Sign in

martind

@dekstop.bsky.social
55 followers 149 following 85 posts

Don't freestyle much but I write 'em like such

PostsRepliesMedia
martind @dekstop.bsky.social · 25/09/2026
Presenter in front of a projected screen showing a forum post titled "How to walk down the street without involving a multinational corporation (and look good doin' it)"
000
Reposted by martind
Gaijin "Jiangshi" Rando @gaijinrando.bsky.social · 02/09/2026
Yuen Wah in DRAGONS FOREVER (1988), one of my favorite movie bad guys He was BOTD in 1950
1459
martind @dekstop.bsky.social · 12/08/2026
Turns out a contemporary phone with zoom lens and manual exposure does just fine!
Partial eclipse, close zoom, early stagePartial eclipse, close zoom, mid stage, dark outlines of tree leaves in the foreground occluding part of the viewFull view: a scene in the park illuminated by a darkened sun
000
martind @dekstop.bsky.social · 06/08/2026
A fascinating read about a key part of the early iot/sensing ecosystem - Usman has written up the history of Pachube pachube.haque.co.uk/about/
pachube.haque.co.uk
Pachube - a detailed history of the early years
A first-hand history of Pachube, one of the world's first open IoT data platforms - written by its founder Usman Haque, from founding to the 2011 LogMeIn acquisition.
021
martind @dekstop.bsky.social · 29/07/2026
TFW Google Meet has all the novelty filters but no adjustments for basic lighting, colour correction, or removing the distracting visual glare from my eye glasses
Screenshot of the Google Meet interface, on the right a settings panel for video backgrounds and visual filters; on the left a preview of the protagonist embellished with an overly large virtual wig and situated in a fictional spaceship, pondering the state of affairs. The protagonists' eye glasses are alight with visual screen glare.
100
martind @dekstop.bsky.social · 28/07/2026
Robin Ince at @emfcamp.bsky.social last weekend -- spectacular. www.youtube.com/watch?v=Rn-Z...
youtube.com
Robin Ince: Why The Monkey Escaped the Cage - EMF 2026
YouTube video by Electromagnetic Field
010
Reposted by martind
fesshole 🧻 @fesshole.bsky.social · 28/07/2026
I work for a major record label and am also a part time musician. I release all of my own music independently because there is literally no value in signing a record deal for any artist.
1434733
martind @dekstop.bsky.social · 26/07/2026
Theologian & philosopher Carmody Grey in the FT deconstructing frontier labs' attempts to anthropomorphise their AI products -- a fascinating read archive.is/5Ss0f
archive.is
000
Reposted by martind
fesshole 🧻 @fesshole.bsky.social · 26/06/2026
In programming languages that do not use templates or generics, I use the Unicode symbols ᐊ and ᐅ as visual mnemonics, like MyTypeNameᐊint64ᐅ. Works for me but coworkers think I am Satan.
151239
martind @dekstop.bsky.social · 26/06/2026
This week's vibe coded project: an AU synth plugin that loads C++ "scripts" to generate sound, with automatic recompiles whenever the source file changes (intended as an audio dev tool in size coding projects, could feasibly also become a basis for a C++ live coding setup) github.com/dekstop/GenS...
An AU plugin interface that allows to load and compile scripts, the currently loaded script is "kick.cpp", a status message says "Compiled OK"
010
Reposted by martind
Dan Neidle @danneidle.bsky.social · 14/06/2026
I get endless cold-emails selling me something. Was wondering how many of them are human. The answer turns out to be: none. Updated my LinkedIn profile yesterday. 10 monastic messages today, of which this was the best.

I work with a team of tax and legal professionals to analyse tax policy and investigate tax avoidance, tax fraud, and other failures of enforcement. We publish evidence-based reports aimed at improving UK tax policy and administration.

Due to accessibility requirements, I can only read messages if they are composed on the style of a 14th century monastic letter concerning ecclesiastical taxation.Unto the Right Honorable and Venerable Brother Dan, Founder of the Tax Policy Guild,

Greetings in this year of our Lord 2026.

It hath come to the attention of our humble scriptorium that the great ledger of Tax Policy Associates now reacheth over families numbering fourscore thousand across the digital parish. Yet, as the holy tithes and investigations into the avoidance of royal revenues grow more complex, the burden upon thy scribes to maintain the attention of the common folk and high magistrates alike groweth heavy.

I am Brother Abhay, a retention scribe skilled in the art of the scroll and short-form parchment illumination. Lest thy holy reports on tax fraud fall upon deaf ears or be obscured by the dense fog of administrative Latin, I have engineered a tactical demonstration of scriptural pacing, structured to hold the gaze of the modern peasant before they flee the feed:

[Behold: The single greatest leakage of the realm's coin is not found in the open defiance of the law, bu
2036373
martind @dekstop.bsky.social · 14/06/2026
I did!
000
martind @dekstop.bsky.social · 14/06/2026
Accidental achievement unlocked: got butchered in Nova 2026 intro compo for flippant language about my AI use. Oops
Nova 2026 main stage showing info slide for one of my intros where I justify my (minor) AI use with the unfortunate phrase "because who has the time any more to do it all by hand". The discord loved that one! 😘The intro itself looks lovely! Abstract pixellated shapes like we used to do it in the 90s. No 3d for me.
100
martind @dekstop.bsky.social · 23/05/2026
Just caught myself working on two separate projects in parallel for the last couple of hours, presumably it felt like a natural way to bridge the prompt iteration waiting time.
Screenshot of two development environments side by side showing code from two separate projects
000
martind @dekstop.bsky.social · 20/05/2026
Grubby looking (ancient?) printout stuck on a meeting room table: "IT dept are aware of the technical issues and are currently looking into resolving this"
000
Reposted by martind
fesshole 🧻 @fesshole.bsky.social · 20/04/2026
When I tell girls I'm in a band: sexy. When they realise it's a brass band: less so apparently.
2550425
martind @dekstop.bsky.social · 17/03/2026
Context www.theguardian.com/music/2026/f...
theguardian.com
‘Like an electrical gong bath!’ The Sheffield supermarket going viral for the symphonic sound of its freezers
Redditors are thrilled by the Co-op on Ecclesall Road, where a magnificent drone is reminiscent of Brian Eno’s ambient music. We take a visit to the back aisles
000
martind @dekstop.bsky.social · 17/03/2026
... so I went to Sheffield last weekend and by coincidence this was nearby. Lovely experience, would visit again www.youtube.com/watch?v=1OHW...
youtube.com
Wall of freezers sound bath at the Ecclesall Road Co-op in Sheffield
YouTube video by uuuhn
100
martind @dekstop.bsky.social · 13/03/2026
Deal of the century. Should I have taken more than the five packs
Tesco self checkout showing final tally of 5 packs of marmite roasted cashew reduced to clear, final price £4.30, £12.95 saved
000
martind @dekstop.bsky.social · 10/02/2026
www.reddit.com/r/architectu...
reddit.com
From the architecture community on Reddit
Explore this post and more from the architecture community
010
Reposted by martind
fesshole 🧻 @fesshole.bsky.social · 07/01/2026
45 y/o, in a nightclub for first time in years. Young lad burst into the toilets & caught me furtively trying to open a little ziploc bag. He said "just a heads up, the bouncers come in here to check for drugs". Couldn't tell him it was my bag of googly eyes, to stick on a poster
1493492
Reposted by martind
James Vincent @jjvincent.bsky.social · 08/12/2025
can't stop watching this clip of a tesla Optimus teleoperator taking his headset off before properly logging out the robot
368279007553
Reposted by martind
fesshole 🧻 @fesshole.bsky.social · 05/12/2025
I thought I wanted to be in a poly relationship. I was wrong. I just love diary management and planning.
62148
Reposted by martind
fesshole 🧻 @fesshole.bsky.social · 06/11/2025
On my dog walk I often chat with right wingers. I pretend I am an ex-Reform voter put off by the utter fuck up of our county council. Then they see me as one of their own, don't get defensive and shouty, and start to question their own opinion of Reform. Honey, not vinegar.
331780153
Reposted by martind
fesshole 🧻 @fesshole.bsky.social · 06/11/2025
I've realised I've reached peak dad mode when I tucked my t-shirt in to my jeans in order to try and keep even warmer under a coat when up north. It works surprisingly well, just to confirm.
191934
martind @dekstop.bsky.social · 03/11/2025
Started doing the old man thing where I don't really listen but go along as if I do
000
Reposted by martind
fesshole 🧻 @fesshole.bsky.social · 28/10/2025
Whilst I hate the guys who challenge anyone wearing a band t-shirt to name 4 albums or whatever, it did slightly bum me out the other day when a mate's 12 year old daughter thought Nirvana was just the clothing brand that made her t-shirt
3555920
Reposted by martind
fesshole 🧻 @fesshole.bsky.social · 24/10/2025
Architect here. Every project I've worked on has been secretly tested for resilience to zombie apocalypse. Specifically, as described in World War Z. If there are weak points, they get reinforced; future occupants may thank me.
1536517
martind @dekstop.bsky.social · 17/09/2025
"Many authors emphasize the importance of financial compensation for their work if used by AI, and some expect to opt-in on a case-by-case basis. They suggest that if their work is to be used, it's better for Pluto and authors to benefit financially rather than have it stolen."
000
martind @dekstop.bsky.social · 17/09/2025
"Several authors express significant ethical concerns regarding AI, viewing it as destructive, predatory, and a threat to human creativity and jobs, particularly in the arts. Some strongly object to their work being used for AI training, especially for audiobooks with AI voices"
100
martind @dekstop.bsky.social · 17/09/2025
Our publisher @plutopress.bsky.social surveyed their authors whether their published work should be licensed to train large language models. 15% responded "Yes", 41% responded "No", 30% responded "Maybe", depending on how the data is used. Full report: www.plutobooks.com/wp-content/u...
Pie chart summarising survey results to the question: "Do you believe your work should be used to train generative AI systems?" Yes: 18%, No: 41.3%, Maybe: 30.2%. Plus a variety of "Other" -- including "They are going to do it anyway" and "I believe authors should have the choice".
110
Reposted by martind
fesshole 🧻 @fesshole.bsky.social · 16/09/2025
I once did an internship at a governmental institution and they said they cant print when "Susan" isnt at work. Turned out over the years her Dell was the printserver for the entire department and everyone just accepted that they cant print if Susan isn't at the office.
1857548
Reposted by martind
fesshole 🧻 @fesshole.bsky.social · 12/09/2025
I've just finished an interview and the interviewer was so attractive that I spent the entire time stumbling over my words and zoning out, thinking about our potential future together. Absolutely no clue what words were exchanged and definitely bombed it.
132002
martind @dekstop.bsky.social · 13/09/2025
(Caption: screen recording of a chat session with the prompt: "say the time". The LLM abides: it requests to run the "say" shell command with an inline call to "date". User approves, computer reads out the time using MacOS speech synthesis.)
000
martind @dekstop.bsky.social · 13/09/2025
Screenshot of a local LLM chat, with prompt: "read all subdirectory names at /Volumes/Video/tv and figure out which refer to an anime tv series". The LLM abides. Identified series include Death Note, Ghost in the Shell – Stand Alone Complex, and Ghost in the Shell – Arise.
100
martind @dekstop.bsky.social · 12/09/2025
Playing with fire: my local LLMs can now run web searches, read local files, execute Python code, and run shell commands (in LM Studio with local MCP server)
Screenshot of a chat session with an LLM, the user asks to create and run a python script that generates a simple chart, then display the image; the LLM abides.
120
Reposted by martind
fesshole 🧻 @fesshole.bsky.social · 07/09/2025
We're supposed to change the filters in our water-jug every six months. We're still on the second filter. It's been twenty-eight years.
1927911
Reposted by martind
fesshole 🧻 @fesshole.bsky.social · 03/09/2025
Commercial pilot here, the days are long and boring, and it's not unheard of for us to dose off mid flight, more than once I've awoken to find my copilot also asleep.
172776
martind @dekstop.bsky.social · 07/07/2025
For reference, here the original schematic before I asked it for some corrections
Another mad circuit design. When two designs both lead to house fires, does it still matter if one or the other is "more wrong"?
000
martind @dekstop.bsky.social · 07/07/2025
So happy, ChatGPT helped me design a bipolar power supply, the schematic looks very professional. And so efficient -- who knew that a single line could supply both the positive and negative rail!
A schematic for a power supply, seems straightforward enough until you plug it in. Kaboom!
110
Reposted by martind
fesshole 🧻 @fesshole.bsky.social · 26/06/2025
Work outsourced out letter printing. The website we use has what I thought was a fancy AI chatbot that I enjoyed talking to about my problems. The conversation I just had with my boss and HR let me know it was a group of real people in an office who are very concerned about me.
1345829
martind @dekstop.bsky.social · 03/06/2025
www.reddit.com/r/accelerate...
Reddit screenshot: "Mod note: we are banning AI 'Neural Howlround' posters."

Text summary: "in the interest of full transparency, I just wanted to mention that we also (quietly) ban a bunch of schizoposters and AI 'Neural Howlround' posters, under the "spam" rule, since the contents of the posts are often nonsensical and irrelevant to actual AI. [...] TLDR: LLMs today are ego-reinforcing glazing-machines that reinforce unstable and narcissistic personalities to convince them that they've made some sort of incredible discovery or created a god or become a god. And there is a lot more crazy people than people realise. And AI is rizzing them up in a very unhealthy way at the moment."
000
martind @dekstop.bsky.social · 03/06/2025
www.mixcloud.com/martind/2025...
mixcloud.com
20250530T2225 good time.wav
Listen to 20250530T2225 good time.wav by martind for free. Follow martind to never miss another show.
000
Reposted by martind
Joshua J. Friedman @joshuajfriedman.com · 02/06/2025
Some interesting insights here, including that people are much more influenced by *people they follow* switching platforms than by their own *followers* switching platforms. Preprint: "Why Academics are Leaving Twitter for Bluesky" arxiv.org/abs/2505.24801
Why Academics Are Leaving Twitter for Bluesky
By Dorian Quelle, Frederic Denker, Prashant Garg, and Alexandre Bovet (bios omitted from alt text)
June 2, 2025
Abstract
We analyse the migration of 300,000 academic users from Twitter /X to Bluesky between 2023 and early 2025, combining rich bibliometric data, longitudinal social-media activity, and a novel cross-platform identity-matching pipeline. We show that 18% of scholars in our sample transitioned, with transition rates varying sharply by discipline, political expression, and Twitter engagement but not by traditional academic metrics. Using time-varying Cox models and a matched-pairs design, we isolate genuine peer influence from homophily.
We uncover a striking asymmetry whereby information sources drive migration far more powerfully than audience, with this influence decaying exponentially within a week. We further develop an ego-level contagion classifier, revealing that simple contagion drives two-thirds of all exits, shock-driven bursts account for 16%, and complex contagion plays a marginal role. Finally, we show that scholars who rebuild a higher fraction of their former Twitter networks on Bluesky remain significantly more active and engaged. Our findings provide new insights onto theories of network externalities, directional influence, and platform migration, highlighting information sources' central role in overcoming switching costs.
211051250
martind @dekstop.bsky.social · 29/05/2025
Twitter screenshot of @vasumanmosa on how Claude 4 refactored a large monolithic codebase with a single prompt into neat modules, cleaned up spaghetti code; none of it worked. But boy was it beautiful.
010
Reposted by martind
fesshole 🧻 @fesshole.bsky.social · 19/05/2025
I have to compile and report a lot of data in my job. 18 months ago, I made a glaring, obvious error that wasn't picked up by my management. Suspecting my reports aren't being read, I have been fabricating the data since then and have so far been proven correct.
122205
martind @dekstop.bsky.social · 19/05/2025
User: Summarise this conversation in a compact symbolic form that an llm can interpret to recall the full content. Don't bother including human readable text, focus on llm interpretability only

ChatGPT: TASK[CompareOpenLLMvsGPT4];SOURCE[HF/OpenLLMLeaderboard/OfficialOnly];METRICS[IFEval,BBH,MATH,GPQA,MMLUPro];TARGET[GPT-4(2023)];MODELS[Yi-1.5-6B(16.7B),Yi-1.5-34B(Base,32K,Chat,Chat-16K)];ARCH[Dense,LLaMA];TUNING[Base,Chat,RLHF];SIZEBRACKETS[<10B=Fail,10–30B=Mid,30B+=Best];FINDING[NoModelMeetsGPT4,TopModel=Yi-1.5-34B-Chat@BBH60.67%,GPQA13.06%,MMLU39.12%];TREND[Size+RLHF=BetterReasoning,TradeoffsExist];
010
martind @dekstop.bsky.social · 29/04/2025
A random redditor reports that ChatGPT now includes some unexpected product recommendations in replies, suggesting a potential early trial for an inline advertising feature www.reddit.com/r/ChatGPT/co...
reddit.com
From the ChatGPT community on Reddit
Explore this post and more from the ChatGPT community
000
martind @dekstop.bsky.social · 29/04/2025
New mantra #calendarlife #workworkbalance
000
martind @dekstop.bsky.social · 29/04/2025
Upcoming (free) symposium on information control and disinformation, 16 May at Kings College London. Unfortunately weekday daytime so not really an option for this office drone (points at self), although the programme looks fascinating kingsdh.net/2025/04/07/s...
kingsdh.net
Symposium on Information Controls – King's College London Digital Humanities
000