AI Weirdness: the strange side of machine learning
When data is messy

When data is messy

There’s a story I tell in my book [https://youlooklikeathing.com] because it’s a great illustration of how AI gets the wrong idea about what problem we’re asking it to solve: Researchers at the University of Tuebingen trained a neural net to recognize images [https://medium.com/
Depixellation? Or hallucination?

Depixellation? Or hallucination?

There’s an application for neural nets called “photo upsampling” which is designed to turn a very low-resolution photo into a higher-res one. This is an image from a recent paper demonstrating one of these algorithms, called “PULSE: Self-Supervised Photo Upsampling via Latent Space Exploration of Generative Models [https://github.
All your questions answered

All your questions answered

I’m previewing OpenAI’s new API, [https://beta.openai.com/] and like GPT-2, it looked at a lot of internet text during training. In my last post [https://aiweirdness.com/post/620645957819875328/this-is-the-openai-api-it-makes-spookily-good] I showed how it can adapt to different prompts in part because of how much it’
Rhyming is hard

Rhyming is hard

Although many people have generated AI [https://twitter.com/ConceptNetPoet] poetry [https://www.centerforthehumanities.org/james-gallery/exhibitions/house-of-dust] and lyrics [https://qz.com/920091/a-west-virginia-teen-taught-himself-how-to-build-a-rapping-ai-using-kanye-west-lyrics/] , you’ll notice that they generally don’t rhyme. That’s because generating a decent rhyme is super hard. You can get an inkling of
You've successfully subscribed to AI Weirdness
Great! Next, complete checkout for full access to AI Weirdness
Welcome back! You've successfully signed in.
Unable to sign you in. Please try again.
Success! Your account is fully activated, you now have access to all content.
Error! Stripe checkout failed.
Success! Your billing info is updated.
Error! Billing info update failed.