Language Log

"The 'Effective Accelerationism' movement doesn't care if humans are replaced by AI as long as they're there to make money from it", Business Insider 12/30/2023:

The Effective Accelerationism movement — a staunchly pro-AI ideology that has Silicon Valley split over how artificial intelligence should be regulated — appears to be walking a razor's edge between being a techno-libertarian philosophy and a nihilistic, even reckless, approach to advancing one of the world's most significant technological developments. […]

A riff on the effective altruism, or "EA," philosophy touted by tech influencers like Sam Bankman-Fried and Elon Musk, e/acc took off in 2023, though its exact origins remain unclear. The movement has attracted a cast of unlikely characters, including venture capitalist Marc Andreessen and convicted fraudster Martin Shkreli.

"EA and e/acc are mostly the same people," Emmett Shear, the former interim CEO of OpenAI, said in an interview with Meridian. "Their only difference is a value judgment on whether or not humanity getting wiped out is a problem."

[This has been drifting down my too-long to-blog list for almost 16 months — but better late than never, I guess, and the world could use some pejorative-flavored humor…] 

Colin Morris, "Compound pejoratives on Reddit – from buttface to wankpuffin", 6/28/2022:

I collected lists of around 70 prefixes and 70 suffixes (collectively, “affixes”) that can be flexibly combined to form insulting compounds, based on a scan of Wiktionary’s English derogatory terms category. The terms covered a wide range of domains, including:

• • scatology (fart-, poop-)
  • political epithets (lib-, Trump-)
  • food (-waffle, -burger)
  • body parts (butt-, -face, -head, -brains)
  • gendered epithets (bitch-, -boy)
  • animals (dog-, -monkey)

Most terms were limited to appearing in one position. For example, while -face readily forms pejorative compounds as a suffix, it fails to produce felicitous compounds as a prefix (facewad? faceclown? facefart?).

Taking the product of these lists gives around 4,800 possible A+B combinations. Most are of a pejorative character, though some false positives slipped in (e.g. dogpile, spitballs). I scraped all Reddit comments from 2006 to the end of 2020, and counted the number of comments containing each.

This is not at all the experience that I've had with multiple-authored papers — but it's funny:

“Thanks everyone for comments on the draft, here it is revised with all your edits.” pic.twitter.com/pSV16wQqca

— Ben Phillips (@benphillips76) September 17, 2023

Despite the evidence of my most recent relevant post, the best current speech-to-text systems still make mistakes that a literate and informed human wouldn't.

In this recent YouTube video on the history of robotics research, the automatic closed-captioning system renders "DARPA" as "Dartmouth":

Perhaps Victor can point us to an analogous ambiguity in Chinese poetico-political history:

Screenshots from @泰拳刚猛 on Weibo:

From the movie "Jak rozpętałem drugą wojnę światową" (How I Unleashed World War II):

The initial Q&A:

Q: Name und Vorname?
A: Grzegorz Brzęczyszczykiewicz.

Nancy Kathryn Walecki, "Sound as Ever: Gram Parsons and Harvard’s hand in country rock", Harvard Magazine July-August 2023:

During Parsons’s Burritos era, Thomas left Harvard to write his dissertation in a cabin on Mount Baldy outside Los Angeles. Now more of an older brother to Parsons than a proctor, he would take study breaks with him in town: “It was a whole different world from Heidegger and Wittgenstein.” Once, they met Janis Joplin in a nightclub parking lot. “This is my adviser from Harvard. He’s into phenomenology,” Gram said. “Wow,” replied Joplin. “I believe in ghosts, too.”

An NFL policy prohibits players from endorsing alcoholic beverages. So Coors found a linguistic work-around:

The next event in the Salon Sanctuary concert series is "Victorious Secret: Love Gamed and Gender Untamed in the Sparkling Courts of the Baroque":

Before the bars of gender binaries caged the mainstream operatic imagination, a golden age of fluidity guided the vocal soundscape. Virility declared itself with the castrato’s clarion high notes, while femininity spoke in earthy tessiture that plunged to shimmering depths.

Texts of the period revel in ambiguity, unfurling genderless narratives of anonymous lovers and unnamed beloveds. Stories of active pursuit and passive reverie remain alike at loose ends, with neat resolutions many movements away.

Please join us for this special program in honor of Pride Month, as the music of the past reveals a golden underground of nonbinary riches, accompanying us in our witness to a new Renaissance.

Current big news around here is the collapse of an elevated section of Interstate 95 due to a tanker truck fire.

As Wikipedia explains, I-95 "is the main north–south Interstate Highway on the East Coast of the United States, running from U.S. Route 1 (US 1) in Miami, Florida, north to the Houlton–Woodstock Border Crossing between Maine and the Canadian province of New Brunswick". It's also an important connection between the Great Northeast and the rest of Philadelphia, which enables this linguistic joke:

By the time I-95 is repaired and contact can be re-established with Northeast Philadelphia, our respective languages are likely to have diverged so much as to be mutually unintelligible.

— Ryan Godfrey (@rgodfrey) June 11, 2023

Benj Edwards has a mirthful article in Ars Technica (6/9/23)

Researchers discover that ChatGPT prefers repeating 25 jokes over and over

When tested, "Over 90% of 1,008 generated jokes were the same 25 jokes."

[includes an AI generated image of "a laughing robot"]

On Wednesday, two German researchers, Sophie Jentzsch and Kristian Kersting, released a paper that examines the ability of OpenAI's ChatGPT-3.5 to understand and generate humor. In particular, they discovered that ChatGPT's knowledge of jokes is fairly limited: During a test run, 90 percent of 1,008 generations were the same 25 jokes, leading them to conclude that the responses were likely learned and memorized during the AI model's training rather than being newly generated.

The two researchers, associated with the Institute for Software Technology, German Aerospace Center (DLR), and Technical University Darmstadt, explored the nuances of humor found within ChatGPT's 3.5 version (not the newer GPT-4 version) through a series of experiments focusing on joke generation, explanation, and detection. They conducted these experiments by prompting ChatGPT without having access to the model's inner workings or data set.

…

[Jentzsch and Kersting] listed the top 25 most frequently generated jokes in order of occurrence. Below, we've listed the top 10 with the exact number of occurrences (among the 1,008 generations) in parenthesis: