Weekly Chronicles #45
Scroll down for the english version
L’IRS non si accontenta
Non devono essere sembrati sufficienti gli 87.000 nuovi dipendenti che l’agenzia fiscale statunitense conta di assumere nel corso dei prossimi, dato che hanno annunciato di voler ricorrere a strumenti d’intelligenza artificiale per stringere la morsa sull’evasione fiscale.
L’agenzia dice che agli algoritmi di machine learning saranno dati in pasto quei casi troppo complessi per l’essere umano, nella speranza di cavare qualche ragno dal buco:
“[…] to target the wealthiest Americans and tackle the kinds of cases that had become too complex and cumbersome for the beleaguered agency to handle.The agency’s new funding is supposed to help the I.R.S. raise more federal revenue by cracking down on tax cheats and others who use sophisticated accounting maneuvers to avoid paying what they owe.”
La soluzione sembra a portata di mano: se è troppo complicato, usiamo l’intelligenza artificiale. Forse però i predatori dell’IRS non sanno che gli algoritmi sono fallibili e spesso danno vita ad elevati tassi d’errore. O magari fanno finta di non saperlo.
Il problema però è grave: se i casi sono talmente complessi da non essere intelligibili da menti umane, chi sarà in grado di verificare che l’intelligenza artificiale non abbia commesso errori? Quale avvocato potrà mai difendere i suoi clienti dalle oscure illazioni di un algoritmo troppo complesso da comprendere? Quale giudice potrà mai decidere nel merito?
Poco importa, d’altronde quando si tratta di fisco siamo tutti colpevoli fino a prova contraria. L’algoritmo t’incastra, l’impiegato IRS ti porta via la casa. O ti spara. O magari entrambe.
Quando compri un’auto nuova, considera anche la privacy
Lo sappiamo. Le auto nuove sono computer con le ruote: wi-fi, bluetooth, sistemi operativi che si aggiornano in automatico, sensori di ogni tipo, assistenti vocali e così via. Un incubo per la privacy. E se non lo sai, ti consiglio di leggere questo articolo sul funzionamento delle nuove auto intelligenti e il futuro distopico che ci aspetta.
Sarebbe quindi buona abitudine quando acquistiamo un’auto nuova, valutare anche il modo in cui trattano i nostri dati (e quali dati) oltre al motore e agli optional. Non è facile: dovremmo leggere decine di lunghe privacy policy scritte in legalese e valutare con attenzione le possibili conseguenze.
Fortunatamente i ricercatori di Mozilla ci danno una mano. Col loro progetto “Privacy Not Included” hanno recensito diversi produttori automobilistici per valutare chi fosse il peggiore in termini di privacy. E beh… se hai una Nissan, ho cattive notizie per te.
La loro privacy policy afferma che l’auto può acquisire e inferire dati di ogni tipo, comprese preferenze e abitudini sessuali. I dati vengono poi rivenduti a broker di vario tipo e finiscono così nel vortice infinito dell’advertising e della profilazione. Fidati, meglio fare sesso nel caro e vecchio letto. A meno che tu non abbia Alexa in camera…
Privacy Chronicles is a reader-supported publication. To receive new posts and support my work, consider becoming a free or paid subscriber.
Privacy Chronicles is a reader-supported publication. To receive new posts and support my work, consider becoming a free or paid subscriber.
Privacy Chronicles is a reader-supported publication. To receive new posts and support my work, consider becoming a free or paid subscriber.
Meno di due settimane alla Privacy Week 2023!
Manca pochissimo al festival dell’anno per appassionati di privacy e tecnologia!
Se non l’hai ancora fatto ti consiglio di seguire la pagina Instagram e iscriverti alla newsletter per non perdere nessun aggiornamento.
Come già anticipato la maggior parte degli eventi saranno trasmessi in streaming su www.privacyweek.it, ma ci sono alcuni momenti d’incontro dal vivo con ingresso libero (previa registrazione sul sito):
- Giovedì 28 settembre, dalle ore 20:30 presso Phyd, Via Tortona 31, (Milano) in cui si terrà la Privacy Night con Diego Passoni di Radio Deejay
- Venerdì 29 settembre, dalle ore 14: presso 21 House of Stories, Via Enrico Nöe 24, Città Studi (Milano) in cui si terrà l’intervista a Max Schrems e anche Bitcoin Beach, un workshop dal vivo per imparare a usare Bitcoin nel tuo esercizio commerciale (o per fare acquisti), a cura di Milano Trustless
Weekly memes
Weekly quote
“Wikipedia is the best thing ever. Anyone in the world can write anything they want about any subject. So you know you are getting the best possible information.”
Michael Scott
Support Privacy Chronicles!
One way to support Privacy Chronicles is to share this article with your friends and acquaintances and leave a like or a comment! Another way is to subscribe (with Bitcoin as well) for the equivalent of a monthly breakfast.
Or, you can make a voluntary donation in sats by scanning this QR Code:
Scan the QR Code or click here!
English Version
The IRS isn't satisfied
The 87,000 new employees that the United States tax agency plans to hire in the coming years must not have seemed sufficient, as they have announced their intention to leverage artificial intelligence tools to tighten the grip on tax evasion.
The agency states that machine learning algorithms will be tasked with handling cases that are too complex for humans, hoping to uncover hidden tax frauds:
"[...] to target the wealthiest Americans and tackle the kinds of cases that had become too complex and cumbersome for the beleaguered agency to handle.The agency’s new funding is supposed to help the I.R.S. raise more federal revenue by cracking down on tax cheats and others who use sophisticated accounting maneuvers to avoid paying what they owe."
The solution seems within reach: if it's too complicated, just use artificial intelligence.
However, perhaps the IRS predators are unaware that algorithms are fallible and often lead to high error rates. Or maybe they pretend not to know.
The problem, however, is significant: if cases are so complex that they are incomprehensible to human minds, who will be able to verify that the artificial intelligence has not made errors? Which lawyer can defend their clients against the obscure speculations of an algorithm too complex to understand? Which judge can make a fair decision?
It doesn't matter much, after all, when it comes to taxes; we are all guilty until proven innocent. The algorithm frames you, and the IRS employee takes away your home. Or perhaps he’ll shoot you. Or maybe both.
When you buy a new car, keep in mind your privacy too
We know it. New cars are computers on wheels: Wi-Fi, Bluetooth, automatically updating operating systems, all kinds of sensors, voice assistants, and so on. A privacy nightmare.
So, it would be a good practice when buying a new car to evaluate how they handle your data (and what data) in addition to the engine and other features. It's not easy: you would have to read dozens of lengthy privacy policies written in legal jargon and carefully consider the possible consequences.
Fortunately, Mozilla researchers are here to help. With their "Privacy Not Included" project, they have reviewed various car manufacturers to determine who is the worst in terms of privacy. Well... if you have a Nissan, I have bad news for you.
Their privacy policy states that the car can acquire and infer all kinds of data, including sexual preferences and habits. The data is then sold to various brokers and ends up in the endless vortex of advertising and profiling. Trust me, it's better to have sex in your good old bed. Unless you have Alexa in the bedroom...
Less than two weeks until Privacy Week 2023!
The year's festival for privacy and technology enthusiasts is just around the corner!
If you haven't already, I recommend following the Instagram page and subscribing to the newsletter to stay updated.
As previously mentioned, most events will be streamed on www.privacyweek.it, but there are some in-person meetings with free entry (registration required on the website):
- Thursday, September 28, starting at 8:30 PM at Phyd, Via Tortona 31, Milan, where the Privacy Night with Diego Passoni of Radio Deejay will be held.
- Friday, September 29, starting at 2:00 PM at 21 House of Stories, Via Enrico Nöe 24, Città Studi, Milan, where the interview with Max Schrems and Bitcoin Beach, a live workshop to learn how to use Bitcoin in your business (or for shopping), will take place, curated by Milano Trustless.
privacychronicles.it/p/weekly-…
