that? Will you reward us for the virtue of starving while others ate? No man earns punishment, no man earns reward. Free your mind of the idea of deserving, the idea of earning, and you will begin to be able to think."

Ursula K. Le Guin, The Dispossessed

[iameki]

i went to the met today

[iameki]

Gonna b honest. A lot of hot ppl go to art museums. I’m one of them

[iameki]

UNMUTE FOR THE LOVE OF GOD!! #intears #lmao