Is that a result of a skewed training set or are people really hard to tell apart from gorillas if there are no obvious tells like large difference in brightness of different areas of the face?
Deep learning, for all its recent glories, still suffers from relatively crude, slow-converging training algorithms compared to other areas of ML and statistics.
Maybe to your typical SGD-type algorithm, working off a dataset filled with mostly light skin toned people, skin tone just looks like a real solid first-order way to distinguish humans and primates, and picking up the black people / primate distinction seems much more marginal and second-order, in terms of impact on the cost function.
If most of the people in the dataset were black, I predict you wouldn't see this.
Consider too what they are likely using for inputs: photos with associated comments.
I don't know Facebook's TOS sufficiently to know whether they are using private groups as source material, but if you're utilizing bigoted content to train pattern recognition, you will replicate bigoted content.
My guess is that the poster was making an assumption that a large part of facebook's images are bigoted content. I am neither agreeing or disagreeing. But apparently some people got a little emotional about the platform being associated with maybe having a heightened amount of bigot content.
Not necessarily a large part, simply enough to identify as its own pattern.
In my experience there are a lot of bigoted things on Facebook. If these are serving as source data, and are sufficiently distinguished from other training material, it may well be user behavior the ML system would replicate.
I’m fairly certain if you showed pictures of both groups to a toddler they’d be able to sort them correctly. It’s really not hard for a human to tell the difference. Which tells me that FB’s AI isn’t really that great.
I saw another article which included a screenshot of the mistake on facebook. The photo is blurry, and contains a side view of the person and iirc the backdrop is in nature. The main indicator that it was a person was the fact they were wearing clothing. Clearly the AI is not good enough, but I will admit that the data it was working on is tricky to achieve perfect accuracy.
Human-like really depends on your interpretation. That's a generous reading of what's going on. If you google Gorilla faces, I don't think you would be confused.
The AI is not that smart and these examples show it.
>Us humans are super good at distinguishing faces.
It would be interesting to test a bunch of midwesterners at their ability to tell Asians apart or to be able to distinguish various Asian ethnicities. My guess is that a lot of the distinguishing features that they look for are altered or missing.
And while that is probably true for most of us and gorilla faces, even those midwesterners would easily distinguish an Asian person from a Gorilla.
It's true that we're good at recognizing faces (even where there are none), and distinguishing on a basic level (type of animal) but specific faces are mostly cultural.
"Human supremacist" attitudes are incredibly common. Look at any discussion of animal intelligence and you'll see the most vehement denials of any possibility that our cognition and emotions aren't unique in the world.
It reminds me of some of the explorers' tales of people who were half-human, half some other animal, or of people covered in hair, the first of which may have originated from seeing people riding animals, and the others to various (actual) primates. If humans can make such mistakes, certainly Facebook's AI can be excused for its confusion.
Are you sure those were honest mistakes and not stories for the sake of storytelling? And no, prehistoric ignorance does not justify this system making it to production.
There is evidence that CNN's use texture features more than shape features, i.e. have a texture bias. It's hard to tell in this case without access to the data/model, but it's very possible colour is being overvalued by the classifier and causes the errors.
The most obvious answer that no one wants to mention is that there just genuinely is some similarity between the two categories which is stronger than the similarity between others.
Maybe to your typical SGD-type algorithm, working off a dataset filled with mostly light skin toned people, skin tone just looks like a real solid first-order way to distinguish humans and primates, and picking up the black people / primate distinction seems much more marginal and second-order, in terms of impact on the cost function.
If most of the people in the dataset were black, I predict you wouldn't see this.