MUSINGS: Words, worldviews, and the multimodal expansion of AI. [Let's have some fun with generative AI in 2024.]

"If language is as inextricably tied up with consciousness as it seems to be, then the continuing diminishment of our inclination to use it to express in letters the times in which we live could mean that an element of human consciousness itself is on the verge of disappearing."   --- Anonymous

I saw this interesting comment above during my readings across the Internet this past week. I guess it triggered in me some thoughts around writing articles these days with the rise of audience engagement with video, the bold new world of AI "generative" technologies, and trust in the sources we read/listen/watch. In this post, let's just "shoot the breeze" with some meandering musings around these themes and navigate through some complex ideas... 🤓

Years ago, during the heat of the MQA Affair, I talked about writing these blog pages using a pseudonym. That was in response to MQA's issue around anonymity; perhaps they were paranoid that I was an industry player or working for a competitor. These days, in the face of increasingly sophisticated artificial agents roaming the Internet, foreign influences, machine-generated content (including artificially generated audio and video), I think it's quickly becoming old-fashioned to worry about just mere human anonymity having nefarious intent!

Now that my writings have spanned more than a decade, I trust that readers know that I'm actually very much a "real" guy behind the keyboard who's passionate about audio reproduction, sharing my perspectives on the high-fidelity pursuit as an audiophile of the early 21st Century. I have my philosophical views, thoughts on the importance of objective hardware fidelity and adequacy, the importance music itself as the origin of the art we admire and the emotions evoked rather than attributing too much of this to the hardware.

Furthermore, I trust my writings have been internally consistent in presenting my audiophile worldview, built up based on objective fundamentals. Comparatively, I think this is quite different from what we typically read elsewhere coming from most magazine writers. My beliefs are often at odds with the interviews given by Industry experts, and I hope these accumulated articles have been vastly more in-depth than superficial opinions on social media.

Worldviews are powerful concepts that shape our interpretation of reality. This vision creeps into the contents of speech, what we do, and how we act in this world; sometimes subconsciously.

As I have suggested a number of times over the years, the audiophile pursuit is a fascinating microcosm into the psychology of humans; a taste of how we interpret our world both through the intellectual understanding of technology and emotional experience of the art. Conflicts are where we see the stress points of differing worldviews - fearlessly delving into those conflicts and controversies is often the most enjoyable part of writing "musings" articles! Within our little corner of the tech universe, we can observe those cultural workings, question whether certain beliefs are based on actual knowledge, appreciate the power of public perceptions, and of course consider the financial motives that drive modern industry. We can see how freedom to speak openly about these conflicts and controversies without risk of getting harmed by any "audiophile mafia", public mob, or censorship is essential.

I hope readers have seen that I have had no qualms with critically calling out companies, their representatives, or members of the media; always done in a respectable manner, I believe. Expressing criticism alone is not enough. I hope my viewpoints have been evidence-based, fair, and I have clearly delineated facts from matters of opinion in a balanced manner.

Recently, I was mulling over whether to create YouTube content since I know many hobbyists these days simply don't read as much anymore. 🤔

I don't blame the public. With all that's happening, pushing and pulling our lives in all directions, it's often easier to just get the news by watching videos on YouTube rather than reading articles. My kids certainly spend a lot of time affected by what they hear and see rather than cracking open the spine of a book or reading through news articles compared to when I was at their age.

Nonetheless, I agree with the melancholic and somewhat sentimental assessment in the anonymous quote above. There is something special about words and expressing ideas "in letters" that we leave behind as evidence of consciousness that might be getting lost (or perhaps diluted). [Some philosophers might even correlate thoughts with a mental language - "mentalese" - as in the Language of Thought Hypothesis.]

Between text articles and videos, I've often wondered, over time, which will have more enduring power to affect the worldview of audiophiles not just for the short term in this generation, but across the decades? Maybe at the end of the day it all depends on context and intent. One important element here is that while text is more easily searched and indexed in the search engines, videos have the ability to "go viral" for rapid transmission of ideas.

[Some interesting comparison articles: here's research for the power of video, this suggests preference for text-based student discussions, and some good points here regarding the benefits of text.]

I think what is clear is that video consumes a lot more data. A figure I've seen is that 1 hour of video is uploaded to YouTube every second of every day, and more new video is put up on YouTube every month than all 3 major US networks have created in 60 years! Imagine how much content that is in stored bits and bytes, and how much is needed to make sure they're safely archived! 😮 This alone could suggest that ideas in text might survive much more easily than maintaining publicly accessible video content into the medium to long-term future.

The power of the viral video I think is interesting. As many turn to video for their primary source of information, with all that massive content just on YouTube (not to forget the extensive social media networks like Facebook, X, Instagram, TikTok, etc.), coupled with the power of algorithms that track our behavioral patterns, interests, purchasing decisions, is it any wonder that we can all be directed to consume content that might be increasingly non-overlapping - catered to each individual's preferences and worldviews?

Just as oceans and mountains divide organisms which in time evolve into different subspecies and species, what happens to the "memes" in our minds if humans are constantly fed different sources of information if not the development of diverging perspectives? Diverging worldviews?

Is it then any wonder why technology can lead to fracturing of societal cohesion down to individual minds given the scale of this "marketplace of ideas" as we're bathed in valuable truths, variable-quality opinions, along side propaganda, falsehoods, and manipulations designed to appear trustworthy often by profit-driven entities? As audiophiles (and by extension members of societies), given the complexity out there, do we have proper "training" when it comes to critical thinking so that we can guard against being stuck on the algorithmic rails? Have we taught our younger generations to be able to stay grounded, to appreciate what is real, and to always make sure to not be lazy when it comes to doing proper research (likely through reading, thinking, integrating ideas from multiple sources - harder work than being fed streams of video content) before we give our minds over to all kinds of beliefs, especially those that lead to false, ignoble, or antisocial ends?

The pursuit of factuality might be one of the greatest challenges for modern humans in this technological age of information excess.

In comparison to broader political and religious controversies, I think audiophilia is at least much easier to navigate. Our answer is in my opinion to be "more objective" because our hobby is rooted in science, engineering, and the physical properties of electronics, transducers, and sound waves. The science of audio has evolved into maturity over decades already.

At the very least, I suspect that if we apply the pursuit of objectivity into other areas of life, we'll be steps ahead of many others in this world who drift along with trends, possibly operating outside fact-based foundations, and who ultimately are probably less capable of affecting the world around them seriously or for the better.

While there is always temptation to publish stuff on YouTube, I think there's something to be said about sticking around and making sure that ideas are clearly laid out and expressed through words, graphs, and infographics which I still think are more thoughtful examinations of truths. Perhaps this is as important as ever given the gradual decline (hopefully not demise) of the traditional print media.

Recently, I saw that Sound & Vision Magazine has closed and subscribers are being compensated with digital Stereophile subscriptions for the remainder of their term. While I am sure that there are overlapping groups of hobbyists who might be served by both magazines, somehow I don't think this substitution will satisfy most home theater enthusiasts!

--------------------

Let's pivot then to a related, and I think very important topic when talking about content on the Internet. I'm impressed by the progress of generative AI over the last year and a half since I posted on the "Audiophile Wisdom of ChatGPT" in early 2023.

As I had already discussed at that time, for audio-related content, subjective "reviews" are now very easy to create. Since subjective impressions and opinions are typically not grounded with measurables, so long as there's an air of truth, AI-generated content could be acceptable so long as the statements do not seem outlandish or bizarre within the culturally-determined belief system of the "tribe".

For example, let's throw this question/suggestion at the freely accessible ChatGPT 4o mini:

Archimago: 

Give me a 500-word review about the amazing sound quality from some US$1500 6' lengths of pure silver speaker cables by the Ultra-Transparent Cable Company terminated with very high quality rhodium-plated banana plugs. Music used to audition the cables were Miles Davis' "Sketches of Spain" album, Radiohead's "OK Computer", and Daft Punk's "Random Access Memories". Mention some songs in those albums that sounded particularly good. The streamer used to play the music was the beautiful $6000 Auralic Aries G2.2, the wonderful sounding $42,000 dCS Rossini Apex DAC, the gorgeous $29,000 Dan D'Agostino Progression Integrated Amplifier connected to the DAC with very high quality $600 Nordost Blue Heaven XLR cables.

ChatGPT 4o mini:

Review of Ultra-Transparent Cable Company’s Pure Silver Speaker Cables

When it comes to high-end audio, the significance of quality cabling cannot be overstated. Recently, I had the pleasure of auditioning a 6-foot pair of pure silver speaker cables from the Ultra-Transparent Cable Company, which are priced at an impressive US$1500. Terminated with high-quality rhodium-plated banana plugs, these cables promise an extraordinary listening experience, and they certainly deliver.

To truly gauge their performance, I paired them with top-tier components: the stunning $6,000 Auralic Aries G2.2 streamer, the exceptional $42,000 dCS Rossini Apex DAC, and the magnificent $29,000 Dan D'Agostino Progression Integrated Amplifier, linked via Nordost Blue Heaven XLR cables. The synergy among these components set the stage for a listening experience that was nothing short of revelatory.

Starting with Miles Davis’ "Sketches of Spain," the cables brought out an astonishing clarity and depth in the music. The opening track, “Concierto de Aranjuez,” was particularly mesmerizing. The delicate interplay between the trumpet and the orchestra was rendered with breathtaking precision. Each note from Davis’s horn had a distinct presence, resonating through the room with a lifelike quality that made it feel as if he were playing right in front of me. The strings, lush and warm, swelled with an emotional weight that lingered long after the music faded.

Transitioning to Radiohead’s "OK Computer," the sheer dynamic range of the album was showcased beautifully. “Paranoid Android” came alive in ways I had never experienced before. The intricacies of the layering—each guitar riff, synth swell, and Thom Yorke's haunting vocals—were separated yet cohesively united. The soundstage felt vast, enveloping me in the sonic landscape the band meticulously crafted. The cable’s ability to convey the nuances in the production was impressive; the cables didn’t just relay sound, they communicated the very essence of Radiohead’s artistry.

Then there was Daft Punk’s "Random Access Memories," an album renowned for its meticulous production and rich textures. Tracks like “Giorgio by Moroder” showcased the cables' incredible detail retrieval. The layered sound design, from the groovy bass lines to the airy synths and the mesmerizing spoken-word sections, felt immediate and vibrant. The cables ensured that every intricate sound had its place, creating an immersive experience that made the album feel alive.

Overall, the Ultra-Transparent Cable Company's pure silver speaker cables lived up to their name, offering a level of transparency and detail that allowed each recording to shine in its own right. The rhodium-plated banana plugs added a touch of elegance while ensuring a solid connection, further enhancing the overall performance.

In conclusion, if you’re looking to elevate your audio system to new heights, these cables are a worthy investment. They not only enhance the sound quality but also enrich the listening experience, bringing you closer to the music in a way that is both engaging and profound. For audiophiles seeking the ultimate in fidelity, the Ultra-Transparent Cable Company's offerings represent an exceptional choice that harmonizes beautifully with high-end components.

Nice, that was 492 words.

Within the prevailing audiophile culture today, which includes a common belief that >US$1000 speaker cables can significantly affect sound quality, that doesn't seem to be a bad writeup, and can easily be used with some minimum editing before publication (I'd adjust the Sketches of Spain portion since "Concierto de Aranjuez" is not a track to use to highlight the sound of strings).

Realize though that for those not in acceptance of the "value" of >$1000 speakers cables, the review sounds highly suspect, even delusional when it claims that these cables will "not only enhance the sound quality but also enrich the listening experience".

All of that can be "written" in less than 30 minutes. Given the time savings and ultimately the same outcome (eg. creating positive sentiments about this Ultra-Transparent Cable Company's product), whether a human wrote each word or machine generated, why wouldn't anyone just use generative AI to assist in creating these purely subjective reviews that do not require much creativity, intellectual insight, or higher-level reasoning?!

We now have even more tools to play with including the image generators (such as video generators like DeepAI, or runway), artificial speech (check out ElevenLabs), music/sound generators (have a listen to Udio), or stuff like NotebookLM which can take links, videos, notes and synthesize them into summaries.

One interesting feature NotebookLM can do already, freely, is to create synthetic 2-person discussions based on these materials. With a little light audio editing, I posted a few of these over the last few weeks based on blog material such as these:

A Dive into "Why Do People Equate High End Audio With Snake Oil":

A Dive into "Bits Are Bits":

A Dive into "Domains of Sound Quality and Important Audio Components":

A Dive into "MQA - Retrospective & Late 2024":

A Dive into "Cautionary Tale: Audiophile's Dream's End - Ken Fritz"

A Dive into "Mark Levinson promoting Daniel Hertz 'C Wave Technology'":

While clearly not perfect with occasional discontinuities, noises, and unusual vocal intonations which will be notable especially to fluent English speakers, I thought these turned out alright. Although the 2-person conversational style is not as nuanced and complete as the written articles they're based on, I think they provide enough to get some basic points across to listeners who might be interested in getting their information via short podcast-style discussions. I found the AI's ability to pick out salient points and present the material in a relatively engaging format surprisingly good in these early days of the technology.

Though we're still witnessing the infancy years of AI tools being available to the general public, the rapidity of "evolution" is obvious. Of course there is hype and there will likely be disappointments in AI and the billions invested already. I've noticed more articles like "The AI bubble is looking worse than the dot-com bubble" out there which can be a cause of concern. However, through the ups and downs ahead, I believe what's going on here is very meaningful and will change society in tremendous ways, like it or not. Obviously, what's happening in the minds of real biological neural networks (in our minds) is being simulated, replicated, gradually with artificial learning across different modalities and tasks. Without resorting to quantum processing nor appealing to some spiritual "ghost in the shell" - neither of these appear necessary. [Clearly, artificial networks are nowhere as efficient as the 20W engine in the human brain, but this too will improve in time.]

Whether folks like Ray Kurzweil is right that the technological "singularity" is near or if AI becomes the last invention for humanity, at this point in history, it's still just a tantalizing dream (or nightmare) that doesn't really seem that far away. I would not be surprised to see some form of Artificial General Intelligence (AGI) in the next decade regardless of the success and decline of various companies in the AI space. Maybe we can revisit this in the next 10 years if I'm still around writing stuff. 😉

Unlike Kurzweil, I'm not sure I'd seek immortality as a major goal. Perhaps it's good enough if one day the contents of "Archimago the Audiophile" on this blog could be imported into a machine to be re-animated as a convincing human-like example of an audiophile who lived, thought, and expressed himself at the start of the 21st Century.

That might be a fun chatbot to debate the merits of speaker cable risers with in the 22nd Century if that's still a thing. 🫣

Exciting stuff ahead happening in our world and a great time to be alive, so much to think about, fellow audiophiles! Hope you've been enjoying the music. Alas, I've got some lectures to attend to for the next little while... Chat later.

To end, let's have some fun! Here's some music to check out, done in collaboration with Udio for songwriting credits:

"A sad blues song about audiophiles who love their expensive cables, speakers, and digital audio converters losing it all because their partners found the machines too ugly and unacceptable in the living room."

Audiophile Blues

[Verse 1]
All those cables in the corner, you know just why
All those cables in the corner, you know just why
You called it clutter, but oh, they made the music fly
Never heard a better sound, those sweet hi-fi
Now they're gone 'cause they didn't match the room
Left behind with only a cryin' tune

[Chorus]
Audiophile blues, baby, audiophile blues
Audiophile blues, baby, audiophile blues
Gone are the sounds that used to make me sway
Left with silence, since you moved 'em away

[Verse 2]
Speakers in the corner, stood proud and tall
Speakers in the corner, stood proud and tall
You couldn't see their beauty, only saw them fall
Tunes played perfectly, but you didn't hear the call
Now the room is empty, echoes of goodbye
Stopped the music, made me say why

[Chorus]
Audiophile blues, baby, audiophile blues
Audiophile blues, baby, audiophile blues
Gone are the sounds that used to make me sway
Left with silence, since you moved 'em away

[Bridge]
Gold cables that sang for me, pulled away
Digital dreams crashed down, cleared the display
Now I'm left listening to a silent plea
Sad melodies, in a room where sound used to be

[Chorus]
Audiophile blues, baby, audiophile blues
Audiophile blues, baby, audiophile blues
Gone are the sounds that used to make me sway
Left with silence, since you moved 'em away

Happier synthpop with some female perspective...

Cables and Dreams (Costly Love)

[Verse 1]
Wires tangled in the dark (Oh)
Sound waves driftin' far apart
(Drive me crazy, drive me crazy, drive me crazy)
Speakers lined up in a row
DACs gleam, let the music flow
(Drive me crazy, drive me crazy, drive me crazy, drive me crazy)

[Chorus]
This is costly love (Oh)
All these cables and dreams above (Oh, baby)
This is costly love (Oh)
All these cables and dreams above (Oh, baby)

[Verse 2]
You said "They gotta go" (Oh)
But our bond broke stereo
(Hold your head high, hold your head high, hold your head high)
I chose tech over your heart
We drifted, fell apart (Oh)
(Hold your head high, hold your head high, hold your head high, hold your head high)

[Chorus]
This is costly love (Oh)
All these cables and dreams above (Oh, baby)
This is costly love (Oh)
All these cables and dreams above (Oh, baby)

[Bridge]
Audiophiles' paradise
Turned into costly sacrifice
Where the music was meant to be
Lost it all, now just you and me

[Chorus]
This is costly love (Oh)
All these cables and dreams above (Oh, baby)
This is costly love (Oh)
All these cables and dreams above (Oh, baby)

[Outro]
(Hold your head down, hold your head down...) 

Okay, not exactly aiming for the Grammies, but some of the ideas and music could be workable and this can be pretty fun and creative. I've already got some ideas around tweaking the lyrics and remixing the sound.

Finally, how about some instrumental classical/soundtrack-like music? Here's Echoes of Valor in 3 movements:

I've seen articles posing the question of whether computers can be creative. IMO, absolutely, probably surpassing the creativity of most humans. However, until there's AGI, we still need some kind of human "central executive" to issue commands as to what to create and whether that output satisfies the intent. The only people who would seriously question the capacity for creativity are likely insecure artists. 🙂

Cheers!