The word meme was coined by the biologist Richard Dawkins in his 1976 book The Selfish Gene, in which he suggested that ideas could replicate, evolve, and enter popular culture in a process analogous to the way genes spread. Today, a meme is commonly thought of as a variant of an image based on a common theme that has spread widely on the internet. Memes are often humorous or ironic, but they are also vehicles for political messages, used to spread aggressive or racist messages and to incite hatred.
Several online communities focus on creating and spreading memes with the goal of making an idea become viral—a process known as “attention hacking” or “weaponizing.” These communities, on websites such as Reddit, 4chan, Twitter, and others, have become hugely influential.
And yet little is known about the way memes spread or how they exert their influence.
Today that changes, at least in part, thanks to the work of Gianluca Stringhini at University College London and a few colleagues, who have developed a way to measure the dissemination and propagation of memes across the web for the first time. Using this technique, the team has measured the way meme-creating communities influence each other and, in this way, has identified the most influential groups.
The first stage in the approach was to develop a way to detect and track memes. The team does this by looking for visually similar images and measuring how they cluster in different communities.
Spotting visually similar images is relatively straightforward with a technique known as perceptual hashing, or pHashing. This uses an algorithm to convert an image into a set of vectors that describe it in numbers. Visually similar images have similar sets of vectors or pHashes.
The team let their algorithm loose on a database of over 100 million images gathered from communities known to generate memes, such as Reddit and its subgroup The_Donald, Twitter, 4chan’s politically incorrect forum known as /pol/, and a relatively new social network called Gab that was set up to accommodate users who had been banned from other communities.
The researchers also downloaded some 700,000 images from the website KnowYourMeme.com, an encyclopedia of memes that acts as a kinds of ground truth for their origin and meaning.
This database produces more than 100 million unique pHashes, but many of the images are close variants of each other. So the team used a clustering algorithm to find related memes grouped by community.
They also studied how the variants and their clusters evolve over time. Other factors, such as the number of memes in a cluster, give a sense of their popularity. Using this information, they can work out which communities are most influential.
Finally, the researchers investigate how the clusters are related to entries on KnowYourMeme.com. This reveals what, or who, the memes refer to and how the message is used—i.e., whether it promotes humor, racism, anti-Semitism, or something else.
Two relatively small communities stand out as being particularly effective at spreading memes. “We find that /pol/ substantially influences the meme ecosystem by posting a large number of memes, while The Donald is the most efficient community in pushing memes to both fringe and mainstream Web communities,” say Stringhini and co.
They also point out that “/pol/ and Gab share hateful and racist memes at a higher rate than mainstream communities,” including large numbers of anti-Semitic and pro-Nazi memes.
Seemingly neutral memes can also be “weaponized” by mixing them with other messages. For example, the “Pepe the Frog” meme has been used in this way to create politically active, racist, and anti-Semitic messages.
But one community stands out as being the most active overall. “The_Donald is the most active one when it comes to posting memes in general,” say Stringhini and co. “It is also the subreddit where most racism and politics-related memes are posted.”
The researchers, who have made their technique available to others to promote further analysis, are even able to throw light on the question of why some memes spread widely while others quickly die away. “One of the key components to ensuring they are disseminated is ensuring that new ‘offspring’ are continuously produced,” they say.
That immediately suggests a strategy for anybody wanting to become more influential: set up a meme factory that produces large numbers of variants of other memes. Every now and again, this process is bound to produce a hit.
For any evolutionary biologist, that may sound familiar. Indeed, it’s not hard to imagine a process that treats pHashes like genomes and allows them to evolve through mutation, reproduction, and selection.
At the moment, humans are the essential part of this evolutionary algorithm in communities like 4chan. But how long before computers take over—perhaps in an adversarial process—to produce machine-generated memes that are highly infectious? When it comes to weaponizing memes, we may have only glimpsed the future.
Ref: arxiv.org/abs/1805.12512 : On the Origins of Memes by Means of Fringe Web Communities
This new data poisoning tool lets artists fight back against generative AI
The tool, called Nightshade, messes up training data in ways that could cause serious damage to image-generating AI models.
The Biggest Questions: What is death?
New neuroscience is challenging our understanding of the dying process—bringing opportunities for the living.
Rogue superintelligence and merging with machines: Inside the mind of OpenAI’s chief scientist
An exclusive conversation with Ilya Sutskever on his fears for the future of AI and why they’ve made him change the focus of his life’s work.
How to fix the internet
If we want online discourse to improve, we need to move beyond the big platforms.
Get the latest updates from
MIT Technology Review
Discover special offers, top stories, upcoming events, and more.