Please wait - loading


This web page is my brainchild, an example of artificial intelligence. It's name, AMR, is an acronym for Associative Message Reasoner. The words making up the default settings are ones that sound enjoyable, or are from memorable conversations.


line(s) with at least word(s) each or quote(s), and vocalize output, or



Click inside bars to edit code and outside to preview it. Comments look like "<!--"this"-->." Time and entry number are encoded as IDs inside generated comments - "<!--BEGIN "[ID]"-->" and "<!--END "[ID]"-->."
= Entry # made [date].

Construct A Link

Please keep in mind that some override others.

Youtube video: - assigned to seed if seed is empty.
Passphrase word count:
Use defaults
Quote count:
Minimum words per line:
Lines (cipher rounds):
Hide generated text
Hash encoding format:
Show etymologies
Show debug info
Background image:
Image orientation:
Key encoding format:
Maximum raw key length:

Generated Link




Words and Numbers
Sentence patterns:
List sizes:

Max words per line:
Lines with max words:


The following sections require advanced computing skills to configure properly. They come preset for optimal behaviour. Sections that are editable have associated controls and use double-quotes ("") in their explanatory notation, to signify literal values.

You can program the page execution using the query string part of the URL (link) that took you here. Use "q="# to make # of quote(s), "q="[text] to show a quote matching [text] (empty if none), "w="# to make # of word(s) per line, "l="# to make # of line(s), "i="[text] to show response to [text], "in="[text] to show interpretation of [text], "s="[state code] to initialize seed to [state code], "r=1" to randomize the state, "pw="# to show a password with # of words, "v=0" to disable vocalizations, "d="[text] to reset to default configuration, or "v="[text] to speak text, which is URL-safe Base64 encoded, "b="[text] to set the background, which is also URL-safe Base64 encoded, "o=h" or "o=w" to set orientation for background to height ("h"), or width ("w") and "y="[video ID] to link title to a Youtube video. Video ID is the variable with the label "v=", and goes up to the "#" or next "&." Another option is to set the title of the pages with "t="[text], which again is URL-safe Base64-encoded. Other possibilities are including links to word etymologies, by using "e=1", and/or the hash of each line's state code, with "h=bin" for binary, "h=hex" for hexadecimal, "h=oct" for octal, "h=b36" for Base36, "h=b64" for Base64 or "h=u64" for URL-safe Base64. The hash parameter value must be URL-safe Base64-encoded. This also supports key generation; use "k="[format] to set key format and "kl="[#] to set maximum key length. The format options for the key are the same as hashing, again URL-safe Base64 encoded, and this feature requires sentence generation. You can hide generated words by passing the "hg=1" input. Another thing you can do is tell it to supply debug info, by using "db=1" in the query string. On top of that, you may specify that you want to link to HTML anchors in another web page, by including "a=1" in the URL of this one.

Input-response List

Format is [inputA1] "|" [inputA2] ":=" [responseA1] "|" [responseA2] "~" [inputB1] "|" [inputB2] ":=" [responseB1] "|" [responseB2] "|" [responseB3]. You must include proper spacing and capitalization in each response.

Word Lists
# - Type (subtype); description <format>
Within format, curved brackets contain (optional) values, [n.]=noun, [v.]=verb, [adj.]=adjective and [adv.]=adverb.

Values are separated by commas.

Joining Words

Joining words are filtered out when quote-matching. Values are separated by commas.

Sentence Patterns

Values are separated by line breaks. A line from these patterns is chosen. Each number of the line is then replaced with the word from the word list matching that number. To see last used pattern, press "find last used" button, right-pointing arrow key on keyboard then the button again if needed.


Values are separated by line breaks.

Word Forms

This lists all root-words and their variations from the linking list. The format is [prefixA1] "|" [prefixA2] "+" [rootA] "+" [suffixA1] "|" [suffix2] ";" [rootB1] "+" [suffixB2] "|" [suffixB3]. Prefixes and suffixes are optional. Accepted text is all alphanumerics and "-;+|'" (double-quotes excluded).

Anchor Linking

Anchor List

This lists HTML anchors in another web page. Values are separeted by ";". Optionally, you may prefix each value with a label. The format is [label]=value. Accepted text is all alphanumerics and "=-;+|:" (double-quotes excluded).

Andrew M. Roberts © 2013-present
All rights reserved