The Prayer

A concept by Diemut Strebe. “The Prayer” is probably the first robot that speaks and sings to God, all Gods. A rough design (inspired to a machine produced by Japanese scientists that replicates the human vocal tract) is combined with a cutting edge neural language model, fine tuned on thousands of prayers and religious books from all over the world. The prayer generates original prayers vocally articulated by Amazon Polly's Kendra voice, and sings religious lyrics to the Divine.

Text by Enrico Santus. More info: Diemut Strebe

Diemut Strebe has made a previous appearance on WU:

Artist Diemut Strebe offered his 3-D-printed re-creation of the famous ear of Vincent van Gogh for display in June and July in a museum in Karlsruhe, Germany--having built it partially with genes from a great-great-grandson/nephew of van Gogh--and in the same shape, based on computer imaging technology. (Van Gogh reputedly cut off the ear, himself, in 1888 during a psychotic episode.) Visitors can also speak into the ear and listen to sounds it receives. [Wall Street Journal, 6-4-2014]
     Posted By: Alex - Wed May 25, 2022
     Category: Music | Religion | Technology | AI, Robots and Other Automatons

It's a nice representation of how prayers without meaning (such as automatically saying "Oh Lord bless this food and those about to eat it amen" just before lunch) are worthless, and feigning to listen (often while emitting white noise such as "yeah, yeah, ok, mmmhmm, oh yeah?") is also worthless.
Posted by Yudith on 05/27/22 at 11:27 AM
I need a grant!

Experiment: make a couple of thousand of these. Give a bank of computers all the phonetics of ancient languages and program them to generate combinations of them. Set each 'speaker' on a table with a plant and some small objects. Point cameras at the tables. Hook the cameras so an AI which can discern movement or changes. Run until every possible word/phrase possible has been 'spoken.' Check AI for any activity.

Goal: Prove (or disprove) there were magical words which could grow/kill plants, move/destroy objects, etc..

I hereby claim copyright of this idea and insist that any study based on it mentions me.
Posted by Phideaux on 05/28/22 at 01:17 AM
@Yudith: "Thoughts and prayers". Rather too poignant these last few days. I don't understand how your country can stand the hypocrisy.
Posted by Richard Bos on 05/28/22 at 02:54 PM
Too late, Phideaux. Mythbusters did it.
Posted by Virtual in Carnate on 05/29/22 at 11:48 AM
@Virtual -- I have no idea what they did (and absolutely would never trust any of their conclusions), but a rough calculation shows it'd take >2^30 seconds (something north of 3,000 years), which is why I'd want thousands of the 'speakers' working simultaneously.
Posted by Phideaux on 05/29/22 at 08:35 PM
Commenting is not available in this channel entry.