phiofx

@phiofx@hachyderm.io

Φ(χ). A unknown function of an unknown argument.

Microblogging about #opensource, #generativeart, #python, #datascience, #physics, #sustainability, #musictheory, #solarpunk, the #fediverse and random combinations thereof.

This profile is from a federated server and may be incomplete. Browse more on the original instance.

kellogh, to random
@kellogh@hachyderm.io avatar

e-bikes are urban 4-wheelers

phiofx,

@kellogh thinking selfishly I would agree (I live in Amsterdam, I don't want them ruining my bike paths) but thinking big about how to save the human race (one of my favorite pastimes), e-bikes are about as good as it can get.

They vastly expand the scope of low impact mobility. A global replace of cars for e-bikes within cities is one of the few low hanging fruits on the journey

stefano, to fediverse
@stefano@bsd.cafe avatar

Big Fedi, Small Fedi

https://evanp.me/2023/12/26/big-fedi-small-fedi/

phiofx,

@stefano

I dont think we have much choice on the matter. The must grow at least to a point it is long term sustainable.

Eking a niche safe space in the shadows of abusive might seem viable right now, but will it be tomorrow or in a few years as technology evolves and creates ever more means of control and abuse?

An , human-centric, connected online world needs to have critical mass and become a shaper of developments, not a passive response.

phiofx,

@smallcircles @stefano the unknown factor is the degree and speed of diffusion. Corporate control of tech follows it being the only economic pattern that cajoled enough brains to tap into its potential. The public sector doing the same is (maybe) the next phase. Yet individual digital sovereignty through tech literacy is the noblest destiny of the fediverse. Should be (imho) an explicit human right / UN as this will shape our future more than anything else.

J12t, to machinelearning
@J12t@social.coop avatar

What's the best approach to follow research in a particular sub-field of / ? Other than the firehose that is arxiv.org?

phiofx,

@J12t

You can step back from the hyperventilation and check once a year the neurips proceedings

https://proceedings.neurips.cc/

LordCaramac, to Lisp German
@LordCaramac@discordian.social avatar

It's a shame almost nobody uses or anymore,
https://www.theregister.com/2023/12/25/the_war_of_the_workstations/

phiofx,

@LordCaramac I have started going down the rabbit hole of algorithmic composition and it turns out the mini-language is modeled on (which I knew nothing about!)

Commercialization of technology in the last decades by-and-large meant dumbification - creating the widest possible addressable market (see idiot "smart"phone devices).

As that approach saturates it is plausible that people will start revisiting useful patterns that have been steamrolled into oblivion

thomasfuchs, to random
@thomasfuchs@hachyderm.io avatar

Few people know this, but USB is the successor of USA

phiofx,

@thomasfuchs alas almost completely dominated these days by USD

dansup, to Pixelfed
@dansup@mastodon.social avatar

After some careful consideration, I have decided to block threads.net on pixelfed.social and .art by default

However, users will have the ability to unblock the domain

Soon we will be selectively enforcing authorized fetch for accounts with domain blocks so as to provide the best of both worlds.

(I'm also shipping a command for :pixelfed: admins to easily add user domain blocks for all local users)

I'm eager to hear your feedback!

PR: https://github.com/pixelfed/pixelfed/pull/4834

#pixelfed #userDomainBlocks

phiofx,

@berkes

This. As an example of the broader collective intelligence and immune system that hopefully will develop organically within the fedi. With developers shaping is software dna and admins, moderators and maybe even new functions being vigilant about what is happening.

@dansup

phiofx,

@berkes I am convinced that if people keep building in line with the human-centric values of the fediverse there is nothing to worry about (no real threats from threads). Adtech based of profiling user behaviors has been an all dominating force not just in social media but almost all personal computing experience, but it has saturated and is end-of-life because thats how capitalism works: no growth = death.

Meta/Threads wants to cannibalize Twitter? Be my guest.

@dansup

phiofx, to random

@smallcircles @astrojuanlu @daniel

They could indeed hijack the technical developments. When you print billions per quarter you can attract serious technical talent (though obviously low on moral values). Open source projects like react for frontend or pytorch for ML have redefined the respective domains.

But their commercial interest is fundamentally misaligned with a (re)democratized web. Not clear how they could extract rents if they support building the tools that empower people not to.

juergen_hubert, to threads
@juergen_hubert@thefolklore.cafe avatar

My two cents on the situation:

  1. Yes, people are right to be wary of Meta, but 2) their adaption of is still a net positive for the .

We know of Meta's blatant data mining, their abominable moderation, and their misuse of their algorithms to fan the flames of hatred. Thus, it's perfectly reasonable to defederate them.

However, plenty of large news and other organizations are in a bind - they need large social media reach for their influence, subscribers, and customer base. And the Fediverse itself is basically a rounding error when it comes to worldwide social media reach.

Threads, on the other hand, is significantly bigger. Thus, it's worth for these organizations to invest some effort into it. But it is by no means guaranteed that Threads will survive - right now, it's one of the bigger players in the fight for the Twitter successors, but we don't know if this will last.

But if Threads ends up fully compatible with ActivityPub, these organizations can set up their own ActivityPub servers and get the best of both world - gain reach and followers on Threads, while remaining independent enough to remain standing if Threads folds.

And other large organizations adapting ActivityPub - such as the recent announcement - should likewise be seen in that vein. They are building an ActivityPub-based network that allows them to remain independent - yet be connected to large platforms such as Threads who are compatible with it as well.

So if you want to defederate Threads, do so - there's certainly enough reason to be leery of them. But there are use cases where federating with Threads makes sense. And these will help make the Fediverse grow.

phiofx,

@juergen_hubert Meta has no incentive to remove the lock-in of their paying clients beyond what might be stipulated by law.

The reasons for them "embracing" the fediverse are thus likely a complex long game.

Still, my feeling is that the fediverse will win not as a social media platform but because it is a fundamentally different proposition, it is a web 3.0 of sorts.

Hopefully this "attention" will be used in a positive way, to get to the next level.

phiofx, to Wikipedia

has changed the world and is written in (at that point of "fractal of bad design" fame).

has changed the world and is written in .

is on its way for similar impact.

The choice of tech platform doesnt seem to have much correlation with whether a project will have impact.

There is an abundance of practically equivalent ways of implementing the same old ideas but a scarcity of really good new ideas...

phiofx, to ai

On the occasion of the fiasco Simon Willison raises the debilitating issue of (lack of) in the so-called industry.

I think he is being naive about the amount of spying that goes on on mobile. Maybe Meta doesnt do it in the in-your-face way others do, but most people have dozens of popular apps from various operators (I have witnessed first hand live conversation showing up as ads minutes later).

The "tech" industry right now is a moral cesspool.

https://simonwillison.net/2023/Dec/14/ai-trust-crisis/

xahteiwi, to markdown
@xahteiwi@mastodon.social avatar

Hivemind, please help.

One thing that still eludes me is a way to lint files to enforce a one-sentence-per-line rule. This is remarkably nontrivial.

Can somebody point me to such a thing? Ideally in , but I'll take any language (even an emacs minor mode will help).

A bajillion bonus points if the thing can also unwrap wrapped paragraphs and automatically turn them into one sentence per line.

Boosts appreciated, thanks!

https://sive.rs/1s

phiofx,

@xahteiwi how about simply replace . with .\n (newline)? This will break some edge cases, e.g., e.g. but visible enough to fix. Alternatively look for a more complex regex that escapes and does not replace full stops that are not followed by a space. Implement as per you favorite editor instructions.

Jetbrains IDE's provide this macro out-of-the box but I only find it useful at the very early stage. I have to put the sentences back together to check whether they read fluently.

hrefna, (edited ) to random
@hrefna@hachyderm.io avatar

I have maintained, and continue to maintain, that my concerns about and the have nothing to do with threads as such and everything to do with how they act around the protocol.

Them releasing a client or a server would be very, very bad news in this space. Them running a multi-tenant (or multi-single-tenant) setup would be very, very bad for the fediverse.

Their presence on a single domain just doesn't say or do all that much in the grand scheme of things.

phiofx,

@hrefna my two cents: chasing the dwindling former twitter addict pool (which seems to be the meta/threads objective) is not the future of the .

Its unique appealing proposition is not social media as we have come to dislike, but it is the federation of very diverse types of servers where people create real and lasting value. Think e.g., all these important communities.

Imho the protocol(s) should leapfrog to the desired end state and leave the mastodon era behind.

phiofx, to cpp

Wow, TIL that , the powerful library for all sorts of numerical computing started as a project to support visualization. Eigen is now a building block in various efforts.

The magic of development is that pieces evolve, recombine and are re-used in novel contexts, almost like a biological process.

seav, to Philippines
@seav@en.osm.town avatar

While doing some editing in , I stumbled onto this dramatic transformation at Domoroy Point in Barangay Matayum, Cataingan, Masbate, 🇵🇭. In a bid to boost tourism, the local government created the huge man-made Matayum Lagoon.

You can see the transformation by switching to different aerial imagery layers. Mapbox at the top shows the area before the lagoon was created, the middle from Bing with the initial excavation, and the bottom from Esri with the finished lagoon.

phiofx,

@seav is there something like the "delta" of , a way to identify major real changes in the underlying topography (not corrections / additions).

With the usage of earlier satellite imagery one could image a stacked chronological sequence that documents the changing landscape.

sb, to python
@sb@fed.sbcloud.cc avatar

I'm several decades familiar with many, many programming languages, but have about a half-dozen that I'd consider daily driver languages. is not one of them.

I'm writing a lil' thing in , and have to keep looking up what modules to include for basic things... and every single time, the answer is exactly what I would have intuited.

"What would I call my datetime module if I were me? I'd call it datetime." It's datetime.

"If I wanted today's date, I'd want 'date.today()'". Yup.

I guess that's why it's so widely recommended as a starter language. Not starter because it's lacks advanced features, but because it lacks unnecessary confusion!

phiofx,

@sb Its now mostly a distant memory but the first impression of after a lifetime of c++ etc. still lingers in my mind.

It was: wow, is it legal to have such an easy time, and dare I say fun, getting things done? :-)

recursive, to random
@recursive@hachyderm.io avatar

Nothing quite taught me the deadly seriousness of forgetting history like the first time I read @cstross 's Glasshouse, which has the quote "Who still talks nowadays about the Armenians?" (Hitler, 1939) right before the novel starts.

I'd been doing California tax filings for a while and had noticed https://gizmodo.com/the-dark-truth-behind-californias-strangest-tax-exempti-1697921957 but never thought much of it.

And this is how I didn't learn about the Armenian genocide until I was 28 years old.

(re: https://infosec.exchange/@lcamtuf/111552632921613751 )

phiofx,

@recursive i dont think evolution granted us many tools to handle large scale societal structures. We went from small clans to millions and now billions in a few thousand years.

In this sense its a miracle that systemic evil is not more common. Maybe it hints that our social nature is more potent than we dare to admit.

But indeed there is so much cultural invention still to be done to enshrine and anchor our good side. There much needless, gratuitous stress that is a leftover from the past.

phiofx, to sustainability

For the entire period we call "History" humans lived unsustainably.

Good example is , an ongoing project since 3000 BC. Its only the rate that accelerated in the industrial era.

While people are not "anti-nature", living in equilibrium with it is not deeply and irrevocably embedded in cultures and economies.

is not a return to something, its something we need to invent: an exciting, beautiful, future. Unlocking a new level.

https://en.wikipedia.org/wiki/File%3A8000_BCE%2B_Loss_of_forest_and_grassland_to_grazing_and_crops.svg

Benhm3, to random
@Benhm3@mastodon.social avatar

@pluralistic has a way with words:

"This "AI debate" is pretty stupid, proceeding as it does from the foregone conclusion that adding compute power and data to the next-word-predictor program will eventually create a conscious being, which will then inevitably become a superbeing. This is a proposition akin to the idea that if we keep breeding faster and faster horses, we'll get a locomotive:"

phiofx,

@Benhm3 @pluralistic

there is a cultish expectation of a Messiah-like emergence of the horse-to-locomotive miracle that is underpinned by "the bitter lesson" theory of AI development [1]

tl;dr it states that "the only thing that matters in the long run is the leveraging of computation"

An ex-post denigration of older, logic focused, AI paradigms once it was clear that supersizing certain models gave unexpectedly interesting results.

[1] http://www.incompleteideas.net/IncIdeas/BitterLesson.html

jonny, to ai
@jonny@neuromatch.social avatar

Molly White is right as usual: "We’ve already tried out having a tech industry led by a bunch of techno-utopianists and those who think they can reduce everything to markets and equations. Let’s try something new, and not just give new names to the old."

trying to articulate new ideologies for computing is where my mind has been at the last few years too. i joke about the 'anti-perf manifesto,' but forging imaginaries that can run on computers that are actively antagonistic to the techno-utopians is all about killing myths of heroism where we are the someone else who goes out and "brings home the spoils." how do we reach a computing that isn't foundationally based on asymmetric power, we serfs at the mercy of the lord of the platform and vice versa, we altrustic platform providers building things the commoners couldn't possibly understand. The language of "scale" where one or a few services need to expand to provide for millions hides futures where we can provide for each other horizontally in overlapping quilts of dozens, hundreds. You could shorthand the "" boom as the continuation of the information conglomerates trying to provide the everything platform, and if our dreams are to meaningfully challenge theirs we can't also aspire to simply "do what they're doing, except it's us doing it."

I tried to articulate this as the cloud orthodoxy vs. a still-nebulous idea i've landed on as vulgarity in computing, but i'll probably be orbiting this idea for as long as i am on line.

re: @molly0xfff
https://hachyderm.io/@molly0xfff/111475137431905986
and
https://newsletter.mollywhite.net/p/effective-obfuscation

The world is asymmetrical and hierarchical. I am a consumer, a user and I trade my power to a developer or platform owner in exchange for convenience. The purpose of the internet is for platform holders to provide services to users. As a user I have a right to speak with the manager, but do not have a right to decide which services are provided or how. As a platform owner I have a right to demand whatever the users will give me in exchange for my services. Services are rented or given away freely56 rather than sold because to the user the product is convenience rather than software. Powerlessness is a feature: users don’t need to learn anything, and platform owners can freely experiment on users to optimize their experience without their knowledge. Information is asymmetrical in multiple ways: platforms collect and hold more information than the users can have and parcel it back out as services. But also, platform holders are the only ones who know how to create their services, and so they are responsible for the convenience prescribed for a platform but not the convenience of users understanding how to make the platform themselves.
Our infrastructures are social. There is no class distinction between “developer” and “user.” We resist concentrated power in favor of mutual empowerment. We don’t seek to cultivate dependence in councils of elders or create new chokepoints of control. Anything worth making is a potential source of power, so anything worth making is worth distributing governance of. We don’t assume the needs of others, but make tools to empower everyone to meet their own needs. We don’t make platforms, we make protocols with rough consensus based on what works. We are autonomous, but neither isolated nor selfish. Our dream is not one of solipsism, glued to our feed, being stuffed with the pellets of our social reality. We are radically responsible for one another, and by organizing together we can provide services as mutual aid. Mutual empowerment means that we are free to come and go as we please, even if we might be missed. We have no love for venerated institutions and organize fluidly, making systems so we can merge and fork105 code and ourselves freely [223, 224].

phiofx,

@jonny language is so multipurpose there are invariably tradeoffs and choices to make in conformity vs freedom dial.

Imagine a scientific or legal context where the speaker or author takes liberties with the naming of things in order to express creatively. Unlikely to be appreciated by their audience. The extreme limit of this: mathematical notation/text.

This struggle plays out here in our very in how people use and abuse hashtags. Are they supposed to be a controlled vocabulary?

phiofx, to fediverse

Some confused and naive human made the terrible mistake (i suppose around 10000 BC) of domesticating some feline species or other and now the has to endure on an recurring seven day pattern.


HeavenlyPossum, to random
@HeavenlyPossum@kolektiva.social avatar

In my town, there is a medieval hospital—not for treating the sick, but for housing the indigent.

It was built in the 15th century and it still serves to house the poor. A wealthy merchant endowed it.

I don’t know why he did it, though. Maybe he was motivated by genuine concern for the poor. Maybe he was hoping to save his immortal soul through charitable works. Maybe he wanted to impress a romantic interest or embarrass a rival merchant. Maybe it was a combination of these things or something else entirely.

Human motivation is, despite what you’ve undoubtedly heard, enormously complex, multifaceted, and often very opaque (even for ourselves and our own motivation).

The fact remains that this merchant did something I’ve been told is impossible, could never happen: he did something for nothing.

1/

phiofx,

@HeavenlyPossum

in my mind we are collectively very naive and oversimplifying our own social cooperation / competition behaviors, to the long-term and systemic detriment of our welfare

I am not at all an expert in this field but even casual awareness of research [1] confirms intuition of very complex patterns: people will punish altruists because, e.g., they are not altruistic enough(!), projecting hidden motives or simply hating their guts (do-gooders)

https://www.science.org/doi/10.1126/science.1153808

emilymbender, to random
@emilymbender@dair-community.social avatar

With the OpenAI clownshow, there's been renewed media attention on the xrisk/"AI safety"/doomer nonsense. Personally, I've had a fresh wave of reporters asking me naive questions (as well as some contacts from old hands who are on top of how to handle ultra-rich man-children with god complexes). 🧵1/

phiofx,

@emilymbender it might be helpful to draw analogies with automated systems and algorithms that have already been used for a long time (e.g in credit scoring, insurance, medical etc) to bring some of the real issues and risks home to people.

There is a track record of regulated use, explainability requirements and rules for overruling that people would be familiar with.

Pretending "AI" is brand new and different is part of the ploy. A repeat of the move-fast-and-break-things strategy

phiofx,

@emilymbender sorry for amplifying your message. you must be a really conceited person. will not happen again.

  • All
  • Subscribed
  • Moderated
  • Favorites
  • •
  • anitta
  • osvaldo12
  • magazineikmin
  • cubers
  • thenastyranch
  • ethstaker
  • Youngstown
  • ngwrru68w68
  • slotface
  • mdbf
  • rosin
  • InstantRegret
  • kavyap
  • DreamBathrooms
  • JUstTest
  • khanakhh
  • Leos
  • modclub
  • normalnudes
  • everett
  • tacticalgear
  • Durango
  • GTA5RPClips
  • provamag3
  • megavids
  • tester
  • cisconetworking
  • lostlight
  • All magazines
    •