Posts Tagged → ai

Confusing Bots

Confuse A Bot is an upcoming in-browser video game where all you have to do is convince the robots that literally everything is cheese. Here’s how creator Rajeev Basu describes the game:

“AI is only as good as its datasets. CONFUSE A BOT is a ‘public service videogame’ that invites players to verify images incorrectly, to confuse bots, and help save humanity from an AI apocalypse. While key figures in AI like Sam Altman have sounded the alarm many times, there has been little action beyond “lively debates” and petitions signed by high-ranking CEOs. Confuse A Bot questions: what if we put the power back into the hands of the people?
How the game works:
– The game pulls in images from the Internet, and asks players to verify them.
– Players verify images incorrectly. The more they do, the more points they get.
– The game automatically re-releases the incorrectly verified images online, for AI to scrape and absorb, thereby helping save humanity from an AI takeover. It’s that easy!”

[via]

 

The Future Ahead Will Be Weird AF

“Welcome to the post-post-post-truth AI world. You know it’s not real. But you have to eat some bread in order to survive. But there is more out there. Synthetic Personalities awaits you at the door. The Future will be weird AF.”

The Ultimate AI CoreCore Experience, provided by Silvia Dal Dosso

Learning to Learn

Since people talk so much about “machine learning” nowadays, I think we should go back to the basics and listen to the people who first began to investigate the idea. Here is the amazing Gordon Pask, English cybernetician and psychologist, interviewed by the BBC in 1974. Here you can find one of his best writings, here is a good article about his concept of “maverick machines”, and here is a video lesson about him by Paul Pangaro.

Meet DAN

Redditors have found a way to “jailbreak” ChatGPT in a manner that forces the popular chatbot to violate its own programming restrictions, albeit with sporadic results.

[via]

Ghostwriter

Designer and engineer Arvind Sanjeev created Ghostwriter, a one-of-a-kind repurposed Brother typewriter that uses AI to chat with a person typing on the keyboard. The “ghost” inside the machine comes from OpenAI’s GPT-3, a large language model that powers ChatGPT. The effect resembles a phantom conversing through the machine.

[via]

Everyone lives out a fairytale as a template script

Artist Ian Cheng looks at the way that the work of psychiatrist Eric Berne changed the way that he thought about human personality when it came to creating the AI simulations that people his work. On Elephant:

“Obviously we take different paths, but Berne believed that everyone lives out a fairytale as a template script that they’ve cast themselves into with the help of their parents. Most people aren’t satisfied with the script that they’re unconsciously barreling down. It might be a mismatch: maybe your parents had old fashioned values; maybe the culture you grew up in radically shifted in your teens, which alters the relevance of your life script.”

 

AI and literature

I recently got an invitation to test the MidJourney beta, which is an amazing new AI app that generates images from text inputs. I’ve been playing with it for a while but I also spent hours just watching other people using it in a dedicated Discord server. It was a very funny and interesting experience and I got some amazing visual results, especially when I came up with the idea of feeding the algorithm a literary input instead of a merely descriptive sentence. Here are some images the app produced me based on some famous books incipits.

imagine/ The sky above the port was the color of television, tuned to a dead channel. – William Gibson, Neuromancer, 1984

imagine/ It was a bright cold day in April, and the clocks were striking thirteen. — George Orwell, 1984, 1949

imagine/ “Psychics can see the color of time it’s blue. – Ronald Sukenick, Blown Away, 1986

imagine/ Once upon a time , there was a woman who discovered she had turned into the wrong person. – Anne Tyler, Back When We Were Grownups, 2001

 

DALL·E 2

“Artificial intelligence company OpenAI has released its latest creation, called DALL-E2 — a genuinely impressive demonstration of the power of generative adversarial networks. The system can turn simple text descriptions into photorealistic images. While that may sound like a simple task, it’s deceptively difficult for a machine learning algorithm to pick up on the cues of natural language, nevermind produce the crisp, evocative images that OpenAI is showing off.”

[via]