Exposure Therapy
Posted on June 5th, 2026
Filed under: General,Music News — Karl Olson @ 12:52 am
In response to:
I DIDN'T WANT TO END THIS THREAD BY JUST SAYING "WELL THAT WAS COOL WHAT'S NEXT," SO I OPENED UP NOTEPAD AND WROTE SOME STUFF. I'M A PRETTY TERRIBLE WRITER, AND IT'S PROBABLY RAMBLING AND AWKWARD TO READ, BUT I DO HOPE THE MESSAGE GETS THROUGH.
During the normal course of being an exceedingly niche, indie artist by every reasonable measure (lifetime streams, social media presence, traditional media presence, any kind of revenue, number of listeners, etc.) there’s few excuses to reflect on main without it falling into “Unwarranted Self Importance.” Of the 2000+ followers I have on atProto, the social platform I’ve had the widest audience on of any of them, I know given what posts or lists got them in the door that most aren’t there for my music. It’s the anime mutuals I’ve had for decades, the folks who have found me from the mouthing off I’ve had on various topics du jour on atProto, and old music friends who were there when I was first putting up songs online, so it’s all old hat to them.
So, when BEAT, a long time mutual dating back to the rhymetorrents era, added me to his big discography review to do list, and asked for a reasonably complete accounting of everything that was at least kind of accessible on online, I was already curious to see where it would go, so I gladly made up a whole spreadsheet. Admittedly, it’s still not perfectly exhaustive, but it was better than it had been in years as it at least had all of the albums that I’d posted on mp3.com by date with the cover art and such, something I only recently got around to fixing by getting all of my old CDs out of storage, then uploading clean rips to Bandcamp, in part because I was already kind of thin on new music in 2025 by my standards. Still, I figured it would be fun, but not necessarily anything more important, especially if he took it on genre by genre, not fully chronologically.
At the time he asked, I was already going through a difficult chase which I’ve already posted at length about, but the day before BEAT’s thread started, I was getting therapeutic botox injections from my maxillofacial specialist to try to loosen up the inflamed facial muscles that were causing constant jaw pain, which also correlated at points with the hyperacusis and chronic tinnitus, as I’d already ruled out so many other options. The thread started right as I needed another distraction the most because I was now going to have to be patient and see what was working.
Thus, to get to spend this past month with such a perfect excuse to engage sincerely with my past in the context of my now and possible future, to have an extended dialogue offered up where I get to simultaneously boast, cringe and meditate on my work and how it now impacts me, is an rare and special blessing. It provided a grounding reminder that amidst these tests and difficulties that are beyond my immediate ability resolve instantly, if at all, there will also yet be more chances for me to at least reflect and speak on my old work, and in that, not only better assess how much I actually have lost vs. my catastrophizing, but also to find the will to proceed, even in the worst case scenarios. I couldn’t have tried to force myself to reflect this way and reassess my core beliefs, but I know that a big reason I’m coming out of the past 6 months of chronic illness feeling like I don’t even care even if every day is still this rough until my last one, has been facing my own discography this month, and feeling better about the past 27 years than I had in while, certainly since this hearing and pain stuff kicked off. This all has made me want a lot more than 27 years more, so thanks to BEAT for giving me the exposure therapy I clearly needed. You’re a great guy.
Comments (0)
re(COVER)ry
Posted on May 28th, 2026
Filed under: Music News — Karl Olson @ 10:56 pm
So, not long after dropping my most recent Drum n’ Bass album, I had the great misfortune to partially lose hearing in my left ear (frequencies over about 5kHz) on the flight descent back home from vacation. It was accompanied unbelievable pain throughout my left side of my neck and face, as well as strong tinnitus and even hyperacusis and distortion in certain ranges (even my own voice could trigger it if I was loud enough, let alone other people, especially in reverberant enviroments.) Now, I wasn’t initially immediately worried as I had similar issues with drainage in that ear on flights, and I finally have a family doctor here, but yet, over 6 months later, it’s all only starting to relent a little just in the past week or so. It’s taken a chase involving various doctor visits, dentist visits, hearing tests, ENT visits, maxillofacial visits, glove in mouth RMT, acupuncture, CAT scans, ER visits, telehealth, MRIs, X-rays, antibiotics, antivirals, muscle relaxants, anti-anxiety meds, night guards and botox. Oh, and I see a chiropractor tomorrow on the advice of my dentist since the tinnitus gets worse when I shoulder check like while driving or look up. I’m still not even fully sure of what this is, more of just what it isn’t – sinuses are clear, nothing suspect in my brain, no major bone spurs/arthritis, probably nothing viral or bacterial as nothing much changed from course of those, thus ruling out ear infections or early shingles. Just seems maybe like TMJ triggered in part by barotrauma and maybe some bad habits/ergonomics, but again, that doesn’t explain the other neck issues necessarily. It sucks living in a House M.D. episode when dealing with real world turnarounds on diagnosis.
Needless to say, I’ve not really been in a mood to do much music as my left ear until relatively recently didn’t really make it a fun process, if not making it outright depressing and dysphoric, as if my senses were alien to myself. Even just wearing headphones on that side of my face wasn’t great some days. However, the “try everything and rule out as much as possible as fast I can approach” seems to have made some progress. It’s not a consistent pathway with plenty of two steps forward, one step back.
You see, an old internet friend has been going through my whole discography on Bluesky/atProto (an endeavor that’s still on going as of this post as he alternates between about 10 releases of mine, then picks up another artist from his “comment on their discography list”,) so I’ve been listening to his favorite songs from each release, adding additional footnotes. I’ve noticed since getting the botox, wearing a night guard to keep from clenching my jaw in sleep and some flexeril, the worst of the tinnitus, the worst of the hyperacusis and even some of the high frequency hearing loss have eased enough that I can both hear some hi-hats again in some my own work (if past me wasn’t too heavy with the hi-pass filters – never again!), and also not be blown by the mid-frequencies being too sharp. Basically, this is first glimmer of light at the end of the tunnel, even if it’s still a long slow walk ahead.
Still, to start getting myself back in motion (beyond just doing throwaway mashups & edits that don’t even need proper hearing, like confirming ADV totally ripped off the Chemical Brothers for their old trailer music, unhinged eccojammed comedy music or dropping emo vocals over a trap classic) I decided to record some new/finish up some old covers. I should really try to gather all of the loose ones I have into a proper playlist/zip file of just that newer material, but that’s for another day:
Anyway, I’m gonna figure this out, bit by bit, even if it’s just to tie up the loose ends sitting half finished in my reason folder and lyrics folder. After all, as that thread going through my discography wears on, I see all the places I’ve only partially covered, and think “no, I’m not done until I do a proper album of this or that.” It’s also been surprising to see new people stumble into the thread, both due to my reshares and the main thread itself. More than I expected have turned around, and bought my whole discography which motivates me despite how rough I feel some days. I still won’t go all in until I’ve given everything more time to settle in, but I’m back in the saddle, even if I’m not off to the races.
Comments Off on re(COVER)ry
Don’t Forget, You’re Here Forever: A Drum & Bass LP
Posted on November 27th, 2025
Filed under: Music News,Videos — Karl Olson @ 12:00 pm
Yeah, this year, I just dropped a couple of Dn’B albums, and a couple of loose tracks. The second LP is Don’t Forget, You’re Here Forever, another 10 junglist jaunt with lots of tech step, breakcore, neurofunk, liquid, fast soul, jump up, glitch and more in the mix. I’m sure drop another album or two like in 2026, as I did spring for some new plugins thanks to the support of my loyal, rad Bandcamp subscribers, who have been enjoying this album now for over a month before the general release, and to my Bandcamp customers who have continued to outpace streaming income last I’d checked.
Still, it’s out now for download, and you don’t even have to wait for Bandcamp Friday, it’s free: Don’t Forget, You’re Here Forever by Ultraklystron.
You can stream it too (click here for a full list of streaming and other download options) but to paraphrase MC Frontalot, “I’m so indie that my [songs don’t get paid out by Spotify. Daniel Ek takes my cut and gives it to big labels, and to himself to put into defense industry investments now as best as I can tell.]”
Maybe cancel Spotify and switch to another service if you can while you’re out listening to this new LP.
Comments Off on Don’t Forget, You’re Here Forever: A Drum & Bass LP
Surrounded By Asymptotes: 4 Papers Showing LLMs & DNNs Won’t Become AGI
Posted on November 23rd, 2025
Filed under: Code,General — Karl Olson @ 1:44 am
(This will be rather different than my usual, but I promise, the next post won’t be so heavy.)
Lately, it’s been wild to me how the markets and mainstream/enthusiast tech news generally have not sounded the alarm about how Apple, OpenAI, Anthropic and Basis have all published papers in past 6 months that, when considered together, underline that far from being anywhere close to the oft-prophesied “Technological Singularity,” Large Language Modeling (LLM) approaches, including so-called Large Reasoning Modeling (LRM), come up short with no path forward given the techniques used, and this even has implications regarding asymptotic limitations on the capabilities of Deep Neural Network (DNN) models generally.
Apple revealed that LLMs and even LRMs can’t truly reason, even when given explicit instructions and the opportunity for infinite run time. OpenAI discovered that LLMs will also always hallucinate, no matter the model size, training goals or data quality due to the intrinsic resolution limitations of any given parameter and the relationships between them in a DNN model. Anthropic showed that LLMs and diffusion models are easily poisoned regardless of the model size, as the amount of bad data required can be nearly constant, yet keep poisoning the output as a model’s parameters exponentially increase. Most damningly, Basis, an AI research firm, demonstrated through a new benchmark suite for World Modeling (put simply, DNNs for physical tasks and spaces) that LLMs and LRMs are all dramatically outperformed by humans in unique problem solving tests, as it can be shown those techniques all rely on their pattern matching abilities alone, unlike humans which can truly learn on the fly, likely further explaining Apple’s results. Just by those four papers, it suggests AGI shall never be born alone from the LLM/LRM/DNN approach: there’s no ghost in the machine to be found here without redressing those papers’ concerns, if that’s possible at all. Given that, it calls into question the gold rush surrounding the technology generally.
Financial viability and general ethics aside (I’ll mostly leave that to Ed Zitron, who you should subscribe to,) these shortcomings explain in some ways why, for some, LLMs can be amazing mirrors and amplifiers of one’s own beliefs, emotions and worldview, even to the point of self-destruction. This has been satirically demonstrated in abstract by comedy youtuber Eddy Burback, but also shown in research specifically buried by Meta which revealed that time off from their Machine Learning-guided feeds improved mental health, and reflected in LLM-induced downward spirals documented by medical journals and made easy to understand by medical youtuber ChubbyEmu. These pattern matching recommendation engines and conversation simulators, which are intrinsically dependent on the end user to guide their generated responses, will trend towards a feedback loop, ultimately pocketing themself and the user into a hyper-focused world in a fashion even the most self-congratulatory forums, chat rooms and skewed social media algorithms could never replicate. As an interactive medium, neither the model nor the human engaging with it are necessarily encouraged or even capable of checking the other by design (or perhaps rather by the lack thereof.) It is entirely upon the user to maintain a grasp of reality, as they are the only participant capable of conceptualizing it in any fashion that matters at all. In practice, the ability to rapidly reroll for a different response encourages gacha-game-like gambling mechanics, brow-beating any model’s responses down whatever path that user wants.
Similarly, that shows why LLMs & LRMs may demonstrate value to some people as a super-charged autocomplete, a draft writer, or an automatic clean-up assistant, especially for specific verticals with heavy repetition like software development, at least when audited by a careful eye that already has the domain knowledge to see the aforementioned impossible to stop hallucinations and artifacts. Barring other externalities, being wrong faster can be surprisingly fine, so long as it’s not always wrong, and what’s wrong is easy to spot and correct. It’s similar to how even classic autocorrection and autocomplete techniques, like those in word processors or Intellisense in Visual Studio, can relieve some cognitive load, even when rather imperfect and limited. Especially in something like unit test writing, which often involves nearly identical, very repetitive blocks of code, that can be a space where supercharged mad-libs might be favorable to a lot of tedious copying, pasting and hand-editing. That said, like those aforementioned traditional autocorrection functions, this means AI augmented professional work is just a feature, not software or services that can be sold or subscribed to alone, and they don’t provide enough value to demand a high fee as just a feature. It’s certainly not worth the investment bubble we’re seeing, nor the approaches being taken for developing and marketing it for mass-adoption.
No, as evidenced by myriad stories of vibe-coding creating more additional work than momentum, especially as code scales in size, if not the more damning and dangerous stories of agentic systems driven by LRMs creating malware vectors in popular operating systems, wasting the limited resolution available for any given problem space by diluting it with lots of unrelated (so effectively poisonous,) and certainly noisy data, is just not the path forward, even for the tasks this has shown some utility. On the contrary, this points towards efficient, secure, locally-run, relatively domain-specific models with highly validated, curated and legitimate training data, and again, only to build super-charged autocomplete and assistance for certain tasks of domain specialists, not to sell a subscription forever to a machine god slave as their various corporate creators often imply lies just beyond the horizon. On the contrary, all state of art LLMs and LRMs currently can have their guard rails consistently bypassed in a single adversarial prompt, so long as it’s in poem form.
That is to say, these companies’ own research, to say nothing of anyone else’s, shows they already are fully boxed in. Definitionally, given the findings of these papers, no matter the volume, curation and annotation of information that is shoved into these models, even given the most perfect training systems and benchmarks intended to avoid all rewarding of hallucinated results, and also given better technology to represent these models in hardware to reduce the discontinuities and increase data resolution to their theoretical limits, the mathematical space these techniques are based upon do not continuously re-balance and update themselves in the sense of reasoning as we know it in biological systems. Given that, trying to build general purpose models in that way will not get these companies and their clients the Swiss Army Knife they’re circulating trillions around. Even outside of LLMs, LRMs, and diffusion models, this likely means any modeling using a DNN approach to generate something, including doing a set of actions, as is being promised with everything from self-driving cars to forthcoming humanoid home robots using World Modeling approaches, will have those issues of constant-rate poisoning, hallucinations from data resolution limitations, and the inability to scale past their inelastic, pattern-matching underpinnings.
Even if we could ignore all of the external issues for this approach and its marketing to the world (and we can’t,) the internal issues are clear as crystal now, and the sooner we can all get on that page, and only use DNNs in the places they have improved things like computational photography or even for light recreation if trained and run ethically and sustainably, the more we can look at what other new frontiers in computing exist, discovering what they can do to genuinely make our lives directly, reliably better.
Comments Off on Surrounded By Asymptotes: 4 Papers Showing LLMs & DNNs Won’t Become AGI
Doot doola doot doo!
Posted on November 6th, 2025
Filed under: Music News,Videos — Karl Olson @ 12:17 am
Did some skits on YouTube with Nursehella for Halloween in our Nardwuar/Wardnuar costumes. They are pretty self-explanatory, enjoy:
Meanwhile, if you were subscribed to my Bandcamp, you could be already listening to a new Drum n’ Bass LP from me that drops later this month. More on that later of course.
Comments Off on Doot doola doot doo!
First Live Show In Over A Decade & A New Song I Won’t Play There.
Posted on July 11th, 2025
Filed under: Music News — Karl Olson @ 1:47 pm
First things first:
Yes, I’m playing live in Victoria, BC at the Phoenix Bar Grill, opening for Nursehella, Bixlee & Skull Cultist on August 1st, 2025. 19+, $20 CAD.
who do i know on here in victoria, bc?
i'm playing a rare* show next month with @karlrolson.com and @skullcultist.bsky.social
*i've literally never done this before**
**oh god i hope it doesn't suck***
***it won't suck
— ALT-BIXLEE (real) and 41 others (@bixlee.bsky.social) Jul 9, 2025 at 4:09 p.m.
Also, dropped this sludgy, pluggnb track about various romance manga that’ll probably go on my next album, which I’ll start real work on after this show, maybe:
Comments Off on First Live Show In Over A Decade & A New Song I Won’t Play There.
More Drum & Bass
Posted on February 20th, 2025
Filed under: Music News — Karl Olson @ 11:52 am
Also on various streaming services (Apple Music, Tidal, Amazon, Spotify, other services as auto-updated) here: https://distrokid.com/hyperfollow/ultraklystron/we-only-have-each-other
Yep, 10 more Dn’B/Jungle Bangers. Lots of fun glitching out the drums. Put a photo of my late cat Hagu on the cover. Miss her.
Comments Off on More Drum & Bass
Collaboration Default Swap (A New Song With Special Guests!)
Posted on August 27th, 2024
Filed under: Music News — Karl Olson @ 10:22 pm
Tired of your favorite businesses, products & services ruining themselves for short term gains, or being bought on credit by shiftless grifters & then ruining themselves to pay that off while handing out golden parachutes to those already, uh, well-gilded? Same! As are funky49 & Nursehella, so here’s a brand new song all about it, available at all of the usual places (including many businesses that themselves are victims/perpetrators of “enshittification,” because we can’t have nice things):
Bandcamp:
Youtube:
Spotify, Amazon, Apple Music/iTunes & everything else are here at hyperfollow.
Comments Off on Collaboration Default Swap (A New Song With Special Guests!)
Embedded Systems (New Music, Retro Merch Drop!?)
Posted on August 10th, 2024
Filed under: Music News — Karl Olson @ 6:00 pm
Another year, another new Dn’B LP, loosely themed around being sick in school & going home early to watch garbage TV. Beats all of the Drum n’ Bass albums I’ve done recently with computer programming names I guess. It’s also on all of the usual places, just check the hyperfollow link here, but here’s the Bandcamp and Youtube embeds:
Meanwhile, after a request from an old friend to reissue the old Ultraklystron “Download A Car” meme shirt from the Animatic LP deluxe pre-orders and one or two live shows, I spruced up my apparently extant and definitely neglected Teespring store. I have everything set to make as little money as Teespring will allow with out fiddling chasing things down to the penny. If they had an easy “whatever current cost is plus IDK 50 cents” setting, I’d have gone for that instead:
Comments Off on Embedded Systems (New Music, Retro Merch Drop!?)
New Music Season, Again
Posted on April 9th, 2024
Filed under: Music News — Karl Olson @ 4:25 pm
Available by 4/24/24 on all the other platforms here: Ultraklystron – Lost Languid Muddle on Spotify, Amazon, Apple Music and more.
Yep, another one.