Automating Multilingual Blog Management with AI | May 03 2024, 18:46

I refined the mechanism for cross-posting to Russian and English blogs. Firstly, I figured out how to group posts by topics, assign them tags, and categorize them. Plus, this is now done on the fly for new posts. I’ll write an article on hybrismart later, but the gist is that openAI vectors are first created for all posts, then they are divided into 50 groups via KMeans and sorted by their distance from the center. Subsequently, the first posts are selected (so as not to exceed the N Kb limit), and the script asks openai about the topic of this cluster of posts. Eventually, I end up with 50 topics, from which I choose, say, Art or Books, and then extract all posts close to the theme of art or books, again sorted by their distance from the theme. The accuracy isn’t very high, especially for posts with little text. Therefore, each post is fed into a local LLAMA3 8B on my laptop, and it decides whether it truly fits the theme or not. Overall, also with rare mistakes, but out of 2000 found by the script based on proximity, it left 600 on the theme of art, and generally quite well.

A separate script iterates over posts on beinginamerica, and there it corrects tags and categories for posts from the list provided by the script above.

I have already distributed posts on themes like art, books, science. Overall, everything is automated, and it’s easy to create another 10 new themes. I will be doing it gradually. For now, tags are only on beinginamerica; I will do it on raufaliev.com later.

Additionally, if a post has ENG in parentheses, it sends the piece after ENG to the English site, and the piece after ENG below in parentheses to the Russian site. This is convenient when I write a post in both languages simultaneously.

The title for my archive was generated through LLAMA3 8B, but openAI is still more powerful, albeit more expensive. For new posts, openAI GPT-4 is now used.

Neither LLAMA3 nor OpenAI GPT-4 excel at creating titles for texts that are too short and uninformative, often producing quite incoherent outputs. Feel free to read and smile.

#TechStories

May 02 2024, 13:05

I wondered, does the name Alex sound like “kitty” (le chat) in French? Or is that a stretch? And what about Nastya and Anastasia in the English-speaking world? I might be wrong, but all sorts of unkind and uncultured people might smirk, recalling nasty and anesthesia. Is this far-fetched, or does it really pose problems?

May 02 2024, 12:02

Patricia Piccinini; р. 1965. Australian artist and sculptor. What do you think about her art? For me, she is a good candidate for my #artrauflikes

P.S. And here are some comments about her from YouTube from haters (the first one is a bomb):

  • “THE RUSSIANS DID IT!!!!”
  • “Sick. Not art”
  • “Never seen something so ugly!”
  • “This is probably what our Government is creating. Just another top secret operation.”
  • “This is EXTREMELY AND SERIOUSLY DISTURBING AS HELL. She says this represents the world we live in. Whose world. That tells us a lot about you and your sick and twisted world. This woman is a creep.”
  • “Supposedly a satanic pedophile favorite…sounds right. They will all soon get exactly what they deserve, god willing.”
  • “No wonder pedophiles buy this garbage…”
  • “She is garbage not artist.”
  • “Stupid talentless whore, please don’t insult the word ‘artist’.”
  • “Sick, sick, sick, I hope this is not something they are experimenting with that the “elite” have knowledge of.”
  • “Disgusting so ugly ughhh please stop calling this art”
  • “KILL THEM ALL WITH FIRE”
  • “By so far John pedestos favorite art. So much pedophilia filled in the world”
  • “The artist was a victim of sex trafficking and is obsessed with this sick twisted interpretation of Art…”
  • “It is not art, it is satanic!!”
  • “They all need 3 bullets in their heads!”
  • “This is the official art of the new world order.”

May 01 2024, 22:00

Major update on Hybrismart: Saturday Evening Post Archives: Engineering a Smart Search Solution.

The Saturday Evening Post (TSEP) is one of the oldest magazines in the U.S., published every Saturday since 1821. For many decades, it was a voluminous magazine with over 100 pages. Virtually none of its articles can be found through a simple Google search. It was immensely intriguing to index nearly 400,000 pages and to experiment with various technologies, from ML reranking to vector search using OpenAI Embeddings.

The article is technical and should be of interest to programmers and solution architects.

https://hybrismart.com/2024/05/01/saturday-evening-post-search-embeddings-solr-smart-search/

Saturday Evening Post Archives: Engineering a Smart Search Solution

May 01 2024, 19:22

Dearest Nadya — happy birthday! Stay as intelligent, beautiful, athletic, energetic, and vibrant as you have always been and I am sure always will be! What’s with the notebooks? Nadya — a volleyball coach, spending hours daily on the volleyball court and hours at home preparing for training. At home, we call this “drawing little squares.” For Nadya, it’s the little squares first, and then everything else, because a coach cannot come unprepared to the court! The gift from her girls “for when your notebook runs out of pages” touched her deeply 🙂 So thoughtful! I tell Nadya — someday you’ll publish a multi-volume series on how to raise superstar volleyball players. Happy birthday!

May 01 2024, 17:44

I really like this illusion. There are no red pixels in the image. Zoom in—the section where the coke can is contains only black and white, and around it is cyan. Color perception is shaped by consciousness.

Red is the opposite color to green. The black stripes make the brain think that the missing color is formed by the same rules as around the jar, and tonal contrast is replaced by color, thus creating the red color.

It’s hard to understand, but this image illustrates it perfectly!

May 01 2024, 14:48

I found a book by Lewis Carroll in which he describes his impressions from visiting Moscow with Henry Liddon in 1867.

He was most kind in answering our questions, and in giving us many hints on seeing Petersburg, pronouncing the language, etc., but gave us rather dismal prospects of what is before us, as he says very few speak any language but Russian. As an example of the extraordinary long words which the language contains, he spelt for me the following:

защищающихся

which, written in English letters, is Zashtsheeshtshayoushtsheekhsya. This alarming word is the genitive plural of a participle, and means “of persons defending themselves.”

May 01 2024, 12:01

This picturesque canvas, displayed in the artists’ exhibition hall as part of the “Youth of Petersburg” exhibit, is titled “Door to the Future” and closely resembles a well-known meme.

The cow is the most interested in the future, judging by the painting. And it looks like something is burning behind the door. It’s also unclear whether he is opening or closing the door here. More likely the latter. The guy by the door is resembles Putin as of 2005. In the lower left corner, an artist whose name history has not preserved was covering something up, either in haste or by order from above.

May 01 2024, 01:54

Artist James Swanson hails from a small town in Wisconsin. Life seemed to guide him towards a career as a forester until he realized that painting trees was much more convenient than sitting on them. His latest project was spurred by the passing of his dog, Amber. In her memory, James decided to initiate a series of paintings featuring dogs, one per day. Eventually, he painted so many pictures of wet dogs that it probably just became uneconomical to paint anything but dogs. By the way, he also has a painting featuring giraffes.

#artrauflikes