Most research on electroplating tries to find ways to make it plate parts more uniformly. [Ajc150] took the opposite direction, though, with his selective electroplating project, which uses an electrode mounted on a CNC motion system to electrochemically print images onto a metal sheet (GitHub repository).

Normally, selective electroplating would use a mask, but masks don’t allow gradients to be deposited. However, electroplating tends to occur most heavily at the point closest to the anode, and the effect gets stronger the closer the anode is. To take advantage of this effect, [ajc150] replaced the router of an inexpensive 3018 CNC machine with a nickel anode, mounted an electrolyte bath in the workspace, and laid a flat steel cathode in it. When the anode moves close to a certain point on the steel cathode, most of the plating takes place there.

To actually print an image with this setup, [ajc150] wrote a Python program to convert an image into set of G-code instructions for the CNC. The darker a pixel of the image was, the longer the electrode would spend over the corresponding part of the metal sheet. Since darkness wasn’t linearly proportional to plating time, the program used a gamma correction function to adjust times, though this did require [ajc150] to recalibrate the setup after each change. The system works well enough to print recognizable images, but still has room for improvement. In particular, [ajc150] would like to extend this to a faster multi-nozzle system, and have the algorithm take into account spillover between the pixel being plated and its neighbors.

This general technique is reminiscent of a metal 3D printing method we’ve seen before. We more frequently see this process run in reverse to cut metal.

hackaday.com/2025/07/24/painti…

One of the categories we chose for the One Hertz Challenge is “Could Have Used a 555.” What about when you couldn’t have, but did anyway? The 555 is famously easy to use, but not exactly the most accurate timer out there — one “ticking” at 1 Hz will pulse just about once per second (probably to within a millisecond, depending on the rest of the circuit), but when you need more precise timing, the 555 just won’t cut it. Not on its own, anyway.
Allan Deviation can be a bit confusing, but generally — lower is more accurate
This entry by [burble] shows us how the humble 555 can hold its own in more demanding systems with some help from a GPS receiver. He used the One Pulse per Second (1PPS) output from a GPS module to discipline the 1 Hz output from a 555 by modulating the control voltage with a microcontroller.

Okay, this sounds a bit like baking a cake by buying a cake, scraping all the icing off, then icing it yourself, but what better way to learn how to ice a cake? The GPS-disciplined 555 is way more accurate than a free running one — just check out that Allan Deviation plot. While the accuracy of the standard 555 begins to decrease as oscillator drift dominates, the GPS-disciplined version just keeps getting better (up to a point — it would also eventually begin to increase, if the data were recorded for long enough). Compared to other high-end oscillators though, [burble] describes the project’s accuracy in one word: “Badly.”

That’s okay though — it really is a fantastic investigation into how GPS-disciplined oscillators work, and does a fantastic job illustrating the accuracy of different types of clocks, and some possible sources of error. This project is a great addition to some of the other precision timekeeping projects we’ve seen here at Hackaday, and a very fitting entry to the competition. Think you can do better? Or much, much worse? You’ve got a few weeks left to enter!

hackaday.com/2025/07/24/2025-o…

Microsoft continua a spingere l’intelligenza artificiale in Windows 11 , come se cercasse di capire quali funzionalità rimarranno e quali causeranno una nuova ondata di scetticismo. L’ultimo aggiornamento ha introdotto una serie di strumenti di intelligenza artificiale, tra cui una nuova variante del già famigerato Recall.

Si chiama Copilot Vision ed è uno strumento con trasmissione attiva dello schermo a server remoti, dove l’immagine viene analizzata in tempo reale da una rete neurale. A parole, tutto questo sembra un passo verso un “assistente intelligente“, ma in realtà assomiglia a un fastidioso osservatore che ti guarda alle spalle senza bussare.

A differenza di Recall, che elaborava i dati localmente e quindi, almeno in teoria, consentiva all’utente di controllare la propria privacy, Copilot Vision invia un flusso di screenshot a server Microsoft remoti. Lì, vengono elaborati tramite riconoscimento ottico dei caratteri e un modello linguistico per ottenere una “comprensione” di ciò che accade sullo schermo e suggerire all’utente cosa fare successivamente. Nonostante l’azienda affermi che le immagini risultanti non vengono conservate a lungo e non vengono utilizzate per addestrare modelli o per pubblicità personalizzata, l’idea stessa di trasmettere il contenuto dello schermo al mondo esterno sta già creando tensioni.

Copilot Vision è attualmente disponibile solo negli Stati Uniti e, come sottolinea Microsoft, sarà disponibile anche in “Paesi extraeuropei“, una dicitura che allude chiaramente a un’incompatibilità legale con il nuovo AI Act nell’UE. Allo stesso tempo, Windows 11 stesso, l’unico sistema desktop ufficialmente supportato da Microsoft, ha ricevuto una serie di innovazioni AI locali. Una di queste è il cosiddetto “agente” basato su un modello linguistico locale chiamato Mu. Questo modello è attualmente disponibile solo su dispositivi con Qualcomm Snapdragon, sebbene l’azienda prometta di estendere il supporto a Intel e AMD.

Mu è un “assistente” integrato nell’interfaccia di Windows 11 che non si limita a rispondere alle richieste, ma offre anche la possibilità di eseguire azioni in modo indipendente. L’utente può, ad esempio, digitare nella barra di ricerca: “collega cuffie Bluetooth” e l’agente non mostrerà un elenco di istruzioni, ma tenterà immediatamente di eseguire il comando. A differenza delle versioni precedenti, in cui l’interazione con il sistema era limitata alle funzioni di riferimento, il nuovo agente è in grado di modificare direttamente le impostazioni, tuttavia solo se comprende ciò che gli viene richiesto.

Gli ingegneri Microsoft ammettono che implementare tali funzionalità non è stato facile. Un esempio è il controllo della luminosità. Se un utente ha due monitor, quale è considerato quello principale? Come interpretare frasi vaghe? Per ridurre il numero di errori, gli sviluppatori si sono concentrati sugli scenari più comuni e hanno gradualmente ampliato il supporto per query più complesse. Tuttavia, il rischio principale di tutti i LLM rimane irrisolto: le cosiddette “allucinazioni“, quando il sistema produce un risultato convincente ma completamente errato. Finora, Microsoft non ha spiegato quali misure siano state adottate per ridurre al minimo i danni derivanti da tali errori, soprattutto nel contesto dell’intelligenza artificiale, che non si limita a fornire consigli, ma agisce per conto dell’utente.

Un altro componente dell’aggiornamento è lo strumento “Clicca per fare”, attualmente in anteprima. Include diverse funzionalità contemporaneamente: riconosce la lettura vocale e aiuta a migliorare le capacità vocali, può attivare la modalità di generazione del testo in Word in base al frammento selezionato, include una modalità di lettura “immersiva” con font regolabili e la possibilità di suddividere le parole in sillabe, e consente inoltre di avviare messaggi e creare richieste in Microsoft Teams direttamente dall’interfaccia.

Anche le applicazioni integrate sono state aggiornate. L’editor di foto ora include la funzione Relight, che simula la retroilluminazione di un oggetto in una foto utilizzando fonti di luce virtuali. Paint ora include un generatore di adesivi basato su una descrizione testuale e un sistema migliorato per la selezione degli oggetti in un’immagine. Anche lo Strumento di Cattura è stato aggiornato: ora può adattare automaticamente la cornice al contenuto dello schermo e include un selettore colore. Tuttavia, è improbabile che quest’ultimo richieda l’apprendimento automatico.

Oltre all’intelligenza artificiale, l’aggiornamento di Windows 11 ha introdotto un’altra innovazione degna di nota: la scomparsa della classica schermata blu di errore. Ora, in caso di errore critico, all’utente viene presentata una schermata nera: formalmente, si tratta della stessa schermata blu di errore, ma con uno sfondo diverso. La nuova versione della schermata di errore differisce non solo nel colore, ma anche nel contenuto: al posto del gergo tecnico, ci sono messaggi brevi e più comprensibili. Al posto del precedente Ripristino all’avvio, ora è presente un sistema di diagnostica automatica rapida, Quick Machine Recovery, che, tramite l’ambiente di ripristino di Windows RE, può riparare da remoto errori di massa senza l’intervento dell’utente.

Tutte le innovazioni vengono distribuite tramite un meccanismo di aggiornamento graduale: il Controlled Feature Rollout. Questo evita situazioni in cui un aggiornamento non preparato raggiunge simultaneamente milioni di utenti, causando problemi su larga scala. Chi desidera ottenere le nuove funzionalità più rapidamente può attivare l’opzione “Scarica gli aggiornamenti non appena disponibili” nel sistema. Gli altri possono solo aspettare, o cercare un’alternativa se il ruolo crescente dell’IA nei sistemi operativi non sembra tanto incoraggiante quanto allarmante.

Insieme alle modifiche software, Microsoft ha introdotto il nuovo Surface Laptop 5G, il primo modello della serie con supporto completo per Copilot+ . Il dispositivo è costruito attorno a uno schermo da 13,8 pollici e sarà dotato di chip Intel Core Ultra di seconda generazione, ciascuno dei quali con un processore neurale da 40 TOPS (trilioni di operazioni al secondo). Viene inoltre annunciato il supporto per un modem 5G integrato, che sarà disponibile entro la fine del 2025. L’azienda stessa sottolinea che questa soluzione fornisce una connessione costante ai servizi di intelligenza artificiale nel cloud, incluso Copilot in Microsoft 365.

L'articolo Windows 11, Microsoft spinge sull’intelligenza artificiale con Copilot Vision proviene da il blog della sicurezza informatica.

😍😍😍

youtu.be/18QQWa5MEcs?si=tFj9-s…

#marvel #Fantastic4

Neha Madhira grew up in North Texas with the TV constantly buzzing with world news. Madhira, now 24, recognized that journalism was key to keeping her family informed on the happenings back home in India. But with state-sanctioned violence limiting journalists on the ground from reporting, and few legacy media outlets with reporters that looked like her left to report on it, Madhira also knew there were gaps to be filled.

Nearly a decade later, Madhira is bridging the gaps in Western media’s health and education coverage of the Middle East-North Africa region, South Asia, and their diasporas. The contacts she’s built have allowed her to expand her reporting focus — she recently collaborated with Freedom of the Press Foundation (FPF) on an article for The Intercept featuring testimonials from journalists targeted by the Israeli military.

“In a time where press freedom is definitely in question in the U.S. right now, and censorship on social media and in newsroom settings is even becoming more and more common, it’s really important for me that I stay true to my values of why I started reporting,” Madhira said. “I use freelancing to try my best to cover that gap in reporting when it comes to Western media, and try to cover the communities that I know deserve a platform.”

Madhira first spent years covering breaking news, the COVID-19 pandemic, and the social movements of 2020 for local newsrooms in Austin, Texas, while studying journalism and women’s and gender issues. But with activism movements roaring overseas, coupled with the lack of coverage on the impacts of the pandemic in India, she saw freelancing as an opportunity to cover issues happening in countries that face extreme press censorship for audiences overseas and in the U.S.

“A big part of my job during the pandemic, and even now, is reaching out to health care officials, regardless of what’s happening, to actually see who it’s affecting, why it’s affecting them, and what resources people need,” Madhira said. “If there’s a possibility that that information is being withheld from the public, that becomes a problem. How are we supposed to continue to inform and educate the public on how to stay safe during a pandemic or epidemic if we don’t even have that information to begin with?”

She has built close relationships with journalists on the ground in Iran, Afghanistan, India, Gaza, and the West Bank, relying on their reporting to reach audiences in the U.S. and abroad. Recognizing the privilege she holds, Madhira does her best to ensure their perspectives are reflected in her writing.

“I covered the Woman, Life, Freedom movement in Iran very extensively,” she said. “Two journalists who I really looked up to were arrested around this time last year. I wrote a story on that, and I noticed that a lot of Iranian and Iranian American activists were reaching out to me, appreciating the fact that the story was written, because even writing about their release and the details of how they are doing and how journalism and activism is continuing within the country is a privilege.”

Over the past year and a half, Madhira has covered the medical and humanitarian infrastructure collapses in Gaza and the West Bank, as well as the campus encampments in protest of Israel’s actions, and the shadowy organizations collaborating with the government to identify and persecute students and others who are critical of Israel.

“As we see less and less news coming out of Gaza, I urge people to not look away.”

Neha Madhira

With more than 180 journalists killed by Israel to date, media blackouts, and censorship on social media, Madhira writes to amplify the voices of her colleagues remaining on the ground, including those whose stories she wrote about in The Intercept.

“I have advocated for Palestine since I was a child, and at the beginning of October 2023, I was horrified at the language being used to dehumanize Palestinians,” she said. “As a journalist, I was seeing the gaps in Western media coverage and its support of Israel, and I wanted to help change this narrative in any way I could. As we see less and less news coming out of Gaza, I urge people to not look away and to pay closer attention to passive voice being used to describe the atrocities Palestinians continue to face every day.”

While the stories that Madhira tells are urgent and deserving of immediate attention, communicating with people on the ground in Gaza is a slow, challenging process. “Most of the people I’ve interviewed, whether that be journalists, or medical workers, or humanitarian workers, there is a small gap every single day that they have access to the internet, and we have used that to communicate with each other every single day,” she said. “I continue to do that because their voices are the most important and the most pertinent.”

Reporting from the U.S. on Palestine has not come without its own battles against censorship. Having experienced “shadow banning” that has limited visibility of her social media posts, she said the public must pay attention to the ways social media platforms moderate content to censor certain news, and she calls for users to consume content carefully. By amplifying journalists’ content on social media, independent reporting can reach wider audiences and fight against algorithmic suppression, Madhira added.

“There are a lot more people who are in the dark about what is happening than you would think,” she said. “There are so many nonprofit, independent newsrooms, not only in the U.S., but around the world who do incredible reporting for very little money, and it’s important to pay attention, because these journalists are some of the most skilled and experienced within their field.”

freedom.press/issues/freelanci…