And of another track: "By marrying a traditional Romantic harmonic language with modern recording clarity, the piece situates itself within the lineage of 19th-century piano miniatures while appealing to contemporary listeners who appreciate high-resolution acoustic sound."
"... Think Chopin’s nocturnes or Schumann’s ‘Kinderszenen’ -- Its slow tempo, expressive rubato, and nuanced dynamic shading invite the listener into an intimate, almost confessional space, while the modern recording quality situates the work within contemporary classical performance practice."
NVIDIA's state-of-the-art 8-billion parameter music intelligence model Music Flamingo says of one track: "... blending the expressive lyricism of 19th-century Romantic piano music with a modern high-fidelity recording aesthetic."
New Christmas album out!
bobltsturm.bandcamp.com/album/en-sv-...
Meta's Audiobox rates its "content enjoyment" of the album at 6.98±0.12 (out of 10), while its rating of fifteen of the greatest Christmas songs ever recorded (e.g., Nat King Cole, Burl Ives, Mariah Carey) at 7.31±0.27.
The problem of "the problem of the blank page"?
Hi Jan! Write to Filippo. I am sure he will find a way to help you.
@article{McNealis2025a,
author = {R. McNealis},
journal = {AI \& Society},
title = {{Shame in the machine: Affective accountability and the ethics of AI}},
year = {2025}}
Sturm, B. L. (2014). A simple method to determine if a music information retrieval system is a “horse”. IEEE Trans. Multimedia, 16(6):1636–1644.
Hicks, M. T., Humphries, J., and Slater, J. (2024). ChatGPT is bullshit. Ethics and Info. Tech., 26(38).
References:
Bender, E. M. and Hanna, A. (2025). The AI Con: How to Fight Big Tech’s Hype and Create the Future We Want. HarperCollins.
But hey, now we can generate videos of bunny rabbits jumping on a trampoline at 2 AM caught by a surveillance camera.
There are too many people deluded that working with these chatbots provides them an advantage over others even though in reality they are just consulting a fancily dressed moral-void truth-ignorant and sycophantic-tuned autocomplete.
We have left one pandemic only to willingly walk into another, but there is no vaccine this time. There are too many opportunists, too many profiteers, too many cons.
among many other things, sexual violence against children, and are given no professional support when they suffer a mental break down from the work and cannot continue—thus taking their trauma back home without any tools to heal.
And to fine tune these horses such that they don't spew out horrific psychologically traumatizing garbage (which is in their training data), people in the third world are paid pennies to listen to and penalize a horse describing, ...
Some cities around the world are clamoring to build massive data centers to build bigger and more powerful horses (they can't be anything else but horses), diverting water, electricity and clean air away from human communities.
Phony Stark is advertising his horse as the most racist, and so the U.S. Department of Defense is paying $200 million to use it.
We are watching a horse correctly answering math questions (Sturm 2014) and then seeing powerful people proposing that that horse should be made available to everyone to perform tasks of all kinds.
to failed legal proceedings due to faked citations, to cheating in education (both students and teachers), to the lack of critical pushback against antisocial and suicidal behaviors (Bender and Hanna 2025).
They just know that one symbol often appears in the presence of some other symbols in a particular order. Cautionary tales of the uncritical use of these chatbots are becoming more numerous every day, from people poisoning themselves, to ruined holiday plans, ...
These chatbots are _not_ trained to output truth, and they are _not_ trained to reason. They are trained only to generate text that is computed to be highly probable, which only sometimes overlaps with truth. These chatbots do _not_ know the meanings of the symbols they traffic in.
The image is a screen grab of an article in The Guardian newspaper from 23 Aug 2025 titled "Deal to get ChatGPT Plus for whole of UK discussed by Open AI boss and minister: Deal that could have cost £2bn was floated at meeting between technology secretary Peter Kyle and Sam Altman"
Dear Lord Almighty—this is beyond mad! ChatGPT and its ilk are not what its creators are selling them as. These chatbots are _nothing more than bullshiters_ (Hicks, Humphries, and Slater 2024) the use of which poses great detriment to anything and everything that revolves around truth and evidence.
I didn't say property rights should be done away with.
Since I'm just an academic too, I don't know. Might universal basic income go some ways to solving many of these thorny problems around IP?
What would specific uses be? I could see regulations about using mug shots to infer "criminality". But what could it be for music?
What do you think of this? papers.ssrn.com/sol3/papers.... It made me rethink my position on this issue.
Thoughts and prayers for the GOP today as they strategize over how to spin their cooperation with a convicted and unremorseful serial child sex trafficker.
