codesanitize

LiveKit and OpenAI with Russ d’Sa

Software Engineering

codesanitize

-

1 May 2025

LiveKit is a platform that gives builders with instruments to construct real-time audio and video functions at scale. It presents an open-source WebRTC stack for creation of dwell, interactive experiences like video conferencing, streaming, and digital occasions. LiveKit has gained vital consideration for its partnership with OpenAI for the Superior Voice function.

Russ d’Sa is the Founding father of LiveKit and has an in depth profession in startups. On this episode he joins Sean Falconer to speak about his startup journey, the early days of Y Combinator, LiveKit, WebRTC, LiveKit’s partnership with OpenAI, voice and imaginative and prescient as the longer term paradigm for pc interplay, and extra.

Sean’s been an educational, startup founder, and Googler. He has printed works overlaying a variety of matters from AI to quantum computing. Presently, Sean is an AI Entrepreneur in Residence at Confluent the place he works on AI technique and thought management. You’ll be able to join with Sean on LinkedIn.

Please click on right here to see the transcript of this episode.

Balancing productiveness and privateness: Safeguarding information within the age of AI-driven instruments

Robotics

codesanitize

-

1 May 2025

0

Balancing productiveness and privateness: Safeguarding information within the age of AI-driven instruments

Taking over repetitive duties, offering insights at speeds far past human capabilities, and considerably boosting our productiveness—synthetic intelligence is reshaping the way in which we work, a lot in order that its use can enhance the efficiency of extremely expert professionals by as a lot as 40%.

AI has already offered an abundance of helpful instruments, from Clara, the AI assistant that schedules conferences, to Gamma, which automates presentation creation, and ChatGPT—the flagship of generative AI’s rise. Likewise, platforms resembling Otter AI and Good Tape, which automate the time-consuming transcription course of. Mixed, these instruments and lots of others present a complete AI-powered productiveness toolkit, making our jobs simpler and extra environment friendly—with McKinsey estimating that AI may unlock $4.4 trillion in productiveness development.

AI’s information privateness challenges

Nevertheless, as we more and more depend on AI to streamline processes and improve effectivity, it’s vital to contemplate the potential information privateness implications.

Some 84% of customers really feel they need to have extra management over how organizations accumulate, retailer, and use their information. That is the precept of knowledge privateness, but this superb clashes with the calls for of AI growth.

For all their sophistication, AI algorithms aren’t inherently clever; they’re well-trained, and this requires huge quantities of knowledge to realize—typically mine, yours, and that of different customers. Within the age of AI, the usual strategy in direction of information dealing with is shifting from “we is not going to share your information with anybody” to “we are going to take your information and use it to develop our product”, elevating issues about how our information is getting used, who has entry to it, and what influence it will have on our privateness long-term.

Information possession

In lots of instances, we willingly share our information to entry providers. Nevertheless, as soon as we do, it turns into troublesome to manage the place it finally ends up. We’re seeing this play out with the chapter of genetic testing agency 23andMe—the place the DNA information of its 15 million prospects will doubtless be bought to the very best bidder.

Many platforms retain the precise to retailer, use, and promote information, typically even after a person stops utilizing their product. The voice transcription service Rev explicitly states that it makes use of person information “perpetually” and “anonymously” to coach its AI programs—and continues to take action even when an account is deleted.

Information extraction

As soon as information is used to coach an AI mannequin, extracting it turns into extremely difficult, if not not possible. Machine studying programs don’t retailer uncooked information; they internalize the patterns and insights inside it, making it troublesome to isolate and erase particular person info.

Even when the unique dataset is eliminated, traces of it should stay in mannequin outputs, elevating moral issues round person consent and information possession. This additionally poses questions on information safety laws resembling GDPR and CCPA—If companies can not make their AI fashions actually ‘neglect’, can they declare to be actually compliant?

Finest practices for guaranteeing information privateness

As AI-powered productiveness instruments reshape our workflow, it’s essential to acknowledge the dangers and undertake methods that safeguard information privateness. These greatest practices can hold your information secure whereas pushing the AI sector to stick to greater requirements:

Search firms that don’t prepare on person information

At Good Tape, we’re dedicated to not utilizing person information for AI coaching and prioritize transparency in speaking this—however that isn’t but the business norm.

Whereas 86% of US customers say transparency is extra vital to them than ever, significant change will solely happen once they demand greater requirements and demand any use of their information is clearly disclosed by voting with their toes, making information privateness a aggressive worth proposition.

Perceive your information privateness rights

AI’s complexity can typically make it really feel like a black field, however because the saying goes, information is energy. Understanding privateness safety legal guidelines associated to AI is essential to understanding what firms can and might’t do along with your information. As an example, GDPR stipulates that firms solely accumulate the minimal quantity of knowledge vital for a particular function and should clearly talk that function with customers.

However as regulators play catch up, the naked minimal is probably not sufficient. Staying knowledgeable means that you can make smarter decisions and make sure you’re solely utilizing providers you’ll be able to belief—Likelihood is, firms that aren’t adhering to the strictest of requirements shall be careless along with your information.

Begin checking the phrases of service

Avoma’s Phrases of Use is 4,192 phrases lengthy, ClickUp’s spans 6,403 phrases, and Clockwise’s Phrases of Service is 6,481. It will take the common grownup over an hour to learn all three.

Phrases and situations are sometimes complicated by design, however that doesn’t imply they need to be neglected. Many AI firms bury information coaching disclosures inside these prolonged agreements—a apply I imagine needs to be banned.

Tip: To navigate prolonged and complicated T&Cs, think about using AI to your benefit. Copy the contract into ChatGPT and ask it to summarize how your information shall be used—serving to you to know key particulars with out scanning by countless pages of authorized jargon.

Push for better regulation

We must always welcome regulation within the AI area. Whereas an absence of oversight could encourage growth, the transformative potential of AI calls for a extra measured strategy. Right here, the rise of social media—and the erosion of privateness brought about attributable to insufficient regulation—ought to function a reminder.

Simply as we’ve got requirements for natural, honest commerce, and safety-certified merchandise, AI instruments have to be held to clear information dealing with requirements. With out well-defined laws, the dangers to privateness and safety are simply too nice.

Safeguarding privateness in AI

In brief, whereas AI harnesses important productivity-boosting potential—bettering effectivity by as much as 40%—information privateness issues, resembling who retains possession of person info or the problem of extracting information from fashions, can’t be ignored. As we embrace new instruments and platforms, we should stay vigilant about how our information is used, shared, and saved.

The problem lies in having fun with the advantages of AI whereas defending your information, adopting greatest practices resembling in search of clear firms, staying knowledgeable about your rights, and advocating for appropriate regulation. As we combine extra AI-powered productiveness instruments into our workflows, sturdy information privateness safeguards are important. We should all—companies, builders, lawmakers, and customers—push for stronger protections, better readability, and moral practices to make sure AI enhances productiveness with out compromising privateness.

With the precise strategy and cautious consideration, we are able to tackle AI’s privateness issues, making a sector that’s each secure and safe.

swiftui – iOS WKWebView exhibits about:clean after app stop and relaunch

iOS Development

codesanitize

-

1 May 2025

0

I am utilizing WKWebView in my iOS app to show an online web page. It really works fantastic when the app is working or backgrounded. Nonetheless, after I stop the app (swipe up) and relaunch it, the WebView solely exhibits about:clean as an alternative of the anticipated URL.

The web site masses simply fantastic after I delete and reinstall the app.

WebView.swift

struct WebView: UIViewRepresentable {
    
    var url: URL?
    
    var redirect: ((URL) -> Void)?
    
    func makeUIView(context: Context) -> WKWebView {
        let webView = WKWebView()
        webView.isInspectable = true
        webView.navigationDelegate = context.coordinator
        return webView
    }
    
    func updateUIView(_ webView: WKWebView, context: Context) {
        if let url = url {
            webView.load(URLRequest(url: url))
        }
    }
    
    func makeCoordinator() -> Coordinator {
        Coordinator(mother or father: self)
    }
    
    class Coordinator: NSObject, WKNavigationDelegate {
        
        var mother or father: WebView
        
        init(mother or father: WebView) {
            self.mother or father = mother or father
        }
    
        func webView(_ webView: WKWebView, didFinish navigation: WKNavigation!) {
            let javascript = """
                if (doc.querySelector('meta[name="viewport"]')) {
                    doc.querySelector('meta[name="viewport"]').setAttribute('content material', 'width=device-width, initial-scale=1.0, maximum-scale=1.0, user-scalable=no');
                } else {
                    const meta = doc.createElement('meta');
                    meta.setAttribute('identify', 'viewport');
                    meta.setAttribute('content material', 'width=device-width, initial-scale=1.0, maximum-scale=1.0, user-scalable=no');
                    doc.getElementsByTagName('head')[0].appendChild(meta);
                }
                true;
            """
            
            webView.evaluateJavaScript(javascript) { consequence, error in
                if let error = error {
                    print("Error injecting JavaScript: (error)")
                } else {
                    print("JavaScript executed efficiently.")
                }
            }
            print("WebView completed loading.")
        }

        func webView(_ webView: WKWebView, didFail navigation: WKNavigation!, withError error: Error) {
            print("WebView failed with error: (error.localizedDescription)")
        }

        func webView(_ webView: WKWebView, didFailProvisionalNavigation navigation: WKNavigation!, withError error: Error) {
            print("WebView provisional load failed: (error.localizedDescription)")
        }
        
        func webView(_ webView: WKWebView, decidePolicyFor navigationAction: WKNavigationAction, decisionHandler: @escaping (WKNavigationActionPolicy) -> Void) {
            if let url = navigationAction.request.url, url.scheme == "oauthapp" {
                mother or father.redirect?(url)  // Cross deep hyperlink to your SwiftUI view
                decisionHandler(.cancel) // Cancel navigation
                return
            }
            decisionHandler(.permit)
        }
    }

}

ContentView

func buildAuthURL() -> URL? {
    var parts = URLComponents(string: authURL)
    parts?.queryItems = [
        URLQueryItem(name: "response_type", value: "code"),
        URLQueryItem(name: "client_id", value: clientID),
        URLQueryItem(name: "redirect_uri", value: redirectURI),
        URLQueryItem(name: "code_challenge", value: codeChallenge),
        URLQueryItem(name: "code_challenge_method", value: "S256"),
        URLQueryItem(name: "scope", value: scopes)
    ]
    return parts?.url
}

var physique: some View {
    if accessToken == nil {
        WebView(url: buildAuthURL()){ url in
            if let parts = URLComponents(string: url.absoluteString),
               let code = parts.queryItems?.first(the place: { $0.identify == "code" })?.worth {
                print("Authorization code: (code)")
                authCode = code
                requestAccessToken(code: code)
            } else {
                print("Didn't extract code from URL.")
            }
        }.ignoresSafeArea()
    } else {
        Textual content("Authorization Code: ((authCode ?? "").prefix(15))...")
        Textual content("Entry Token: ((accessToken ?? "").prefix(15))...")
    }
}

What’s occurring:
After quitting and relaunching the app:

The WebView seems.
But it surely solely exhibits a clean white display screen with about:clean because the URL.

Notes:

This solely occurs after a chilly begin (i.e., app was pressure stop).
There is not any crash, and every thing else within the app works fantastic.

Query:

How can I be sure that WKWebView correctly masses the supposed URL as an alternative of displaying about:clean after a chilly begin?

Robots-Weblog | Inklusionsprojekt mit Low-Price-Roboter gewinnt ROIBOT Award von igus

Robotics

codesanitize

-

1 May 2025

0

Robots-Weblog | Inklusionsprojekt mit Low-Price-Roboter gewinnt ROIBOT Award von igus

Wittekindshofer Werkstätten realisieren behindertengerechten Arbeitsplatz mit Low-Price-Roboter für 4.970 Euro

Köln, 10. April 2025 – Daniel Hillebrand leidet an einer Tetraspastik, die eine kontrollierte Bewegung der Extremitäten unmöglich macht. Trotzdem kann er selbstbestimmt arbeiten – dank eines automatisierten Arbeitsplatzes, den die Diakonische Stiftung Wittekindshofer Werkstätten aus Unhealthy Oeynhausen trotz engen Budgets mit einem Low-Price-Roboter von igus realisiert hat. Für dieses kreative Inklusionsprojekt erhielt die Stiftung jetzt den ROIBOT Award. Der Wettbewerb zeichnet progressive und wirtschaftliche Automatisierungsprojekte aus, die mithilfe von igus Produkten erfolgreich umgesetzt wurden. Zu den weiteren Preisträgern zählen das niederländische Unternehmen Paperfoam, das französische Forschungsinstitut CNRS und die Universität Politecnico aus Mailand.

Daniel Hillebrand sitzt im Rollstuhl und bewegt mit seinem Kinn einen Joystick. Damit steuert er einen Roboterarm, der Kunststoffbauteile sortiert. Mehrere Stunden professional Tag, ohne fremde Hilfe. „Daniel ist es gewohnt, in seinem Leben quick vollständig auf Hilfe angewiesen zu sein“, sagt Torsten Jeschke, Elektriker und Erzieher in den Wittekindshofer Werkstätten. „Dank der neuen Anlage kann er nun trotz seiner schweren Lähmung selbstbestimmt arbeiten.“ Das sei für ihn der Himmel auf Erden. „Der Roboter ist cool“, bestätigt Daniel Hillebrand. „Ich musste in die Technik erst reinkommen, aber mittlerweile läuft alles richtig intestine. Am schönsten ist es, wenn der Sack nach langer Arbeit voll ist.“

„Ein Automationsprojekt, das für uns bei igus besonders ergreifend ist.“
Marktübliche Industrieroboter wären für die Wittekindshofer Werkstätten unerschwinglich und in der Steuerung zu komplex gewesen. Jeschke hat deshalb eine günstigere Lösung zusammengestellt, die sich ähnlich leicht bedienen lässt wie ein Computerspiel – mithilfe der Low-Price-Robotik-Plattform RBTX von igus. Herzstück und Daniel Hillebrands Armersatz ist dabei der ReBeL, ein Gelenkarmroboter aus Hochleistungskunststoff für nur 4.970 Euro. igus hatte den ROIBOT-Wettbewerb zum mittlerweile dritten Mal ausgeschrieben, um Unternehmen und Organisationen auszuzeichnen, die mithilfe des RBTX-Marktplatzes besonders smarte und wirtschaftliche Automationsprojekte realisieren. Die Gewinner erhalten Gutscheine für Robotik-{Hardware} im Wert von bis zu 5.000 Euro. „Für uns ist es wirklich ergreifend zu sehen, wie es die Wittekindshofer Werkstätten geschafft haben, mit begrenzten finanziellen Ressourcen und dafür umso mehr Fantasie ein Automationsprojekt auf die Beine zu stellen, welches das Leben eines Menschen so sehr verbessert. Wir hoffen, dass sie den 5.000 Euro-Gutschein nutzen können, um in Zukunft noch weitere Projekte dieser Artwork umzusetzen“, sagt Alexander Mühlens, Leiter des Geschäftsbereichs Low-Price-Automation bei igus und Schirmherr der ROIBOT Awards. igus selbst hat die Good Work Constitution des Verband Deutscher Maschinen- und Anlagenbauer unterschrieben und sich damit dem positiven Beitrag von Robotik zur Gesellschaft verpflichtet. Die Charta betont, dass Robotik und Automatisierungstechnologien nicht nur die Produktivität steigern, sondern auch das Leben der Menschen verbessern können, indem sie Arbeitsbedingungen optimieren und neue Möglichkeiten schaffen.

Die weiteren Preisträger: Roboterkomponenten für die Qualitätssicherung, Astroteilchenphysik und automatisierte Obsternte
Platz zwei und 2.500 Euro für Robotik-{Hardware} gehen an Paperfoam. Die niederländische Firma hat den Gelenkarmroboter ReBeL von igus mit einer Kamera ausgestattet, um ihre biobasierten und recycelbaren Verpackungen stichprobenweise auf Produktionsfehler zu prüfen. Die Lösung reduziert die körperliche Belastung der Mitarbeiter und erhöht gleichzeitig die Qualität der Produktion. Über Platz drei und 1.000 Euro freut sich das französische Forschungsinstitut Centre nationwide de la recherche scientifique (CNRS) für die Entwicklung einer Kalibriervorrichtung eines Teleskops für die Astroteilchenphysik. Durch den Einsatz von schmierfreien Linearachsen von igus erreichen die Konstrukteure eine hohe Präzision und Wartungsfreundlichkeit. Der Sonderpreis für Bildungseinrichtungen und ebenfalls 1.000 Euro gehen an die wissenschaftliche-technische Universität Politecnico in Mailand. Sie hat mit dem ReBeL Roboterarm einen mobilen Manipulator konstruiert, der die Obsternte durch Automatisierung effizienter und weniger arbeitsintensiv gestaltet. „Die Gewinner beweisen, dass Automation heute nicht mehr nur eine Frage des Geldes ist“, so Mühlens abschließend. „Auch mit kleinen Budgets und Kreativität lassen sich wirtschaftliche Automationslösungen mit einem schnellen Return on Make investments realisieren. Wir freuen uns schon darauf weitere spannende und kostengünstige Automatisierungsprojekte beim nächsten ROIBOT Award kennenzulernen.“

Erfahren Sie mehr über den ROIBOT Award und die Gewinner auf:
https://www.igus.de/automation/service/gewinner-roibot

BurgerBot in Los Gatos, California, automates the fast-food line

Robotics

codesanitize

-

1 May 2025

0

BurgerBot in Los Gatos, California, automates the fast-food line

It is taking place: The robots are taking our jobs. No sick days, lavatory breaks, and no extra curly hairs in your buns. Simply chilly, arduous effectivity. Extra particularly, BurgerBot is a brand new fast-food joint the place robots are doing all of the work that people aren’t fascinated with, like burger meeting traces.

In Los Gatos, California, one of many San Francisco Bay Space’s extra prosperous areas, a shiny new fancy fast-food idea has simply popped up within one in every of its fashionable upscale brunch spots. ABB Robotics and BurgerBots have teamed up and unleashed a pair of IRB 360 FlexPickers and YuMi cobots (collaborative robots) to slap out some tasty burgers for the plenty – in 27 seconds, flat.

These machines do not simply stack US$18 all-beef patties, particular sauce, lettuce, cheese, pickles, onions on a sesame seed bun with surgical precision onto a QR-coded tray – they’re claimed to make completely constant burgers each single time with zero perspective. A far cry from little Billy, who’s by no means had a job earlier than and is on day one in every of coaching and experiencing rush hour for the primary time.

However earlier than we pull out the protest indicators and begin a picket line, here is some meals for thought: A full employees of people is employed on the restaurant. The bots (for now) solely deal with the burger manufacturing operation – from grinding the meat and griddling it as much as tossing it onto a conveyor belt meeting line. They then assemble the substances and kick out a whole, ready-to-eat burger again to a human server to be dished out to a ready visitor.

The robots do the repetitive stuff whereas supposedly leaving individuals extra time for hospitality and different people-y issues.

BurgerBots final assembly: an $18 burger, still in its QR-coded tray — BurgerBots remaining meeting: an $18 burger, nonetheless in its QR-coded tray

“The meals service business is dynamic and demanding, and our know-how brings industrial-grade consistency, effectivity and reliability to this area,” stated Marc Segura, President of ABB Robotics Division. “By taking up repetitive and time-consuming duties, robots permit employees to give attention to what issues most – creating memorable eating experiences.”

ABB surveyed 1,250 hospitality staff and located that 67% truly need robots to take over boring, gross, and harmful duties, and 63% had been excited by the prospect of a robotic making their job simpler. Ultimately, automation is not essentially about changing people, it is about upgrading all the system.

When the washer changed the washboard, individuals all over the place rejoiced.

BurgerBots is not only a Silicon Valley tech gimmick. It is designed for scalability, hygiene, and effectivity. ABB’s compact robotic cell combines the FlexPicker 360 – which grabs and stacks veggies and the like – and the YuMi robotic for remaining meeting as a field rolls down a conveyor. The system makes use of real-time stock monitoring from lettuce to condiments and all the things in between.

Robotic burger-making in 27 seconds!

BurgerBots has solely been open for roughly 24 hours on the time of writing, so time will inform the way it performs … although this is not ABB’s first foray into robotic meals prep. In 2021, ABB’s IRB 4600 ‘bot helped energy Roboeatz’s ARK (Autonomous Robotic Kitchen) in Latvia – claimed to be the world’s most superior autonomous kitchen that may whip up over 1,000 recipes from 80 recent substances.

In line with 2025 information from the World Financial Discussion board, automation and AI might result in the lack of roughly 92 million human jobs (about 8% of all jobs) by 2030. In direction of the highest of the potential record are positions like cashiers and fast-food staff. The primary duties to seemingly substitute human staff might be harmful, repetitive, or tedious jobs – and roles that do not significantly require excessive social or emotional intelligence.

That being stated, on the BurgerBot web site, the corporate is accepting resumes from certified people. Simply not for the burger-making place.

Supply: ABB Robotics

codesanitize

LiveKit and OpenAI with Russ d’Sa

Sponsors

Balancing productiveness and privateness: Safeguarding information within the age of AI-driven instruments

AI’s information privateness challenges

Information possession

Information extraction

Finest practices for guaranteeing information privateness

Search firms that don’t prepare on person information

Perceive your information privateness rights

Begin checking the phrases of service

Push for better regulation

Safeguarding privateness in AI

swiftui – iOS WKWebView exhibits about:clean after app stop and relaunch

Robots-Weblog | Inklusionsprojekt mit Low-Price-Roboter gewinnt ROIBOT Award von igus

BurgerBot in Los Gatos, California, automates the fast-food line