Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

Songtell

songtell.com

Post reply on HN
Print view View original

71–80 of 83 posts

Re: Songtell

#71

Post by YeGoblynQueenne »

>> Meaning of the song "Rebirth of the Nemesis (Enuma Elish Rewritten)" by Melechesh

>> The song "Rebirth of the Nemesis (Enuma Elish Rewritten)" by Melechesh is a reimagining of the ancient Babylonian poem Enuma Elish, which tells the story of the creation of the world by the god Marduk. The song focuses on the mythical figure of Tiamat, a primordial god often associated with chaos and destruction, and the forces arising from her return. The lyrics describe a powerful storm that stomps out the weak and leads to a million lost souls, symbolizing Tiamat's destructive power. The lyrics also depict her as a "Serpent Nemesis" who asserts her reign and links the polarities of conjecture from the celestial and terrestrial realms. As she ascends "with wings of liquid fire," the lines proclaim "Enuma Elish is re-written" and "Fate was never sealed in stone", emphasizing the power of Tiamat to rewrite destiny and exert her chaotic power. In the end, a powerful refrain warns, "Don't fear, don't, this deity, don't/Rebirth of the Nemesis has been foretold". Through the lyrics, Melechesh evokes the ancient legend of Tiamat as an intimidating figure of chaos and destruction— one to be feared and respected at the same time.

https://www.songtell.com/melechesh/rebirth-of-the-nemesis-en...

Hmmm... maybe. It lacks a certain je-ne-sais-quoi as an explanation. It's nice that it captured the bit about "wings of liquid fire" that the vocalist clearly loves repeating (I get goosebumps every single time) but it doesn't seem to really feel what this is all about. Enuma Elish is rewritten, man. Tiamat is back. Don't you get it? The Serpent Nemesis has returned! Returned with wings of liquid fucking FIRE!

OK, I guess I can't explain it either.

But I get goosebumps, every. single. time.

Edit: Btw, Tiamat identifies as a woman. Not "god" but "goddess".

Re: Songtell

#72

Post by azakai »

I tried a few nontrivial examples and it gets them completely wrong - either missing the point entirely, or having the exact opposite of the intention.

One issue is this appears to only process the lyrics of the song. But the actual audio matters a lot too in conveying emotion and meaning: the tempo, the vocal subtleties, the chords, the notes, etc. etc.

This also just doesn't look very deeply into relevant background. It doesn't look for references to other songs or works, nor does it use information about the life of the author at the time they wrote it, etc.

But, not to be too negative, in principle some of those things could be improved on eventually. I think it could be useful some day.

Re: Songtell

#73

Post by whatshisface »

wpietri wrote: post #58

I tried out Tom Waits's "The Piano Has Been Drinking", which it gets spectacularly wrong: https://www.songtell.com/tom-waits/the-piano-has-been-drinki... It saddens me that we're entering into an age of infinite machine-generated bullshit (and here I mean "speech produced without regard for truth"). A lot of it has negative information value, kind of like how misshapen molecules can block cell receptors. I hope someb…

>speech produced without regard for truth

People are not any different from transformer models in this regard even if they are well-intentioned. Every time there is a physics or finance thread on HN, a lot of well meaning software engineers transformer out absolute nonsense with all of the same "confident, language that signals an intelligent author, completely wrong" qualities that we see in these AIs.

Re: Songtell

#74

Post by wpietri »

whatshisface wrote: post #73
wpietri wrote: post #58

I tried out Tom Waits's "The Piano Has Been Drinking", which it gets spectacularly wrong: https://www.songtell.com/tom-waits/the-piano-has-been-drinki... It saddens me that we're entering into an age of infinite machine-generated bullshit (and here I mean "speech produced without regard for truth"). A lot of it has negative information value, kind of like how misshapen molecules can block cell receptors. I hope someb…

> speech produced without regard for truth People are not any different from transformer models in this regard even if they are well-intentioned. Every time there is a physics or finance thread on HN, a lot of well meaning software engineers transformer out absolute nonsense with all of the same "confident, language that signals an intelligent author, completely wrong" qualities that we see in these AIs.

I get your point -- people also produce bullishit -- but people are significantly different in that they are not only capable of finding the truth but mostly do so with some regularity. I expect that even the most "confident but wrong" software engineer has areas of their life where are perfectly able to recognize the truth and correct their statements for it. Indeed, observing children, I think the feedback loop of "confidently say wrong thing and self-correct over time" is how a lot of actual learning happens. Transformer models concern me because they are entirely missing that loop.

Re: Songtell

#75

Post by whatshisface »

wpietri wrote: post #74
whatshisface wrote: post #73

Earlier quoted context omitted.

> speech produced without regard for truth People are not any different from transformer models in this regard even if they are well-intentioned. Every time there is a physics or finance thread on HN, a lot of well meaning software engineers transformer out absolute nonsense with all of the same "confident, language that signals an intelligent author, completely wrong" qualities that we see in these AIs.

I get your point -- people also produce bullishit -- but people are significantly different in that they are not only capable of finding the truth but mostly do so with some regularity. I expect that even the most "confident but wrong" software engineer has areas of their life where are perfectly able to recognize the truth and correct their statements for it. Indeed, observing children, I think the feedback loop of…

ChatGPT also has areas where it can recognize the truth. I agree that some people can hedge their confidence in some areas, but it's not a universal trait that everyone exhibits all the time. I think this shows we're only sometimes generally intelligent.

Re: Songtell

#76

Post by ljw1001 »

whatshisface wrote: post #75
wpietri wrote: post #74

Earlier quoted context omitted.

I get your point -- people also produce bullishit -- but people are significantly different in that they are not only capable of finding the truth but mostly do so with some regularity. I expect that even the most "confident but wrong" software engineer has areas of their life where are perfectly able to recognize the truth and correct their statements for it. Indeed, observing children, I think the feedback loop of…

ChatGPT also has areas where it can recognize the truth. I agree that some people can hedge their confidence in some areas, but it's not a universal trait that everyone exhibits all the time. I think this shows we're only sometimes generally intelligent.

The real difference is in scale. Automations can be coordinated to produce self-affirming bullshit at a scale that drives real discussion out of view. You already see this on twitter with troll farms and primitive bots. Now it will be a tidal wave

Re: Songtell

#77

Post by bloaf »

another-dave wrote: post #33
jmathai wrote: post #2

I spent a lot of time in the early 2000s on discussion forums and songmeanings.net learning about and contributing to interpretations and meanings of songs. This is an interesting concept (AI lyric interpretation) and my comment is something a bit deeper than this concept itself. Perhaps I'm showing my age, but what makes lyric interpretation so cool is that it's the intersection of two people's life experiences - th…

We often hold AI to the bar of a highest submission in a field (critique this song to the level of a music/English verse critic) but if you think about the average quality it's performing much better in that context. (Just get it to output "this song is really about drugs" and you're now up to the bar of "average Song Meanings" submission!)

I agree, just think about how many people play Leonard Cohen songs (esp. Hallelujah) in religious/spiritual contexts and completely miss the heavy eroticism.

I checked out "Show me the place" which is incredibly erotic. Leonard asks to be shown "where the suffering began" (painful childbirth is the first punishment God mentions for Eve) and "where the word became a man" (i.e. where Jesus was conceived-Mary's womb) because "his head is bending low." The AI, just like a lot of people, can't get past that use of religious terminology to realize the sexual innuendo.

https://www.songtell.com/leonard-cohen/show-me-the-place

Re: Songtell

#78

Post by whatshisface »

ljw1001 wrote: post #76
whatshisface wrote: post #75

Earlier quoted context omitted.

ChatGPT also has areas where it can recognize the truth. I agree that some people can hedge their confidence in some areas, but it's not a universal trait that everyone exhibits all the time. I think this shows we're only sometimes generally intelligent.

The real difference is in scale. Automations can be coordinated to produce self-affirming bullshit at a scale that drives real discussion out of view. You already see this on twitter with troll farms and primitive bots. Now it will be a tidal wave

It already is a tidal wave just with people.

Re: Songtell

#79

Post by jmathai »

another-dave wrote: post #33
jmathai wrote: post #2

I spent a lot of time in the early 2000s on discussion forums and songmeanings.net learning about and contributing to interpretations and meanings of songs. This is an interesting concept (AI lyric interpretation) and my comment is something a bit deeper than this concept itself. Perhaps I'm showing my age, but what makes lyric interpretation so cool is that it's the intersection of two people's life experiences - th…

We often hold AI to the bar of a highest submission in a field (critique this song to the level of a music/English verse critic) but if you think about the average quality it's performing much better in that context. (Just get it to output "this song is really about drugs" and you're now up to the bar of "average Song Meanings" submission!)

Perhaps the difference is that I'm much less interested, perhaps not at all, in learning about a song's meaning as much as I am interested in connecting with others and learning more about myself through their experiences.

There's definitely a place for ingesting information purely for the sake of learning. That's rarely been my relationship with music though.

Re: Songtell

#80

Post by wpietri »

whatshisface wrote: post #75
wpietri wrote: post #74

Earlier quoted context omitted.

I get your point -- people also produce bullishit -- but people are significantly different in that they are not only capable of finding the truth but mostly do so with some regularity. I expect that even the most "confident but wrong" software engineer has areas of their life where are perfectly able to recognize the truth and correct their statements for it. Indeed, observing children, I think the feedback loop of…

ChatGPT also has areas where it can recognize the truth. I agree that some people can hedge their confidence in some areas, but it's not a universal trait that everyone exhibits all the time. I think this shows we're only sometimes generally intelligent.

> ChatGPT also has areas where it can recognize the truth.

Do you have some examples here?

Post reply on HN