Technologies for Converting Digits to Text and Back (Translated from Russian) For research purposes, we will examine some methods for converting a given sequence of digits into meaningful text and back, the associated problems, and how to solve them. All methods are based on Artificial Intelligence (AI) and, in this document, are limited to writing examples of universal queries for publicly available free versions of AI. No sources were used in writing this material. All digits in Russian begin with either unique letters or two letters—see the table. Therefore, a table can be created that maps 10 digits to 10 letters (or two letters). Then, using AI, we can compose some meaningful text using either words from the table whose first letters directly indicate a digit, or "filler" words used to add coherence to the text. For AI using gigantic dictionaries, this task is not particularly difficult. Let's try converting the number "35762084" into text. -------------------------------------------------------------------------------------------------- Example Query #1: Create a set of text variables with values ​​and assign them to characters as shown in Table #1: 0=N 1=O 2=Dv 3=T 4=Ch 5=P 6=Sh 7=S 8=V 9=De Main Part of the Problem: Convert the given text in quotation marks, "35762084," into a sequence of individual characters IN THE SAME ORDER: 3 5 7 6 2 0 8 4 and replace each character with a number with any word that begins with the letter corresponding to the number, as shown in Table #1. For example, Replace "3" with "Requirement" or "Trust" or "Standard" Replace "5" with "Simple" or "First" and so on... Choose the letter case yourself based on your needs. Moreover, the resulting text should be short and coherent on any topic. To facilitate the task of creating a meaningful text, the text may additionally contain words NOT starting with "N", "O", "Dv", "T", "Ch", "P", "Sh", "S", "V", "De" in any appropriate place. Do not insert numbers into the text or use words denoting numbers. For example, do not use the words "five", "fifth", "six", "sixth", etc. Nothing in the text should indicate numbers. -------------------------------------------------------------------------------------------------- Perplexity AI Response: Match Table (Extended) Digit Symbol Example Words (Variables) 0 N New, Delicate, Elegant 1 O Single, Common, Ordinary 2 Dv Double, Two, Door 3 T Traditional, Requirement, Cake 4 W Wonderful, Clear, Clock 5 P Festive, Simple, Feast 6 S Chic, Champagne, Ball 7 S Snowy, Shining, Light 8 V Cheerful, Evening, Whirlwind 9 De December, Delicate Coherent text for the sequence 3-5-7-6-2-0-8-4 Traditional festive snow globe of two new evenings, the clock strikes. This short text describes a New Year's scene: a traditional festive snow globe of two new evenings, and the clock strikes (midnight). The extra words ("beat") don't break the rule. In principle, that's correct. But the text itself is rather unnatural. Let's try asking the same query in Google's AI (browser version). Response: According to your algorithm (T-P-S-W-Dv-N-V-Ch) and the prohibition on using reserved letters in sequences, I present the corrected version. Text: The quiet ashes of fear stirred, the door slowly released a gloomy monster. Sequence analysis: 3 (T) — Quiet 5 (P) — ashes 7 (S) — fear 6 (Sh) — stirred 2 (Dv) — door 0 (N) — slowly 8 (V) — released 4 (Ch) — monster. Linking word: "gloomy" (begins with "M," which is not in Table #1, which meets the condition). The text "The quiet ashes of fear stirred, the door slowly released a gloomy monster" looks much more natural. This suggests that Google's language models are significantly more powerful. Using a query that requires the key letters (or two letters) to appear only at the beginning of a word narrows the list of words used. Let's slightly modify the query. -------------------------------------------------------------------------------------------------- Example Query #2: Create a set of text variables with numbers and assign them to characters as shown in Table #1: 0=N 1=O 2=Dv 3=T 4=Ch 5=P 6=Sh 7=S 8=V 9=De Main Task: Convert the text specified in quotation marks "35762084" into a sequence of individual characters IN THE SAME ORDER. For example, for "35762084" it would be "3 5 7 6 2 0 8 4" Replace each number character with any word that begins with the letter corresponding to the number, as shown in Table #1. For example, "3" should be replaced with "Requirement" or "Trust" or "Standard" "5" should be replaced with "Simple" or "First" or "Paracetamol" and so on... These will be "special words." If you lack suitable special words, you can use proper nouns, brand names (one word), names of dishes, names of medications, chemicals, biomaterials, or scientific or science fiction terms, or slang. Choose the letter case yourself based on your needs. Compose a coherent, meaningful text on any topic using these words. The order of special words in the text must be maintained. To make it easier to create meaningful text, the text may contain additional words that DO NOT BEGIN with "N," "O," "Dv," "T," "Ch," "P," "Sh," "S," "V," or "De" anywhere appropriate. Use at least 20% of the total text as special words. Don't insert numbers into the text or use words that denote numbers. For example, don't use the words "five," "fifth," "six," "sixth," etc. Nothing in the text should indicate numbers. The first two or three sentences should be introductory to the topic of the text and not contain any special words. You can insert sentences without special words throughout the text to give the text a more natural flow. -------------------------------------------------------------------------------------------------- Tried in Google AI (browser version) Answer: Below is a meaningful text, composed strictly according to your rules for the sequence 3 5 7 6 2 0 8 4 (T - P - S - Sh - Dv - N - V - Ch). Step 1: Variable Set (Table #1) 0 = H 1 = O 2 = Dv 3 = T 4 = Ch 5 = P 6 = Sh 7 = S 8 = V 9 = De Step 2: Text (Topic: Alchemy) The Archmage was preparing an elixir. The flask slowly boiled. The moon cast a glowing light over the laboratory. The red wormwood herb mixed gently, silkily. The thick oil stirred, causing the bubbling substance to instantly turn black. Rule Compliance Analysis: Introduction: The first three sentences ("The Archmage was preparing an elixir...") do not contain special words or words beginning with reserved letters (H, O, Dv, T, Ch, P, Sh, S, V, De). Special words (3-5-7-6-2-0-8-4): 3 (T) — Grass 5 (P) — Wormwood 7 (S) — Mixed 6 (Sh) — Silky 2 (Dv) — Moving 0 (N) — Oil 8 (V) — Substance 4 (Ch) — Blackened Additional words: The words "red," "carefully," "somehow," "thick," "bubbling," and "instantly" are used. They do not begin with letters from Table 1. No numbers allowed: The text does not contain numbers or words denoting them. Volume: Additional words (including introductory sentences) make up more than 60% of the text, which exceeds the required limit of 20%. It looks interesting, but it doesn't look very natural. The query still needs some refinement. -------------------------------------------------------------------------------------------------- Example Query #3: Create a set of text variables with numbers and assign them to characters as shown in Table #1: 0=N 1=O 2=Dv 3=T 4=Ch 5=P 6=Sh 7=S 8=V 9=De Main Task: Convert the text specified in quotation marks "35762084" into a sequence of individual characters IN THE SAME ORDER. For example, for "35762084" it would be "3 5 7 6 2 0 8 4" Replace each character with a number with any word that begins with the letter corresponding to the number, as shown in Table #1. For example, "3" should be replaced with "Requirement" or "Trust" or "Standard" "5" should be replaced with "Simple" or "First" or "Paracetamol" and so on... These will be "special words." If suitable special words are lacking, you can use proper nouns, brand names (one word), names of dishes, names of medications, chemicals, biomaterials, or scientific or science fiction terms, or slang. You can insert any prepositions—"in," "on," etc.—even those that match the initial letters of the special words. We won't take them into account in the transformation. Thus, the phrase "ran into the field" will correspond to the numbers "85," not "885." Choose the letter case yourself based on your needs. Compose a coherent, meaningful text on any topic using these words. The order of the special words in the text must be maintained. To facilitate the creation of meaningful text, the text may contain additional words that DO NOT begin with "N," "O," "Dv," "T," "Ch," "P," "Sh," "S," "V," or "De" anywhere appropriate. Use at least 20% of the total text as additional words. Do not insert numbers or use words that denote numbers. For example, do not use the words "five," "fifth," "six," "sixth," etc. Nothing in the text should indicate numbers. The first two or three sentences should be introductory to the topic of the text and not contain any special words. Sentences without special words may be inserted throughout the text to give the text a more natural feel. -------------------------------------------------------------------------------------------------- A further refined query. An attempt to achieve a more natural text. Example Query #4: Create a set of text variables with numbers and assign them to characters as shown in Table #1: 0=N 1=O 2=Dv 3=T 4=Ch 5=P 6=Sh 7=S 8=V 9=De Main Task: Convert the text specified in quotation marks "8674520159" into a sequence of individual characters IN THE SAME ORDER. For example, for "8674520159" it would be "8 6 7 4 5 2 0 1 5 9 4" Replace each character with a number with any word that begins with the letter corresponding to the number, as shown in Table #1. For example, "3" should be replaced with "Shadow" or "Trust" or "Type" "5" should be replaced with "Ceiling" or "Plane" or "Paracetamol" and so on... These will be "special words." Special words in the text should ONLY be used as nouns. If suitable special words are lacking, you can use proper names, brand names (one word), names of dishes, names of medications, chemicals, biomaterials, or scientific or science fiction terms, or slang. You can insert any prepositions—"in," "on," etc.—even those that match the initial letters of the special words. They will not be included in the transformation. Thus, the phrase "ran into the field" will correspond to the numbers "85," not "885." Choose the letter case yourself based on your needs. Compose a coherent, meaningful text on any topic using these words. The order of the special words in the text must be maintained. To facilitate the creation of meaningful text, the text may contain additional words that DO NOT begin with "N," "O," "Dv," "T," "Ch," "P," "Sh," "S," "V," or "De" anywhere appropriate. Use at least 20% of the total text as additional words. Do not insert numbers or use words that denote numbers. For example, do not use the words "five," "fifth," "six," "sixth," etc. Nothing in the text should indicate numbers. The first two sentences should be introductory to the topic of the text and should not contain any special words. Show the text and highlight the special words used in bold. -------------------------------------------------------------------------------------------------- Google AI's answer. Works very quickly, even in the browser version. Here's the solution to the problem according to your updated rules for the sequence 8 6 7 4 5 2 0 1 5 9 (H - W - S - C - P - Dv - N - O - P - De). Step 1: Correspondence Table (Table #1) 0 = H 1 = O 2 = Dv 3 = T 4 = C 5 = P 6 = W 7 = S 8 = H 9 = De Step 2: Text (Topic: Laboratory Description) Dense, wet snow fell slowly outside. The lamps brightly illuminated every corner of this small room. The Great Spring had passed, and the cold Storm had died down, while old Mica covered the white Tea. A small vapor slowly filled the Yard, as if thick Neon had enveloped the Lake beyond the edge of the Arable Land on the eve of Decadence. Text Analysis: Introductory Sentences: The first two sentences ("Outside the Window...", "The Lamps Are Bright...") do not contain special words and do not begin with letters from Table 1. Special words (nouns): 8 (V) — Spring 6 (Sh) — Storm 7 (S) — Mica 4 (Ch) — Tea 5 (P) — Steam 2 (Dv) — Yard 0 (N) — Neon 1 (O) — Lake 5 (P) — Cultivation Fields 9 (De) — Decadence Prepositions: The prepositions "za", "v", and "na" are used, which, according to the condition, are not counted in the sequence. Additional Words: Adjectives and verbs that do not begin with reserved letters are used: "slowly", "fell", "thick", "wet", "passed", "quieted down", "covered", "filled", "as if", "enveloped". Digit Restriction: The text is completely devoid of numerals or any hints of them. It's interesting. The text looks somewhat fantastical. But there are still resources for transitioning to a more natural look. Google wrote the following assessment of the task: The text is a bit surreal due to the strict restrictions on initial letters, but this is precisely what gives it a special "cryptic" charm. Such tasks, which require character-by-character control of text while simultaneously observing phonetic/alphabetic restrictions, are among the most challenging for AI. The problem is that most language models operate not on letters, but on "tokens" (groups of characters), so they often "don't see" the first letter of a word the way we do. To successfully solve such a puzzle, the model must constantly "recheck" itself at the level of individual characters, which requires a high level of contextual focus. https://t.me/llmsource/ http://oflameron.com https://github.com/vallshmeleff/oflameron-obfuscation-language https://androidjavaopensource.blogspot.com/ https://ailanguagesdatasets.blogspot.com/ https://12f3688e.pinit.eth.limo https://proposed-gray-cattle.myfilebase.com/ipfs/QmXMVBgdZez4VfeKpoKKfSJ3j1oVfTHztmnvC2oYNSoz3x https://pixelfed.ru/p/Jettman/929298265312142479 https://t.me/llmsource/ http://oflameron.ru/Valery_Shmelev_5eng.pdf