You have commented 375 times on Rantburg.

Your Name
Your e-mail (optional)
Website (optional)
My Original Nic        Pic-a-Nic        Sorry. Comments have been closed on this article.
Bold Italic Underline Strike Bullet Blockquote Small Big Link Squish Foto Photo
Cyber
American neural network policies
2026-07-21
Direct Translation via Google Translate. Edited

Text taken from a NY Times article

Commentary by Russian military journalist Boris Rozhin is in italics.

[ColonelCassad] Neural Network Politicians: American politicians at all levels are already fully utilizing AI services to manage their reputations and those of their competitors. Politicians are trying to change what chatbots say about them. An industry known as Answer Engine Optimization (AEO) has emerged in response to changes in how AI-powered chatbots generate their responses. The first chatbots, released several years ago, were trained on massive amounts of information already published online. This meant that many of their responses to news events became outdated almost immediately.

Modern chatbots have solved this problem by searching the internet and curating fresh content to answer questions about current events, including midterm elections and candidates. "There's a certain panic around this because we understand so little about it, and everything changes so quickly." "Even if there are people who claim to understand it, even if they've mastered some techniques, the volatile and unpredictable nature of LLMs means it's very difficult to control," said Beth Simone Noveck, a professor at Northeastern University who studies democracy and artificial intelligence tools known as large language models (LLMs).

Politicians also weigh another risk: being tainted by AI-generated search results that are not only unflattering but also downright inaccurate. An experiment conducted before the Scottish general election in May found that more than a third of AI-generated responses to questions about the upcoming election were inaccurate.

Researchers from the British think tank Demos found that many chatbots invented candidates, fabricated accusations of nepotism, or invented financial scandals. Some falsely claimed that a sitting MP was running when they weren't, while others misrepresented the candidates' positions. Caucus AI conducted an experiment to determine how long it takes for content recently published on Wikipedia to be responded to by a chatbot. The result: about 12 minutes.

Deus Ex is getting closer.
Whatever that means, Borya
Deus ex machina, perhaps. Literally a stage direction in Classical Greek and Roman plays, informing the reader that a god was to be lowered to the stage by machine to resolve the plot. Figuratively now used to suggest that something unexpected or artificial is introduced to the situation to resolve a difficulty. Derisive when describing badly written fiction, astonished when it happens in real life.
Meanwhile, Xi Jinping announced the formation of a Chinese artificial intelligence coalition involving 29 countries, including Russia, Belarus, Brazil, Kazakhstan, and Indonesia.

Steps in this direction were to be expected, given the attempts of the US and several Western countries to consolidate institutional superiority in AI over other countries.


Posted by:badanov

#1  comments); ?>werwer
Posted by: Grom the Reflective   2026-07-21 00:14  

00:00