Notices by phiofx (phiofx@hachyderm.io), page 14

Researchers discover thousands of nanoplastic bits in bottles of drinking water. 90% were small enough to qualify as nanoplastics: so small that they can be absorbed into human cells and tissue, as well as cross the blood-brain barrier.
https://www.latimes.com/environment/story/2024-01-08/thousands-of-nanoplastics-found-in-bottled-drinking-water

@Sustainable2050 a fractal pattern (happening at all scales) of environmental degradation seems the "natural" path for the unfolding of the unnatural, human engineered disasters.

After all those years and the infinite number of invented markup formats and #markdown variations, if you want to organize textual information with non-zero complexity you still need to use a #spreadsheet

Something that could improve social media (and the #fediverse) would be an automated prompt before final "submit" with a checklist

Are you sure you are not being stupid, ignorant, vile, vacuous, vindictive, naive, pretentious, impatient, pedantic, boring, double speaking, egomaniac, abusive, gratuitously offensive, psychopathic, conspiratorial, wasting yours and everybody else's attention and energy and diverting it from experiencing something good and worthwhile or at least having a good time?

@ssutch #MastoAdmin hashtag is a good resource to find them. I've got help and pulled great tools from it. Also, general discussions. I also found some by browsing Mastodon's GitHub, by looking at profiles of coders when activities happen to the code.

@GrrlScientist with the Artic warming up and the ice sheet melting, destitute fishermen can redeploy to support the growing the oil and gas infrastructure and maritime routes opening up. /S

There are always economically beneficial ways to accelerate environmental collapse...

@elefanto I see activitypub as maybe the currently most visible of a large number of innovative but niche efforts to "re"decentralize the internet that dont yet amount to a complete redesign either in technical, economical or political terms

How things will play out depends on how comprehensive the deprecation of the current tech oligopolists and their ugly business models, what sort of separation of commercial interests from public good, but above all, the degree of societal #digitalliteracy

Another approach would be to offer alternatives to the timeline. In an email or rss client type view where new emails or posts are automatically slotted per account you dont have to worry about missing that rare but valuable new post. A simple visual indicator (eg boldface) immediately guides you to the new content

Manually curating a #timeline with follows of accounts and #hashtags feels like a first, baby step towards solving the hard information firehose problem.

But it fails almost immediately: some channels are much more chatty than others and dominate. Its like listening to a band where some instruments swamp all others.

One approach would be to implement a sort of #equalizer of content density: Downsample to a common frequency of displayed posts, to give quieter channels some visibility.

@davidho clean pollution is the future

@simon the mathematical structure of algorithms is as objective as it gets in terms of classifying them and "#AI" in its current #llm form is an incremental evolution of a vast prior body that ultimately goes back to linear regression.

Fitting functions to data, extrapolating and doing something with the outcome is bread and butter in many industries.

I suspect that one factor reinforcing the (ab)use of the term "AI" is to decouple any regulatory discussion from historically established norms

@risottobias @morgan @djnavarro imho a challenge for the #fediverse is the still partial understanding of the dynamics of online information exchange. How design interacts with and shapes human behavior: E.g., how actors and content are discovered, moderated, amplified. Feedback loops and the coupling to "real life".

There is a vast range of options and a sample of a handful #adtech implementations hardly spans what is desirable. Patterns like #email and #rss point to a much wider space.

@inthehands i think the abstract, macro, idea is that companies must treat their clients well (trust as a condition for staying in business). But we know that at the micro level this is breached all the time. Surveillance capitalism did not invent greed, just the next level of it.

For-profit search business models where clients pay directly should minimally be heavily regulated and monitored, like any entity that has access to such vast amounts of user data.

@djnavarro @risottobias everything is political and political discussions are good for the soul and society. But conducting them in constructive civility requires attributes from both forum and participants that are sorely missing in online digital incarnations.

I am not too hopeful that the current fediverse is structurally different. Just a selection effect of early adopters. It will take a lot of vision, thinking and implementing to invent something fit-for-purpose.

@inthehands i suspect that just like mozilla and any other niche operator facing the concentration of power and money in bigtech they dont have many options. Money is scarce for wannabee disruptors of the worst cartel in history, existentially scarce.

Not excusing any choices, just pondering the conditions under which they are made.

Imho we are now in a phase where only search-as-public-good can change the status quo. Essentially only governments/states can kill the monster they created.

@command_tab #github is for social media, #gitlab is for work

@jaz its a fantastic way to make the #fediverse more tangible. People are territorial animals they immediately relate to place :-)

Something that would be nice for future directions is to include the various upcoming non-#mastodon type servers (#peertube, #pixelfed etc). Some of them, like #mobilizon even have an explicit geospatial dimension (meetups, events etc)

Taste, in the aesthetic, artistic sense is such an opaque concept and so unhelpfully enmeshed in the nature versus nurture debate.

Sometimes people advance nurture arguments, such as "we grow to like music forms we have been exposed to in childhood"

But this can hardly be the whole picture. If you compile any list of artifacts that have seen a long sequence of visual appearance changes (clothes, cars, buildings, fonts etc) what people "like" (disclosure: n=1) seems pretty random.

@simon in the ten minutes saved using AI one can ponder the meaning of #wordclouds, whether they need to exist at all and in what sense they might be an improvement over a nicely ordered table of word frequencies :-)

@jonny yes those multi-page author lists are a physics thing (some astronomy work linked to major observatories is similar) but it wasn't immediately obvious how it interacts with the pattern of ever increasing author "productivity". More like a step function than exponential.

In any case if they would filter out large collaboration papers (with more than X authors), I suspect the remainder would be quite similar to chemistry and biology...