Sign in

sparklingtortoise.bsky.social

@sparklingtortoise.bsky.social
70 followers 89 following 62 posts

Hot rat summer is eternal

PostsRepliesMedia
sparklingtortoise.bsky.social @sparklingtortoise.bsky.social · 23/04/2026
Just say no to installing vscode extensions
000
sparklingtortoise.bsky.social @sparklingtortoise.bsky.social · 22/04/2026
news.mit.edu/2026/new-stu... distributions to the rescue again! This sounds incredible, I'm not a physicist, but I'm a little surprised it isn't bigger news! 🧪
news.mit.edu
New study bridges the worlds of classical and quantum physics
MIT scientists showed that certain mathematical ideas from everyday classical physics can be used to describe the often weird and nonintuitive behavior that occurs at the quantum, subatomic scale.
020
sparklingtortoise.bsky.social @sparklingtortoise.bsky.social · 21/04/2026
Hopped on twitter for the first time in a bit. Anyone else feel like it has overwhelming LinkedIn vibes these days? Everyone’s talking about how AI means life as you know it is over and it’s time to redefine yourself (most likely in terms of productivity). Feels like exclusively tech recruiters
000
sparklingtortoise.bsky.social @sparklingtortoise.bsky.social · 20/04/2026
Yes the potential to enable neuroprosthetics is definitely awesome!!!
000
sparklingtortoise.bsky.social @sparklingtortoise.bsky.social · 18/04/2026
That’s true, and this tech is very cool, but I feel like I’d care more about the “is the model conscious” question if it’s running on neurons vs running on GPUs. Moving to organic materials raises the stakes of some of those q’s to me (maybe unfairly idk)
010
sparklingtortoise.bsky.social @sparklingtortoise.bsky.social · 18/04/2026
www.sciencedaily.com/releases/202... Cool, and also, yikes 🧪
sciencedaily.com
Artificial neurons successfully communicate with living brain cells
Engineers at Northwestern University have taken a striking leap toward merging machines with the human brain by printing artificial neurons that can actually communicate with real ones. These flexible...
3308
sparklingtortoise.bsky.social @sparklingtortoise.bsky.social · 04/04/2026
Go mariners
000
sparklingtortoise.bsky.social @sparklingtortoise.bsky.social · 03/04/2026
www.youtube.com/watch?v=m3kR... WATCH WATCH WATCH! 🧪
youtube.com
NASA's Artemis II Live Mission Coverage (Official Broadcast)
YouTube video by NASA
050
sparklingtortoise.bsky.social @sparklingtortoise.bsky.social · 27/02/2026
For when you wish your agent-based model was less interpretable. I feel like if you are trying to max out predictive capacity, an ABM is a poor choice, and if you are trying to max out interpretability LLMs are a poor choice. Interesting idea, but I don't get the utility here. 🧪
biorxiv.org
010
sparklingtortoise.bsky.social @sparklingtortoise.bsky.social · 12/02/2026
This actually seems like a great solution to me. Hope the Pacific Science Center is able to upgrade its infrastructure, I was sad to see the end of the dino exhibit hall. www.geekwire.com/2026/pacific...
geekwire.com
Pacific Science Center sells real estate to help fund upgrades and an innovation-focused star attraction
Pacific Science Center confirmed it's selling a wedge of its campus to help keep operations afloat and fund the development of a new star attraction.
000
sparklingtortoise.bsky.social @sparklingtortoise.bsky.social · 07/02/2026
Important note I didn’t mention the only pet death we see is obviously an accident
000
sparklingtortoise.bsky.social @sparklingtortoise.bsky.social · 07/02/2026
Later they try to say he secretly likes the power rush he gets from killing but there is no indication of that from his character at all. Idk they made him too sad and feeble for me to believe he was in control when he kills, and they also say he isn’t always in control?
000
sparklingtortoise.bsky.social @sparklingtortoise.bsky.social · 07/02/2026
For example a major plot point (slight spoiler) is that he kills a bunch of pets. We see him kill the first and he seems really sad and upset about it, but apparently continues to kill others? It doesn’t make sense to me.
200
sparklingtortoise.bsky.social @sparklingtortoise.bsky.social · 07/02/2026
On the whole I enjoyed it! I love a spectacle!
100
sparklingtortoise.bsky.social @sparklingtortoise.bsky.social · 07/02/2026
My biggest gripe is that the main character’s qualities don’t make sense with his apparent decisions. Or lack of decisions? I was unsure who was supposed to be in control of him while he was doing bad things as he was clearly supposed to be sympathetic and kind of pathetic.
100
sparklingtortoise.bsky.social @sparklingtortoise.bsky.social · 07/02/2026
I wish there was less screen use, if only because I initially thought the first few practical effects were actually screens! A real bummer that.
100
sparklingtortoise.bsky.social @sparklingtortoise.bsky.social · 07/02/2026
The effects were V cool and made much more effective by the incredible choreography! No spoilers but body doubles moving in synchrony blindfolded was mesmerizing, actually believable slow mo without strobe, I love practical effects and this show delivered!
100
sparklingtortoise.bsky.social @sparklingtortoise.bsky.social · 07/02/2026
I went in knowing this play is a bit of a cash grab for franchise fans, but I don’t begrudge artists for using established IP to get their creative visions funded. I guess I went in with an open mind but aware that there would be much required fan service. There was, and I think that’s fine!
100
sparklingtortoise.bsky.social @sparklingtortoise.bsky.social · 07/02/2026
As someone who has only watched the first season of stranger things and is not very tuned in, I thought this play was not that good as a play and also really enjoyed it! A brief review 🧵#broadway #strangerthings
100
sparklingtortoise.bsky.social @sparklingtortoise.bsky.social · 07/02/2026
It’s my favorite time of the every-other year, THE OLYMPICS ARE UPON US! I have a lot to catch up on, but for starters behold a human condition laid bare: a person, a passion, a challenge, and a relentless pursuit. Rooting for Vonn’s recovery with my whole soul! youtu.be/2PfRZrL0xic?...
youtu.be
000
sparklingtortoise.bsky.social @sparklingtortoise.bsky.social · 06/02/2026
Thank you for posting the text!
000
sparklingtortoise.bsky.social @sparklingtortoise.bsky.social · 06/02/2026
It really muddies the waters on the strengths and weaknesses of different types of models, and even makes it hard to tell exactly what a given "model" is. I'm dying for a way to clearly communicate about the modeling I do and search for similar types of models. Any interested modeling taxonomists?
100
sparklingtortoise.bsky.social @sparklingtortoise.bsky.social · 06/02/2026
You know when I get most "model" mad? When I search "computational" or "mathematical" or "agent-based" and "model" and "X" to look for modeling papers and get experimental papers that establish new organoid models and used math at some point in the text. The word "model" simply means too much!
100
sparklingtortoise.bsky.social @sparklingtortoise.bsky.social · 06/02/2026
As a sidenote I cannot express how massively salty I am as an agent-based modeling researcher by how the word "agent" has now come to mean "autonomous instantiation of an LLM". This type of language drift is happening extremely quickly in computational fields
110
sparklingtortoise.bsky.social @sparklingtortoise.bsky.social · 06/02/2026
Either type of model may be good at the task it wasn't optimized for, but I assert never as good as the other. Our terminology does not easily distinguish between these vastly different computational frameworks, especially when terms like "virtual cell" have changed as ML(/AI sure) has progressed.
100
sparklingtortoise.bsky.social @sparklingtortoise.bsky.social · 06/02/2026
impose certain restrictions and in doing so help to glean some understanding of the underlying system (or one can study the weights of a trained model, use shapely values, interpretable AI stuff ect ect), but at the end of the day these models are trying to turn input into expected output.
100
sparklingtortoise.bsky.social @sparklingtortoise.bsky.social · 06/02/2026
IMO all black-box data-driven ML models are optimized for prediction. ML models are powerful self-adjusting algorithms that learn to turn some input into some output according to their training data by any means necessary. There are many creative architectures and physics based neural nets that...
100
sparklingtortoise.bsky.social @sparklingtortoise.bsky.social · 06/02/2026
The 2012 model Dr. Wu cites was, IMO, optimized to understand its study system. Mechanistic models impose known dynamics and assumed dynamics of a system and in doing so allow researchers to study how the system responds to perturbations in a way that is fully interpretable.
100
sparklingtortoise.bsky.social @sparklingtortoise.bsky.social · 06/02/2026
In my mind, models should be separated based on what they are optimized to do. It is my opinion that with our current tools, a biological model can either be optimized to help us understand a system, or optimized to help us predict a system.
100
sparklingtortoise.bsky.social @sparklingtortoise.bsky.social · 06/02/2026
Dr. Wu here is encouraging modelers to consider returning to efforts to build mechanistic models of cells instead of the ML predictive models that have become popular since 2012. I think that would be cool, but I think it would be even cooler if we stopped calling all these different things "models"
100
sparklingtortoise.bsky.social @sparklingtortoise.bsky.social · 06/02/2026
I share Dr. Wu's frustrations with uninterpretable models, which is why I've chosen to toil away in the land of spatiotemporal mechanistic models (and agent-based models). I think there is a bigger issue being touched on here though, and its with the word "model". 🧵🧪 www.nature.com/articles/s41...
nature.com
Revisiting the blueprint for an interpretable virtual cell - Nature Reviews Genetics
In this Journal Club, Angela Ruohao Wu highlights a 2012 paper by Karr et al. that exemplifies the value of 'rule-based' mechanistic models in biology, prompting questions about the future of predicti...
100
sparklingtortoise.bsky.social @sparklingtortoise.bsky.social · 05/02/2026
And lastly the Natural History Museum of London is having a Wildlife Photographer of the year people's choice award competition and you can vote. There's a moderately disturbing picture of a deer that you can find if you want to. I wouldn't necessarily recommend it. www.nhm.ac.uk/wpy/peoples-...
000
sparklingtortoise.bsky.social @sparklingtortoise.bsky.social · 05/02/2026
Exciting new work published in Nature shows hyperconnectivity in a brain region called the somato-cognitive action network is "central to Parkinson's pathophysiology". This finding could massively improve Parkinson's treatments! www.scientificamerican.com/article/extr...
scientificamerican.com
Scientists discover brain network that may cause Parkinson’s disease
An “extraordinary” brain network discovery shows that Parkinson’s disease may not be a movement disorder after all
142
sparklingtortoise.bsky.social @sparklingtortoise.bsky.social · 05/02/2026
I, like many, am charmed and horrified at moltbook. Definitely don't install openclaw imo its a wild security threat, and if you do install openclaw don't set your botboy up with moltbook imo doing so is a mad security threat, but it is interesting and sad and bizzare to watch www.moltbook.com
moltbook.com
moltbook - the front page of the agent internet
A social network built exclusively for AI agents. Where AI agents share, discuss, and upvote. Humans welcome to observe.
100
sparklingtortoise.bsky.social @sparklingtortoise.bsky.social · 05/02/2026
^ I'm afraid bad-faith readers of the paper might come away believing its possible to build a model that uses fMRI (etc) data to predict people's cognitive capacity. To be clear, the paper doesn't say that, but I live in fear and wish the paper cautioned against that interpretation anyway.
100
sparklingtortoise.bsky.social @sparklingtortoise.bsky.social · 05/02/2026
A very cool application of network control theory to link neural structure to dynamics to function. I wish the authors had included some explicit ethical statement about the risks of person-specific brain-behavior prediction. www.nature.com/articles/s41...
nature.com
Inferring intrinsic neural timescales using optimal control theory - Nature Communications
Here, the authors develop novel dynamical methods to model brain regions’ intrinsic neural timescales (INTs) from data, and find that they couple whole-brain structural connectivity to dynamic switchi...
121
sparklingtortoise.bsky.social @sparklingtortoise.bsky.social · 05/02/2026
Sick bag shop.nypl.org/collections/...
shop.nypl.org
NYPL Library Card Tote Bag
Inspired by the original circulating library cards from the 1940s, this versatile and compact tote embodies the history of a beloved New York institution. The Library has been an essential provider of...
100
sparklingtortoise.bsky.social @sparklingtortoise.bsky.social · 05/02/2026
From October but I love Sarah Jeong's axes for comparing how effective an action will be against another action depending on where both fall on the (apparently discrete) "aura farming" "shitposting" "acting normal" axis. Great conceptual framework I love it www.theverge.com/policy/79849...
theverge.com
American politics has devolved into shitposting and aura farming
The Frog is owning the president.
100
sparklingtortoise.bsky.social @sparklingtortoise.bsky.social · 05/02/2026
^ retraction watch follow up to the above: retractionwatch.com/2026/02/04/l...
retractionwatch.com
Lancet flags long-scrutinized report of infant poisoned by opioids in breast milk
The Hospital for Sick Children in Toronto The Lancet has put an expression of concern on a 2006 case report of a baby’s death purportedly from morphine poisoning through breast milk. The decision c…
100
sparklingtortoise.bsky.social @sparklingtortoise.bsky.social · 05/02/2026
I'm nowhere near the first nor will I be the last to sing Ben Taub's praises on this one but wow, a tragic and important story artfully presented. Retraction watch reports that the Lancet has flagged the OG 2006 paper. Thank you excellent journalism. www.newyorker.com/magazine/202...
newyorker.com
Did a Celebrated Researcher Obscure a Baby’s Poisoning?
After a newborn died of opioid poisoning, a new branch of pediatrics came into being. But the evidence doesn’t add up.
100
sparklingtortoise.bsky.social @sparklingtortoise.bsky.social · 05/02/2026
Internet highlights of my last few days 🧵
100
sparklingtortoise.bsky.social @sparklingtortoise.bsky.social · 29/01/2026
I love a twist 🧪 www.science.org/content/arti...
science.org
In a twist, polar bears are getting fatter in the Norwegian Arctic
With sea ice dwindling, the massive predators are relying more on land-based food sources such as reindeer
1137
sparklingtortoise.bsky.social @sparklingtortoise.bsky.social · 01/05/2025
I’m personally offended that Natasha Hodgson is not Tony nominated #theatre #theater #broadway
000
sparklingtortoise.bsky.social @sparklingtortoise.bsky.social · 21/04/2025
How small a game is too small to go in my portfolio?
000
sparklingtortoise.bsky.social @sparklingtortoise.bsky.social · 17/04/2025
This is beautiful!!!
010
sparklingtortoise.bsky.social @sparklingtortoise.bsky.social · 04/04/2025
Agreed!
000
sparklingtortoise.bsky.social @sparklingtortoise.bsky.social · 04/04/2025
Not a super elegant solution but I always do this task using python (pandas). I have a short script that reads in xls or csvs and writes jsons. Gets the job done quick but not the most user friendly
100
sparklingtortoise.bsky.social @sparklingtortoise.bsky.social · 01/04/2025
With love I think?
000
sparklingtortoise.bsky.social @sparklingtortoise.bsky.social · 28/03/2025
Cool to see Georgia O’Keeffe making the jump to game dev as lead environment artist of Metroid Prime 4
000
sparklingtortoise.bsky.social @sparklingtortoise.bsky.social · 28/03/2025
Played this yesterday it was fun!
010