Wyniki wyszukiwania:

In "Outliars," Angie Waller examines the role of click-friendly headlines in amplifying misinformation during the Trump administration, using data from a 28-year archive of The New York Times. Connecting the themes of outliers and liars, the book depicts the heightened media coverage of Trump relative to past US presidential administrations and the impact technologies such as search engine optimization and headline A/B testing may have on this phenomenon. The book annotates headlines in a manner suited for an algorithmic lie detector, highlighting the futility of filtering out the noise...

Facebook's Automatic Alternative Text feature offers image captions to assist visually impaired users. However, the automatic image recognition and captioning system has limitations and can overlook important content and context. It typically lists recognizable objects in a mechanical way, such as "one person, smiling" or "person, outdoor, nature." Seeing Like a Computer highlights these shortcomings with a sampling of snapshots from Mark Zuckerbergâs Facebook timeline.

Sorry AI Took Your Job depicts a dystopian future where artificial intelligence (AI) has replaced all jobs, from radiologists to Zamboni drivers. Each page features dialogues or correspondences that illustrate the tough conversations that might happen if an AI takeover occurs. In a deliberate conflict of interest, ChatGPT, an artificial intelligence (AI) interface, was used to generate all of the following contents.

In "Grifting the Amazon,â Angie Waller follows a group of book sellers aiming to generate passive income on Amazon.com. To make a profit, the entrepreneurs in this book rely on the unseen labor of overseas workers producing content on a wide range of topics, from juicing to capybara care. This book offers a humorous tour of how you might encounter this "grift" through dubious authors, overly enthusiastic reviews, and mysteriously priced print-on-demand books selling for thousands of dollars.

Trippy Training Data introduces a unique dataset of labeled images for training a computer to mimic the experience of taking LSD. Images are sourced from a Reddit subreddit where people discuss their experiences getting high and document their hallucinatory visions using low-quality smartphone cameras. The images often showcase ordinary objects, such as leftover pizza or bathroom tiles, captured under harsh fluorescent lighting with exclamatory captions like, "WOH !" When an algorithm is fed training data, it will generalize rules, establish relationships, and detect patterns, even if there...