Twitter admits bias in algorithm for rightwing politicians and news outlets | Twitter | The Guardian
<https://www.theguardian.com/technology/2021/oct/22/twitter-admits-bias-in-al...> Twitter has admitted it amplifies more tweets from rightwing politicians and news outlets than content from leftwing sources. The social media platform examined tweets from elected officials in seven countries – the UK, US, Canada, France, Germany, Spain and Japan. It also studied whether political content from news organisations was amplified on Twitter, focusing primarily on US news sources such as Fox News, the New York Times and BuzzFeed. The study compared Twitter’s “Home” timeline – the default way its 200 million users are served tweets, in which an algorithm tailors what users see – with the traditional chronological timeline where the most recent tweets are ranked first. The research found that in six out of seven countries, apart from Germany, tweets from rightwing politicians received more amplification from the algorithm than those from the left; right-leaning news organisations were more amplified than those on the left; and generally politicians’ tweets were more amplified by an algorithmic timeline than by the chronological timeline. According to a 27-page research document, Twitter found a “statistically significant difference favouring the political right wing” in all the countries except Germany. Under the research, a value of 0% meant tweets reached the same number of users on the algorithm-tailored timeline as on its chronological counterpart, whereas a value of 100% meant tweets achieved double the reach. On this basis, the most powerful discrepancy between right and left was in Canada (Liberals 43%; Conservatives 167%), followed by the UK (Labour 112%; Conservatives 176%). Even excluding top government officials, the results were similar, the document said. Twitter said it wasn’t clear why its Home timeline produced these results and indicated that it may now need to change its algorithm. A blog post by Rumman Chowdhury, Twitter’s director of software engineering, and Luca Belli, a Twitter researcher, said the findings could be “problematic” and that more study needed to be done. The post acknowledged that it was concerning if certain tweets received preferential treatment as a result of the way in which users interacted with the algorithm tailoring their timeline. “Algorithmic amplification is problematic if there is preferential treatment as a function of how the algorithm is constructed versus the interactions people have with it. Further root cause analysis is required in order to determine what, if any, changes are required to reduce adverse impacts by our Home timeline algorithm,” the post said. Twitter said it would make its research available to outsiders such as academics and it is preparing to let third parties have wider access to its data, in a move likely to put further pressure on Facebook to do the same. Facebook is being urged by politicians on both sides of the Atlantic to distribute its research to third parties after tens of thousands of internal documents – which included revelations that the company knew its Instagram app damaged teenage mental health – were leaked by the whistleblower Frances Haugen. The Twitter study compared the two ways in which a user can view their timeline: the first uses an algorithm to provide a tailored view of tweets that the user might be interested in based on the accounts they interact with most and other factors; the other is the more traditional timeline in which the user reads the most recent posts in reverse chronological order. The study compared the two types of timeline by considering whether some politicians, political parties or news outlets were more amplified than others. The study analysed millions of tweets from elected officials between 1 April and 15 August 2020 and hundreds of millions of tweets from news organisations, largely in the US, over the same period. Twitter said it would make its research available to third parties but said privacy concerns prevented it from making available the “raw data”. The post said: “We are making aggregated datasets available for third party researchers who wish to reproduce our main findings and validate our methodology, upon request.” Twitter added that it was preparing to make internal data available to external sources on a regular basis. The company said its machine-learning ethics, transparency and accountability team was finalising plans in a way that would protect user privacy. “This approach is new and hasn’t been used at this scale, but we are optimistic that it will address the privacy-vs-accountability tradeoffs that can hinder algorithmic transparency,” said Twitter. “We’re excited about the opportunities this work may unlock for future collaboration with external researchers looking to reproduce, validate and extend our internal research.”
Ciao Alberto, grazie della segnalazione. On October 23, 2021 6:26:33 AM UTC, Alberto Cammozzo wrote:
Twitter has admitted it amplifies more tweets from rightwing politicians and news outlets than content from leftwing sources.
Dunque si sono accorti che non possono contenere tutti i leak e cercano di giocare d'anticipo, in modo da orientare la narrazione. Qualcuno si chiederà quale costo ha avuto questo bug per le democrazie negli scorsi anni? Come è stato scelto il periodo da analizzare? ;-) E come sono andate le elezioni? [0]
“Algorithmic amplification is problematic if there is preferential treatment as a function of how the algorithm is constructed versus the interactions people have with it. Further root cause analysis is required in order to determine what, if any, changes are required to reduce adverse impacts by our Home timeline algorithm,” the post said.
Attenzione che qui Twitter sta buttando fumo negli occhi. L'amplificazione automatica[1] è problematica ogni volta che non è sotto il completo controllo di chi la subisce. Ogni volta che una terza parte può tecnicamente amplificare un messaggio (e dunque smorzarne altri) ricevuto da una persona, quel sistema di comunicazione è "broken beyond repair".
Twitter said it would make its research available to outsiders such as academics and it is preparing to let third parties have wider access to its data, in a move likely to put further pressure on Facebook to do the same.
Attenzione che Google non è per nulla diverso, solo più invisibile. È persino riuscito ad entrare nel vocabolario come sinonimo di "cercare" [2], per cui gli utenti lo associano con una esperienza (la ricerca di un documento in un archivio o di un calzino in un armadio) oggettiva, non manipolabile: il calzino o c'è o non c'è, il documento che trovo può non essere il più attinente alla mia ricerca, ma nessuno lo sceglie per me, contando di farmi sostenere una tesi piuttosto che un'altra.
Facebook is being urged by politicians on both sides of the Atlantic to distribute its research
A volte ho l'impressione che abbiano deciso di sacrificare lo scemo del villaggio come capro espiatorio.
Twitter said it would make its research available to third parties but said privacy concerns prevented it from making available the “raw data”. The post said: “We are making aggregated datasets available for third party researchers who wish to reproduce our main findings and validate our methodology, upon request.”
Attenzione al passaggio! I dati personali sono EMESSI inconsapevolmente dalle persone. I contenuti invece sono dati ESPRESSI consapevolmente. Se io raccolgo personalmente i miei dati personali (ad esempio le mie posizioni GPS o i miei dati medici) e DOPO averli analizzati personalmente record per record ed ALTERATI per tutelare la mia privacy, li cedo alla comunità scientifica per fini di ricerca, quei dati smettono di essere una mia emissione e diventano mia espressione. SOLO se progettata in questo modo, con un passaggio di revisione obbligatoria da parte del soggetto emittente, una politica di "solidarietà dei dati" è compatibile con libertà individuali e democrazie. Lo stesso però vale per aziende e governi: i dati "anonimizzati" o "privacy friendly" fornite da un'organizzazione non sono EMISSIONI di quell'organizzazione ma loro ESPRESSIONE. Come tali, non descrivono l'azienda (o i software che esegue sugli utenti), ma l'immagine che l'azienda vuole dare di sé (e dei suoi software). La differenza fondamentale fra dato descrittivo emesso da un'entità e dato espresso (aka "contenuto") da quella stessa entità, è che l'espressione è una scelta consapevole che permette legittimamente[3] di omettere e mentire.
Twitter added that it was preparing to make internal data available to external sources on a regular basis. The company said its machine-learning ethics, transparency and accountability team was finalising plans in a way that would protect user privacy.
“This approach is new and hasn’t been used at this scale, but we are optimistic that it will address the privacy-vs-accountability tradeoffs that can hinder algorithmic transparency,” said Twitter. “We’re excited about the opportunities this work may unlock for future collaboration with external researchers looking to reproduce, validate and extend our internal research.”
Immaginatevi l'autorevolezza dei lobbisti a Bruxelles, se potessero sostenere che le piattaforme di sorveglianza sono privacy friendly facendo riferimento ad articoli "scientifici" persino riproducibili! Quale politico gli farà notare che gli "scienziati" in questione hanno analizzato e tratto conclusioni solo sui dati che queste piattaforme hanno fornito loro? Giacomo [0] https://en.wikipedia.org/wiki/2021_Madrilenian_regional_election [1] NON "algoritmica": questi sono SOFTWARE reali, pieni di BACHI, non algoritmi. [2] https://www.treccani.it/vocabolario/googlare_res-6255901c-89c4-11e8-a7cb-002...) [3] è la possibilità di mentire a rendere preziosa l'onestà e necessaria la fiducia nonché possibile la libertà. chi propone un mondo senza tale possibilità è consapevole delle asimmetrie di potere che tale sistema non impedirebbero, per cui chi ha potere potrebbe mentire o omettere, ma la maggioranza delle persone no.
participants (2)
-
Alberto Cammozzo -
Giacomo Tesio