For instance, here’s the first thing the AI generated when I asked it to show me an example of the “ Is This a Pigeon?” meme: So meta it hurts. The results are often hilarious, and they certainly feel like real memes. Neural networks have made great strides in recent years, giving us everything from fake movies to fake articles and, of course, fake porn.īasically, Imgflip’s neural network processed a lot of memes and then tried to predict what memes should look like. The computer gorges itself on a lot of data and then teaches itself, through lots of repetition, how to predict what that data should look like. ![]() If you’re not familiar with a neural network, it’s basically a computer that uses an algorithmic technique known as deep learning. No, this generator uses its data about memes that do exist to conjure up memes that don’t exist. The tool is not to be confused with other meme generators, which merely curate existing popular memes. You can let the network generate a random meme for you, or you can preselect your meme from one of many popular templates, from Mocking SpongeBob to the Gatsby toast. But these memes aren’t actually “real.” They’re being created on the spot by a neural network, an artificial intelligence (AI) that predicts what it thinks a meme might look like. For proof, look no further than “ This Meme Does Not Exist,” a meme-creation tool created by meme-template website Imgflip.Īt a glance, it looks like your average random assortment of meme templates. The Dall-E generator did great in showcasing what this would look like - the cows actually look like they’re on the field in this one.It’s official: The coronavirus quarantine may have well and truly made the concept of a “meme” obsolete. Spectators just may feel less inclined to eat a hamburger while watching the next game. Pigeons always seem to wander onto the field during football games - how fun would it be to look out onto the field and see a bunch of cows roaming? They would definitely make sure the grass was cut so our team could score, and they’d distract the other teams with their adorable nature. Let’s hope he’s out by the next football season, because we really need his cheer! The AI had an easier time making the panther’s face as well as clearly showing the bars he’s behind, so it definitely had a much better execution than the images above, earning it the number two spot overall. While we can’t imagine what crimes Roc must have committed to have landed in jail, the look of anguish that the Dall-E website generated on the poor panther’s face shows that it must have been pretty bad. What do Kanye West’s tour staging and the Cathedral of Learning have in common? They are both extra - but in a great way! Cathy would make the perfect setting for Kanye’s next tour - it already looks like a church, which would make it the best place for Ye to perform his album “Jesus is King.” The AI also did a bit better in making this image - we can see both Cathy and Kanye more clearly, albeit a bit fuzzy on Ye’s face. Kanye West in the Cathedral of Learning. ![]() ![]() Great job on the idea, less so on the execution. He sort of looks like a blue and yellow blob in a vaguely Rite Aid background. While it would be really cool to see our very own Kenny Pickett browsing the aisles of Rite Aid for snacks or shampoo, the Dall-E AI didn’t do so well at making out the shapes of the store or of Pickett. Here are our rankings for the best ones with a Pitt theme. on Instagram created some Pitt-themed memes using Dall-E with some outlandish scenarios bound to make Pitt students laugh. It was only a matter of time before Pitt students got into the mix. Truly, you can make whatever your heart desires with this AI. Some examples include Rocky Balboa fighting the Michelin Man and Gollum playing tennis. Dall-E, the AI trained to create images based on text, has taken the internet by storm this week, with new accounts on Twitter and Reddit popping up to post about the weirdest combinations that can be made using the app.
0 Comments
Leave a Reply. |