Tanguy Fardet
boosted
Showing activities from people in this circle
@Fischblog ... weis nicht, das "6*7=42" wahr ist, obwohl er es berechnen kann. Das ist doch nicht schwer zu kapieren, das das ganze einfach nur eine "Phrasendreschmaschine" ist. Halt eine gute. Ein solches System kann halt die Strukturen von Erkenntnis sprachlich reproduzieren, ohne selbst Erkenntnis im menschlichen Sinn zu vollziehen.
@DanKen Ja, eben. Das ist einfach ein außerordentlich ausgefeilter statistischer Algorithmus mit sehr viel Rechenpower dahinter. Deswegen klappt das ja auch mit dem Prompt Engineering nicht. Das ist auch nur ein Input mit bestimmten statistischen Eigenschaften.
Voted best Burger place in Wales
Man, Wales must be depressing to live in if this counts as a good burger. The order terminal didn't give me options to add extras (except paid-for fries that I didn't want) and even if I could have added some sauce.
I kinda left in a hurry, so maybe there were some, that doesn't excuse a complete lack of lettuce, tomatoes, pickles, cheese, or idfk.
I've been overpaying for food all week but man I feel robbed right now 😒 #emfcamp #emfcamp2026
@kookie what was the name of the place?
Ich bin so müde.
Das ist alles nur Gewäsch und hohle Phrasen. KI ist "auf dem Level des Menschlichen Gehirns" - nach welchem Maßstab? Hm?
Stattdessen nur "alles voll komplex und entwickelt sich total schnell". Heißt was? Guckt man hin, heißt das einfach Rechenpower. Aber 3000 Computer sind halt nicht intelligenter als ein Computer.
Letztlich nur wieder ein Deppensyllogismus:
Das Bewusstsein ist superkomplex. KI ist superkomplex. Also KI = Bewusstsein. 🙄
https://www.theguardian.com/technology/2026/jul/19/could-ai-be-conscious
@Fischblog
Hochgeschwindigkeitstrottel
They lied to you when they told you the genocide had ended.
They lied to you when they told you it was the 'most moral army.'
They lied to you when they told you there were safe zones in Gaza.
They lied to you about the death toll.
They lied to you about the number of the wounded.
They lied to you, and you were even foolish enough to believe them
@joynewacc and we believed it? Supported it? Enough of us did.
Course, general public awareness lags behind reality, but yeah, we complied. We inherited this world, and we also made it.
andy is surely the final prime minister before we just all admit the uk was a massive mistake and we'd be better off sinking the whole place into the sea
@nico Make thr UK Doggerland
"The lefties are starting to win and now they are going to run amok."
You speak like someone you think is smart.
@CartyBoston They're speaking to people that think that the speaker is smart.
Benjamin Sonntag-King
boosted
Want an update on the Weekend at Bernie’s clown show that is W Social? No? Me neither, but here we are…
So before we start, let’s recount the whole dystopian raison d’être of these Davosian clowns: to be a EU®opean “alternative to X” that – cue the 1984 theme tune – verifies your identity to ensure that only “real people” are on it (real is what your government says it is).
Now, let’s fast forward to last week, when TikTok set up an account on W Social.
That’s nice, isn’t it? One surveillance capitalist lending their support to another? Kumbaya and all that. The only problem? (spoiler alert), it wasn’t TikTok. (Shocking, I know.)
The account, Identity Verified™ by EU X – sorry, W Social, was a spam account and immediately labelled as such by BlueSky moderators. (Don’t you love a “decentralised” system where a post on an EU server, accessed from another EU server, is moderated by our benevolent Big Brother in Trumpland, by the way? I know it just warms the cockles of my heart. So much sovereignty! But I digress…)
So @_elena calls them out (someone has to while @EUCommission is busy preparing for their honeymoon after expressing their undying devotion to W Social recently on every social platform¹):
https://mu.social/profile/elenarossini.com/post/3mqob7shxks2t
Anna Zeiter – W CEO, Stanford’s own, and ex-eBay executive (such sovereignty) – replies with this absolute zinger:
> “We are just verifying if some [sic] is a human and at least 18 years old. Users are then free to choose their own user name. This (human) user apparently chose the name TikTok.“ (https://mu.social/profile/anna.wsocial.eu/post/3mqoepm7m2k2z)
Read that as many times as you like until it really sinks in.
These incompetent clowns (dangerous competent clowns but incompetent clowns nevertheless) make people go through a ridiculous 13-step identity verification stage² and then… umm… you get to choose whatever identity you want.
🤡🤡🤡
Don’t get me wrong, I’m glad they’re incompetent. It would be worse if they were competent, given they’re a closed-source venture-capital-funded startup determined to prove your identity.
> “Yes we have a business model, it’s not a philanthropy, we are not a B corp, we actually want to make money.” – Anna Zeiter³
But, what I want to know is this: Why has @EUCommission embraced a proprietary, closed-source, corporate ghost town with 14,300 accounts, 90% of which have not posted even once (in other words, with only 1,259 accounts that have ever posted)⁴ when there is a thriving non-corporate fediverse with millions of active people on it.
Could it have anything to do with identity verification, surveillance, and control?
And if so, why can’t they even get that right?
Looking forward to seeing how long @EUCommission can drag the corpse of W Social around before people realise it’s dead.
Oh, and how does this make the whole “European alternatives” thing look? Not good. Not because there aren’t excellent European alternatives but because the folks at the commission are either too clueless or too corrupt to recognise and support them.
So, yeah, how’s your Monday going?
¹ e.g., https://mastodon.ar.al/@aral/116770333053823108, etc.
² https://aseachange.com/@elena/statuses/01KVD55YBYVM3B46ACQTECDMT2
³ https://blog.elenarossini.com/the-untold-story-about-w-social-unconventional-beginnings-strategic-pitches-conflicting-signals/ (from the YouTube video that they since deleted. Nice transparency there, clowns.)
⁴ Run this script in Node 18+ to reproduce the numbers from their PDS yourself: https://codeberg.org/aral/gists/src/branch/main/wsocial-pds-stats.js Also see https://blog.elenarossini.com/w-social-a-broken-ghost-town/ and https://blog.elenarossini.com/w-social-fictional-metrics-and-the-beauty-of-open-data/
#WSocial #EU #sovereignty #socialMedia #socialNetworking #mastodon #fediverse #activityPub #ATProto #startups #EUCommission
Increasingly, the greatest value in culture lies not simply in creating beloved characters or stories but in owning the experiences built around them. https://www.japantimes.co.jp/commentary/2026/07/20/japan/japanese-culture-ownership/?utm_medium=Social&utm_source=mastodon #commentary #japan #cooljapanfund #pokemon #pokemongo #nintendo #manga #anime
@pallenberg
Dachte ich mir schon. Er stand so schön am Rand, da kann man das natürlich leicht bewerkstelligen. Sowas kann sonst das Bild verhandeln.
@lordmortis I don't, nor do I want one.
Glad to see #pragmatism advancing arguments and evidence about #intuition and #reflection. I've applied it to
- #CogSci and #metaphilosophy broadly: https://doi.org/10.1111/meta.12534
- #AI and #PhilMind: https://doi.org/10.48550/arXiv.2505.22987
A new paper extends it to #Sociology: https://doi.org/10.1111/jtsb.70053
![Pragmatism, Dual-Process Theory, and the Study of Reflective Cognition
Gordon Brett
This article advances dual-process theory (DPT) by developing a pragmatist framework for the study of reflective cognition. Although dual-process models were introduced to sociology via Bourdieusian practice theory, Bourdieusian DPT has proven inadequate for theorising reflective cognition and action, with sociologists looking to pragmatism as an alternative. To support these efforts, I provide a synthetic, elemental account of the pragmatist theory of reflection, delineating reflective cognition's defining characteristics, core components, its relationship with non-reflective cognition, and its interindividual variation. I then outline the merits of the pragmatist theory of reflection for dual-process research, arguing that it not only addresses the major limitations of Bourdieusian DPT but that it has broad interdisciplinary support [and] provides advantages [to both science and scholarship].](https://media.nerdculture.de/media_attachments/files/116/952/037/437/968/302/original/bd048e1267141433.png)
@urlyman @GhostOnTheHalfShell Those kind of campaigns can backfire by feeding into the fomo they're parodying.
There was a travel agency doing a serious "go before it's too late" campaign in the 90s over images of iconic travel spots destroyed by future progress.
(Ooh, found it! https://www.luerzersarchive.com/work/kilroy-travels/)
The eel's listed as critically endangered, but people still "need" to eat it every year here because "traditions" without realizing they could save the tradition by doing it less often.
christopher
boosted
Another incredible #EMFCamp thank you so much to all of the organisers, volunteers, contributors and everyone I met who showed, as always, incredible enthusiasm to share the things that excite the with everyone.
4 media
hypebot
boosted
Today, we fined AliExpress €550M for breaching its obligations under the DSA to diligently assess and mitigate risks relating to the sale of illegal, unsafe or counterfeit products on its e-commerce platform.
Scale is not an excuse; risks must be identified and addressed systematically to ensure consumers can safely shop online.
Today, we are holding AliExpress to this standard and request it to take action.
I am trying to write a page for our website, for when people ask me to review third party (including AI) generated legal documents, and it is not much fun.
Expectations are, so often, misaligned.
The fact that it was cheap, fast, and easy to generate these documents sometimes leads to a misconception that it will be cheap, fast, and easy to have them reviewed and "signed off.
In fact, it is often *more expensive* for me to review third party generated documents than to deal with the matter from the beginning.
*sigh*
@neil I feel the same way about LLM-generated code and pull requests!
How is this journalism?
There is war in Europe, there is a genocide in Gaza, the world economy is deeply challenged by Trump’s war in Iran. Yet the story on the third most popular news site in Denmark is
“Woman in chock after scary incident in forrest”
Now, you may be thinking, “Was she attacked by a pack of wolves?”, but no… The story from the TV2 journalist is about a woman who saw a branch fall from a tree… you know… as branches do when you walk in the forrest.
https://nyheder.tv2.dk/samfund/2026-07-18-kvinde-i-chok-efter-ny-ubehagelig-oplevelse-i-skov
@kawaiipunk I tried to get rid of all the grass but I will be repacking equipment before putting it away
@Workshopshed I had to pack quickly so I brought a couple of bales oh hay back with me in my socks
@wanwizard @LordWoolamaloo This is AI slop. (Evidence: try reading the sign board to the bottom left, or the illuminated sign in the window behind the pillars.)
@cstross @LordWoolamaloo Most likely, yes. But still funny, unlike most AI slop.
