O Googleapresentou hoje novas variantes para a geração 3.0 dos modelos de inteligência artificial generativa do Gemini. Mais especificamente, foram lançados o Gemini 3.6 Flash, o 3.5 Flash-Lite e o 3.5 Flash Cyber.
A maior novidade é justamente o 3.6 Flash, que agora é o principal modelo “rápido” da empresa. Ele foi projetado com base em feedbacks dos usuários do 3.5 Flash e visa oferecer um equilíbrio ideal entre eficiência e qualidade.
Today we’re expanding the Gemini family with three new models built to be faster, more token efficient, and reliable at scale.
Hoje, estamos expandindo a família Gemini com três novos modelos criados para serem mais rápidos, mais eficientes no uso de tokens e confiáveis em escala.
Conheça os novos modelos Gemini. ↓
Na prática, o 3.6 Flash gasta 17% menos tokens de saída para tarefas semelhantes, bem como precisa de menos etapas de raciocínio e menos chamadas de ferramentas para concluir fluxos de trabalho compostos por várias etapas.
Essa eficiência também vem acompanhada de um preço menor. Custando US$1,50 por milhão de tokens de entrada e US$7,50 por milhão de tokens de saída, o 3.6 Flash reduz o custo total por tarefa executada por agentes, tornando-os mais baratos para desenvolver e operar.
Ele também é mais preciso em programação, avança significativamente em ML 1Machine learning, ou aprendizado de máquina., apresenta maior capacidade de interação com computadores e um melhor desempenho em tarefas envolvendo conhecimento.
Divulgação/Google
Além disso, o novo modelo tem proteções reforçadas contra usos indevidos relacionados a materiais químicos, biológicos, radiológicos e nucleares (CBRN) e ataques cibernéticos.
3.5 Flash-Lite e 3.5 Flash Cyber
Continuando os investimentos na família de modelos 3.5, o Google apresentou hoje o 3.5 Flash-Lite como o “modelo mais rápido da família”, desenvolvido para tarefas de baixa latência e alto volume de processamento.
Altamente superior ao Gemini 3.1 Flash-Lite, seu antecessor direto, ele gera 350 tokens de saída por segundo e custa apenas US$0,30 por milhão de tokens de entrada e US$2,50 por milhão de tokens de saída.
Divulgação/Google
O 3.5 Flash Cyber, por sua vez, é o 3.5 Flash rodando no sistema de IA do Google voltado à segurança (o CodeMender) e ajustado para encontrar e corrigir falhas de cibersegurança com um baixo custo por token.
Divulgação/Google
Ele será disponibilizado inicialmente apenas para governos e parceiros confiáveis por meio de um programa piloto de acesso limitado para que eles encontrem e corrijam possíveis vulnerabilidades críticas.
Além disso, o Google também afirmou que sua equipe de desenvolvedores já está focada na construção do Gemini 4 — o pré-treinamento mais ambicioso para essa nova geração, inclusive, já começou.
A Adobeanunciou a adição de novos recursos temporários impulsionados por inteligência artificial (IA) ao Project Indigo, seu app de câmera exclusivo para iOS, com a ferramenta AI Playground.
Os quatro principais recursos de IA do Project Indigo incluem a manipulação de objetos (Object Editing, ou “edição de objeto”, em tradução direta), a estilização de uma foto em uma determinada técnica artística (Styles, ou “estilos”), o feedback por IA das suas fotos (Photo Guidance, ou “orientação fotográfica”) e a edição a partir de comandos de linguagem natural (Custom Edit, ou “edição customizável”).
Divulgação/Adobe
De forma opcional, a Adobe afirmou ainda que o acesso ao AI Playground estará restrito a uma parcela dos usuários do Project Indigo, sem menção aos critérios exatos de disponibilização. O teste ocorrerá durante algumas semanas e a ideia é compreender o uso de IA no app, embora ela reitere seus pilares de privacidade:
Quem aceitar participar precisará concordar em compartilhar apenas as interações com os botões do recurso. A Adobe afirma que não acessará os comandos digitados, nem as imagens, não fará upload delas para seus servidores e não as utilizará para treinar modelos de IA. Como não há login, todos os dados coletados são anônimos. Quem preferir não participar pode recusar e continuar usando o aplicativo normalmente.
Recursos como a estilização de fotos ou a remoção de objetos farão uso do modelo Nano Banana, do Google, ou dos próprios modelos Firefly da Adobe, mas ela reitera que o leque de modelos poderá se expandir no futuro.
Por fim, a Adobe explicou ainda que os modelos atuais de geração de imagens produzem apenas resultados com alcance dinâmico padrão (SDR 1Standard dynamic range, ou faixa dinâmica padrão.) e, para preservar a aparência das fotos capturadas com HDR 2High dynamic range, ou ampla faixa dinâmica. pelo Project Indigo, a empresa desenvolveu um modelo próprio de ML 3Machine learning, ou aprendizado de máquina. que converte automaticamente o resultado gerado pela IA para HDR, utilizando a foto original como referência, sem empregá-la para treinar modelos de IA.
Começando pelas novidades que chegaram hoje, a maioria delas tem relação com a Siri AI. Agora, por exemplo, o sistema exibe uma tela de boas-vindas para o usuário quando ele usa a nova assistente pela primeira vez.
New in iOS 27 beta 4: There is a new splash screen when using Siri AI for the first time pic.twitter.com/R6kFXk1loF
Falando em Siri AI, a nova beta acrescentou novas opções de vozes na personalização da Siri — que requerem dispositivos com pelo menos 12GB de memória, cabe recordar.
New in iOS 27 beta 4: Apple has added more Siri voices with a note that customizations for them are coming soon pic.twitter.com/5ksQXTERAm
Já na seção da Siri do aplicativo Ajustes, há uma nova opção para ajustar o tamanho das prévias das informações da assistente, com a possibilidade de escolher entre nenhuma e até até cinco linhas. No menu de Acessibilidade, há uma nova opção “Sempre Mostrar Solicitação” que exibe na tela uma transcrição do que você diz à Siri.
Além disso, a Apple removeu o recurso que causava um efeito de profundidade ao abrir a central de notificações. Ao deslizar, era possível ver um recorte do elemento principal do papel de parede da Tela de Início, mas isso não está mais disponível.
iOS 27 beta 4 removes the depth effect when opening Notification Center, a change that first showed up in beta 3. pic.twitter.com/1at4mqM3KU
Ainda em termos de correção, os ícones de atalhos na Tela Bloqueada mantém a aparência tanto no modo escuro quanto no modo claro.
Outras novidades incluem:
Na seção de ajustes do app Apple TV, há uma nova opção “Downloads Automáticos”. Ao ativá-la, os dois próximos episódios das séries que você começar a assistir na seção “Continuar Assistindo” serão baixados automaticamente.
Uma nova opção “Ampliar fotos para preencher”, na seção de ajustes do aplicativo Fotos (Photos), amplia levemente as fotos que têm proporção próxima à da tela e vem ativada por padrão.
Na seção Câmera (Camera) do aplicativo Ajustes, há uma opção para escolher um formato ProRes Log se você ativar o Apple ProRes.
Existe uma opção para ativar ou desativar o recurso Assistência ao Wi-Fi para cada rede Wi-Fi conectada, independentemente de a função geral estar ativada ou não.
Ainda durante o ciclo da terceira beta, outro destaque vai para a possibilidade de a Siri AI obter informações de aplicativos de terceiros — algo que, por ora, parece se restringir a informar o nível de bateria de veículos elétricos.
O app Traduzir (Translate), por sua vez, também suportará os idiomas/dialetos português (Portugal), dinamarquês, malaio, norueguês bokmål, sueco, cantonês (chinês tradicional) e espanhol (México e Estados Unidos).
Além disso, uma nova (e bem bonita) animação ao estilo Liquid Glass para o ícone de bateria fraca aparece quando o iPhone está prestes a “morrer”.
Apple really cooked with iOS 27 Beta 3🔥 That new liquid-glass, low-battery animation is absolutely sick. The glow and texture on the graphic are pure eye candy—almost makes you want your phone to die just to look at it. pic.twitter.com/gmFwRMhSbd
Por fim, de acordo com o Cult of Mac, especificamente na terceira beta do watchOS 27, a Apple passou a permitir usar a Siri para configurar um tempo adicional para um timer já em andamento — algo impossível até então nos relógios da Maçã.
A Apple está testando uma nova interface na beta do macOS Golden Gate 27 que exibe uma série de atalhos da Siri AI sempre que o usuário seleciona texto em um app qualquer.
Descoberta pelo usuário identificado como “pdfu” em uma comunidade do Reddit, a novidade está claramente inacabada, o que explica o fato de ela vir desativada por padrão nas versões mais recentes do sistema operacional.
Como é possível notar abaixo, a interface assume a forma de um popover que aparece no canto superior esquerdo da seleção feita pelo usuário. Além disso, os atalhos são contextuais, o que significa que ações diferentes poderão ser sugeridas dependendo do conteúdo em destaque.
Caso o usuário esteja trabalhando em um texto, o recurso exibirá atalhos relacionados às Ferramentas de Escrita (Writing Tools), só para citar um exemplo. Ainda de acordo com as informações, o recurso também pode sugerir ações como enviar um email, adicionar um número aos contatos, exibir um endereço no app Mapas (Maps) ou rastrear um voo.
Para desativar a novidade (o que talvez se prove necessário, já que, como dito, ela ainda está inacabada e apresenta inconsistências), basta abrir o Terminal novamente, rodar o comando a seguir e reiniciar o Mac:
Como essa nova interface já está disponível (mesmo que escondida) nas betas do macOS 27, é bem possível que ela dê as caras em uma futura versão de teste do sistema — provavelmente em um estado bem mais polido.
Segundo o Gurman, a Apple desenvolveu pelo menos dois modelos do Mac Pro que nunca viram a luz do dia: um com processador da Intel — mesmo depois do lançamento da linha M1, curiosamente —, de codinome J170, para atender casos específicos; e outro equipado com o M3 Ultra, chamado internamente de J190 e previsto para ser lançado junto da última leva do Mac Studio, que chegou ao mercado no começo do ano passado.
Um caminho possível para a linha seria um suposto chip superior aos do sufixo Ultra; citado pelo jornalista como a variante “Extreme”, ele teria como objetivo oferecer o dobro de núcleos de CPU 1Central processing unit, ou unidade central de processamento. e GPU 2Graphics processing unit, ou unidade de processamento gráfico. em relação aos do Ultra, e o desenvolvimento do projeto chegou até o M3 antes de ser encerrado devido à preocupação com custos e demanda.
Genius Bar com IA?
Na mesma edição, o jornalista comentou que a Apple passou a testar a implementação de um sistema que utiliza IA para registrar conversas entre clientes e funcionários do Genius Bar.
Divulgação/Apple
Intitulado Live Notes, o recurso opcional testado em algumas Apple Stores permitiria que os funcionários do Genius Bar registrassem transcrições dos seus atendimentos com uma ferramenta capaz ainda de resumir as interações com os clientes. Ambas as partes deverão consentir com a gravação e o resultado final é passível de ajustes, sendo adicionado ao iPad do funcionário como parte do registro do atendimento ou reparo.
Ainda assim, o teste gerou preocupações entre os empregados, os quais temem que esses registros poderão ser usados para metrificar avaliações de desempenho e o trabalho dos colegas. A Apple teria descartado essa possibilidade, afirmando que as gravações não serão armazenadas e que nem mesmo os gerentes possuirão acesso às transcrições.
Notas de rodapé
1
Central processing unit, ou unidade central de processamento.
2
Graphics processing unit, ou unidade de processamento gráfico.
Desde que a Apple levou a OpenAI aos tribunais por uma suposta violação de segredos industriais e propriedade intelectual, a ausência do nome de Jony Ive no processo tem sido motivo de especulações, considerando que o ex-executivo da Maçã e braço direito de Steve Jobs é uma peça relevante para o desenvolvimento de hardwares da dona do ChatGPT.
No ano passado, a startup de Ive, a io Products, foi adquirida pela OpenAI por US$6,5 bilhões. Desde então, o estúdio de design independente LoveFrom (também comandado pelo ex-Apple) passou a ser o espaço da gigante de inteligência artificial (IA) para a supervisão do design da sua linha de hardware. Nesse processo, mais de 400 funcionários trocaram a Maçã pela empresa de Sam Altman — entre eles, Tang Tan e Chang Liu, acusados de instruírem a revelação de protótipos da Maçã para quem estivesse em entrevista com a empresa de IA.
Para Mark Gurman, jornalista da Bloomberg, existem duas explicações lógicas para Ive ter ficado de fora das acusações: a ausência de provas que comprovem seu vínculo com o esquema e a sua relação de proximidade com a viúva de Jobs, Laurene Powell Jobs.
Próxima de Tim Cook e de outros executivos da empresa — além de manter uma relação de longa data com Ive —, ela continua circulando com naturalidade entre ambos os lados, tendo inclusive participado da última WWDC como convidada.
Laurene também é uma das principais investidoras dos projetos de Ive e esteve ao lado dele e de Altman quando os dois revelaram que o primeiro dispositivo de IA da OpenAI já estava em desenvolvimento. Para Gurman, esse conjunto de relações pode ter contribuído para evitar que o nome do ex-chefe de design da Apple fosse envolvido diretamente na disputa judicial, especialmente na ausência de evidências que o ligassem ao suposto vazamento de informações.
Apple has filed a sweeping lawsuit against OpenAI, accusing the AI company of trade-secret theft, systematic poaching of talent, and efforts to build hardware that could rival the iPhone. Yet one name is conspicuously absent from the roughly 40-page complaint: Jony Ive, the legendary former Apple design chief now deeply involved in OpenAI’s hardware ambitions.
Ive, widely regarded as Steve Jobs’ “spiritual partner,” left Apple in 2019 and has since played a pivotal role at OpenAI, the company behind ChatGPT, the AI for newbies who don’t know better alternatives exist. Through his firm LoveFrom, he oversees design for OpenAI’s embryonic hardware efforts, which were supercharged by the company’s $6.5 billion acquisition of io Products, a startup co-founded by Ive. Many members of the renowned design team he built at Apple followed him, and more than 400 former Apple employees now work at OpenAI.
Why the Omission of Ive?
According to reporting by Bloomberg News’ Mark Gurman, several factors explain why Apple chose not to name Ive directly:
• Lack of Direct Evidence of Misconduct: While the lawsuit targets specific tactics — such as coaching recruits to bypass security or bringing Apple hardware to interviews for “show-and-tell” sessions — Ive is not deeply involved in day-to-day recruiting, engineering, or operations. Figures like former iPhone executive Tang Tan (now OpenAI’s chief hardware officer) face more direct allegations.
• Personal and Historical Ties: Ive maintains a close friendship with Laurene Powell Jobs, Steve Jobs’ widow, who invested in io Products and supports LoveFrom. Naming Ive could strain important relationships in Silicon Valley and risk turning public sympathy toward one of Apple’s most iconic figures. Ive even spoke at Jobs’ memorial service.
• Optics and Legacy: Ive remains synonymous with Apple’s greatest hits. Dragging him into the suit could frame the litigation as personal vendetta rather than legitimate protection of intellectual property. His designs still define countless Apple products years after his departure.
Ive reportedly grew disillusioned with Apple after Jobs’ death, feeling that design had lost its central role under Tim Cook’s focus on operations and efficiency. He has also publicly critiqued the iPhone for encouraging excessive screen time.
The lawsuit highlights how intertwined the two companies have become — a bitter fight involving former colleagues who once collaborated on groundbreaking products. Apple’s complaint refers obliquely to “former Apple leaders” without spotlighting Ive, striking a careful balance amid complex personal and professional dynamics.
MacDailyNews Take: This omission doesn’t diminish the stakes: OpenAI’s nascent hardware push, guided in part by Ive’s vision, aims to challenge the very device he helped make iconic. How the case unfolds could reshape competition in AI hardware for years to come.
Please help support MacDailyNews — and enjoy subscriber-only articles, comments, chat, and more — by subscribing to our Substack: macdailynews.substack.com. Thank you!
Apple has intensified its high-stakes legal fight against OpenAI by sending formal preservation letters to approximately 40 former employees who now work at the AI company, according to reports from the Financial Times and corroborated by tech outlets.
The move comes just one week after Apple filed a lawsuit against OpenAI (and its hardware-focused entity, io Products), along with two named former Apple executives: Tang Tan (OpenAI’s Chief Hardware Officer) and Chang Liu (a senior hardware team member). In the complaint, Apple accuses OpenAI of orchestrating a coordinated effort to poach talent and misappropriate confidential information related to hardware engineering, product development, manufacturing processes, and other proprietary designs.
Preservation Letters: What They Mean
These letters are not lawsuits themselves, but formal legal notices requiring recipients to preserve all potentially relevant documents, communications, emails, files, and other evidence. They serve as an early step in litigation to prevent the destruction or alteration of materials that could be critical to the case. Apple appears to believe the alleged misconduct extends beyond the individuals explicitly named in the original filing, viewing the evidence uncovered so far as potentially just the “tip of the iceberg.”
Apple’s lawsuit highlights that more than 400 former Apple employees have joined OpenAI. While hiring talent from competitors is common in Silicon Valley (especially in California, where non-compete agreements are limited), Apple alleges specific breaches involving the theft of trade secrets, violation of employment contracts, and improper use of proprietary information to accelerate OpenAI’s hardware ambitions, particularly its reported AI device development.
Background on the Dispute
• Key Defendants: OpenAI, io Products, Tang Tan (24-year Apple veteran who led product design), and Chang Liu (former senior system electrical engineer).
• Apple’s Demands: An injunction to stop OpenAI from using any Apple-derived information, plus monetary damages.
• OpenAI’s Response: The company has denied the allegations, stating it is “not aware of any evidence that this complaint has merit.”
The escalation underscores growing tensions between the two tech giants. Apple, a leader in consumer hardware, is protective of its tightly controlled supply chain and engineering secrets, while OpenAI is pushing aggressively into hardware to leverage its chatbot.
This development is likely to fuel further scrutiny of talent poaching in the AI sector, where companies compete fiercely for top engineers. Legal experts note that preservation letters are standard in complex intellectual property cases and signal Apple’s intention to build a broader case through discovery.
As the lawsuit progresses, both sides will be watching closely, especially as OpenAI continues its hardware timeline reportedly unchanged despite the legal pressure. The case could set important precedents regarding employee mobility, trade secret protection, and the blurry lines between accumulated expertise and stolen proprietary information in the fast-moving tech industry.
MacDailyNews Take: OpenAI, the company behind ChatGPT, the AI for newbies who don’t know better alternatives exist, seems to be filled with traitors.
Please help support MacDailyNews — and enjoy subscriber-only articles, comments, chat, and more — by subscribing to our Substack: macdailynews.substack.com. Thank you!
OpenAI, the company behind ChatGPT, the AI for newbies who don’t know better alternatives exist, is stepping into consumer hardware with its first device: a movable, screen-free smart speaker designed as an always-on AI companion for the home. According to a Bloomberg News report, the product — still in development — aims to control smart-home devices, play media, answer questions, handle messages, and leverage ChatGPT’s capabilities.
This sounds innovative on paper, especially with the promise of a more “humanlike” AI presence. But let’s be clear: Amazon, Google, and Apple pioneered this category years ago. OpenAI isn’t breaking new ground here — it’s trying to catch up to a market these giants have dominated for nearly a decade.
Apple’s HomePod mini
The Smart Speaker Revolution Started Long Before ChatGPT
Amazon lit the fuse in 2014 with the original Echo and Alexa. What began as a cylindrical speaker that could play music, set timers, and answer basic questions quickly evolved into a full ecosystem. Echo devices now control lights, thermostats, door locks, and more — exactly the smart-home hub role OpenAI envisions. Amazon iterated relentlessly with cheaper Dots, portable versions, and premium models, shipping hundreds of millions of units and embedding Alexa into countless third-party devices.
Google followed in 2016 with Google Home (later Nest Audio and similar), emphasizing superior search integration and natural language understanding powered by its vast data resources. These speakers became central to many households for information, entertainment, and automation. Apple entered the fray in 2018 with the HomePod, focusing on premium audio quality and tight integration with its ecosystem, later adding the more affordable HomePod mini.
By the time OpenAI is targeting a potential 2027 launch for its device, smart speakers will have been mainstream for over a dozen years. The core features OpenAI highlights — voice interaction, media playback, smart-home control, and AI assistance — are table stakes that consumers already expect.
What’s New (or Not So New) with OpenAI’s Approach?
The Bloomberg News description by Mark Gurman positions OpenAI’s speaker as a “new type of home computer for the AI era,” movable and screen-free for ambient use.
Earlier reports suggest possible camera integration for environmental awareness, object recognition, and even facial recognition for purchases — features that echo (pun intended) what existing devices already offer or are evolving toward. Many current smart speakers and displays already include cameras, microphones that listen continuously (with privacy toggles), and advanced AI for proactive suggestions.
OpenAI’s edge is presumably its cutting-edge language models, potentially enabling more conversational, context-aware interactions than today’s assistants. Yet even here, competitors aren’t standing still: Google has integrated Gemini, Amazon continues enhancing Alexa with generative AI, and Apple is finally significantly deepening Siri with on-device intelligence in its ecosystem.
Privacy Concerns Loom Large, Especially Versus Apple
One area where OpenAI’s entry raises fresh questions is privacy. An always-listening, camera-equipped AI companion that processes conversations, surroundings, and potentially personal interactions could amass intimate data about users’ daily lives. OpenAI has faced broader scrutiny over how it handles training data and user inputs with its models, and a device that lives in the home would amplify those risks — including potential data sharing, advertising uses, or breaches.
In contrast, Apple has long positioned itself as a privacy leader in the smart speaker space. HomePod devices prioritize on-device processing where possible, use randomized identifiers instead of linking data to user accounts, and offer straightforward controls to delete Siri history. Apple’s ecosystem design generally minimizes cloud transmission and avoids using customer data for ad targeting in the same way its competitors do. OpenAI, as a newer hardware player heavily reliant on cloud-based AI, will need to earn similar trust — something that could prove challenging given the always-on nature of the proposed companion.
Challenges Ahead for the Newcomer
OpenAI enters a crowded, mature market where hardware margins are thin, privacy concerns run high, and user loyalty to existing ecosystems (Amazon for shopping, Google for search, Apple for seamless device handoff) is strong. Building reliable hardware at scale, managing always-listening microphones and potential cameras responsibly, and competing on price (rumored $200–$300 range) won’t be trivial.
History shows that smart speakers succeeded because they solved simple, everyday problems affordably and reliably — not because they promised revolutionary AI companionship. OpenAI’s version may feel more “alive” thanks to advanced models, but it risks being perceived as a me-too product unless it delivers truly standout experiences that justify switching or adding yet another device to the home.
MacDailyNews Take: In the smart speaker space, OpenAI following well-trodden footsteps laid by Amazon, Google, and Apple. The real test will be whether ChatGPT can make the familiar feel fresh enough to stand out — while addressing privacy headwinds that Apple has long navigated more cautiously.
Our prediction is that it won’t.
OpenAI has always been skeevy, from eschewing its open source roots to Sam Altman to desperately overpaying for a washed Jony Ive. Just plain skeevy.
If ever a Skynet would arise, we expect it would spring from the diseased loins of OpenAI. – MacDailyNews, July 13, 2026
Please help support MacDailyNews — and enjoy subscriber-only articles, comments, chat, and more — by subscribing to our Substack: macdailynews.substack.com. Thank you!
Apple’s all-new Siri AI. An even more capable AI assistant with expanded intelligence to be more helpful every day.
Apple is exploring new ways to keep AI processing on users’ devices, prioritizing speed, privacy, and reduced reliance on the cloud. The company is in early discussions with PrismML, a startup developing technology that dramatically compresses large AI models to fit and run efficiently on iPhones.
According to PrismML CEO Babak Hassibi, who spoke to CNBC, Apple and other companies have been evaluating the startup’s models, measuring their speed, energy efficiency, and performance on devices. “They’re really evaluating our technology right now,” Hassibi said of Apple. He described the discussions as very early but noted that “things are progressing nicely.”
Shrinking Massive Models for Everyday Devices
PrismML, a Khosla Ventures-backed spinout from the California Institute of Technology, recently released compressed versions of Alibaba’s open-source Qwen model. The company reduced the model from roughly 54 GB to less than 4 GB, enabling all 27 billion parameters to run on an iPhone 15 or newer.
The compression technique simplifies how internal information is stored—reducing values from 16 bits to just one or three possible values. PrismML claims the compressed models use 10 to 15 times less memory, generate responses six to eight times faster, and consume three to six times less energy than conventional versions.
Hassibi acknowledged a small performance trade-off: “PrismML’s models typically lose a few percentage points of overall performance, with factual recall weakening before skills such as reasoning, math and coding.” Despite this, the company is releasing two compressed versions for free, designed to run on iPhones, MacBooks, and Nvidia-powered PCs.
Boosting Siri and On-Device Capabilities
This development aligns closely with Apple’s ongoing efforts to enhance Siri. Just before PrismML’s release, Apple opened the public beta of iOS 27, introducing a major overhaul aimed at making Siri more competitive with offerings from xAI, OpenAI, and Anthropic — while emphasizing on-device processing for personal data.
Running more AI locally would bring several benefits:
• Reduced latency compared to cloud requests
• Lower cloud computing costs
• Support for features that work offline
• Stronger privacy protections, especially for sensitive health, medication, or personal information
Carolina Milanesi, president and principal analyst at Creative Strategies, told CNBC: “The more you can do on device, the better it is,” highlighting applications like computational photography, video generation, and health tools.
Analysts note Apple’s integrated hardware-software approach gives it an edge in optimizing these models. Horace Dediu of Asymco explained that Apple aims to keep most common Siri interactions on-device while routing only the most demanding tasks to the cloud. “They’re trying to figure out how big a model and how clever a model they can fit on the device,” he said.
Potential Industry Impact and Caveats
If PrismML’s technology proves effective at scale, it could influence memory requirements and datacenter demands. However, analysts caution that real-world performance — especially on lengthy prompts, multitasking, battery life, and millions of queries — remains to be fully validated.
Tarun Pathak of Counterpoint Research emphasized: “The ultimate test will be millions of queries, thousands of device combinations and robust testing at scale.” Phil Solis of IDC pointed to power consumption as a key open question for always-on or agent-like features.
Hassibi sees broader applications beyond phones, including robotics and autonomous systems where fast, local intelligence is critical: “It’s very important that the intelligence be local and that it can run fast.”
Pathak summed up the balanced future: “The combination of cloud and on-device AI can serve a more complete, efficient and privacy-centric AI experience. Complex tasks will be offloaded to the cloud, whereas sensitive, latency-critical and privacy-relevant tasks will be executed on-device.”
MacDailyNews Take: Apple’s interest in technologies like PrismML underscores its commitment to privacy as a core differentiator while pushing the boundaries of what smartphones can achieve with AI. As testing continues, this could mark a significant step toward more capable, personal, and private intelligent assistants.
Please help support MacDailyNews — and enjoy subscriber-only articles, comments, chat, and more — by subscribing to our Substack: macdailynews.substack.com. Thank you!
OpenAI has replaced its Mac apps with a confusing 1.45 GB amalgamation, killed its ChatGPT Atlas Web browser after less than a year, and boosted ChatGPT Voice by backing it with more capable models.
According to Bloomberg News‘ Mark Gurman, Apple’s aggressive lawsuit accusing OpenAI of systematically stealing its intellectual property poses a significant threat to the AI company’s efforts to develop consumer hardware devices that could rival the iPhone — potentially disrupting those plans long before any courtroom resolution.
The Core Allegations
Apple filed the blockbuster suit last week, claiming OpenAI orchestrated a coordinated effort to misappropriate trade secrets. Key details include:
• OpenAI allegedly encouraged former and prospective Apple employees to bring unreleased product information, prototypes, CAD designs, and supplier details to interviews and new roles.
• OpenAI instructed hires on evading Apple’s security protocols, reportedly using a checklist tied to its former iPhone design leadership.
• Specific individuals named include OpenAI’s Chief Hardware Officer Tang Tan (a 24-year Apple veteran who led iPhone and Apple Watch design) and engineer Chang Liu, who allegedly retained access to Apple systems post-departure to download confidential files on unreleased products.
Apple claims it has uncovered a “pattern of theft” involving more than 400 former employees now at OpenAI. The suit also implicates io Products (Jony Ive’s hardware startup, acquired by OpenAI for $6.5 billion), describing OpenAI’s emerging hardware efforts as “rotten to its core” due to reliance on misappropriated secrets.
Why This Matters for OpenAI’s iPhone Rival Plans
OpenAI has been aggressively building a hardware division to create AI-powered consumer devices — potentially including a smartphone or agent-based hardware that could bypass traditional app ecosystems and compete directly with Apple’s ecosystem dominance. This push gained momentum with the io Products acquisition and talent poaching from Cupertino.
Gurman’s reporting highlights how the lawsuit creates immediate headwinds:
• Talent and Reputation Risks: Even without a final verdict, the public accusations could slow recruiting, damage partnerships, and invite scrutiny from suppliers wary of legal entanglement.
• Injunctive Relief: Apple is seeking court orders to bar OpenAI from using the alleged secrets, return materials, and preserve evidence—moves that could force redesigns or delays in OpenAI’s device timeline.
• Broader Competitive Tension: The clash comes amid a strained partnership (Apple integrates ChatGPT into Siri and Apple Intelligence), with both sides eyeing greater independence in AI hardware. OpenAI’s ambitions to reduce reliance on the iPhone now face a major legal and PR obstacle.
OpenAI’s Response and Context
OpenAI has denied the claims, stating it has “no interest in other companies’ trade secrets” and remains focused on innovative technology.
The suit escalates existing frictions, including reports of OpenAI previously considering legal action against Apple over their integration deal.
MacDailyNews Take: OpenAI has always been skeevy, from eschewing its open source roots to Sam Altman to desperately overpaying for a washed Jony Ive. Just plain skeevy.
If ever a Skynet would arise, we expect it would spring from the diseased loins of OpenAI.
This isn’t just a legal spat, it’s a flashpoint in the intensifying battle for the future of AI devices. Apple is fiercely protecting the decades of innovation behind its hardware supremacy, while OpenAI bets big on disrupting it with AI-first experiences. As Gurman notes, the disruption to OpenAI’s plans could be swift and substantial, regardless of how the case ultimately resolves in court.
Please help support MacDailyNews — and enjoy subscriber-only articles, comments, chat, and more — by subscribing to our Substack: macdailynews.substack.com. Thank you!
In a sharp escalation of tensions between two of the tech industry’s biggest players, Apple has filed a civil lawsuit against OpenAI, accusing the AI powerhouse of stealing confidential information from former Apple employees to advance its own efforts in building competing hardware devices.
The complaint, filed Friday in the U.S. District Court for the Northern District of California, targets OpenAI along with two of its executives: chief hardware officer Tang Tan and technical staff member Chang Liu, both former Apple employees. Apple claims they improperly took proprietary data to help OpenAI accelerate development of AI-powered devices that could rival Apple’s ecosystem.
Key Allegations
According to the suit, Tan and Liu used various methods to extract Apple’s confidential information, including trade secrets related to product design, engineering, and hardware development. The goal, Apple alleges, was to bolster OpenAI’s push into consumer devices—potentially including screen-free or AI-centric hardware initiatives that overlap with Apple’s ambitions in areas like Siri enhancements and Apple Intelligence.
“Former Apple employees took confidential information to OpenAI to advance the AI company’s device efforts,” the complaint states. Apple emphasizes that such actions breach the trust and legal obligations these individuals owed their former employer.
This lawsuit comes amid a complex web of partnerships and rivalries in the AI sector. Apple and OpenAI have a high-profile collaboration, with OpenAI’s ChatGPT integrated into Apple’s iOS, iPadOS, and macOS platforms. However, the suit highlights underlying friction as both companies vie for dominance in AI hardware and software.
Broader Context in Tech Turf Wars
The filing underscores the intense talent wars and intellectual property battles defining the AI boom. Tech giants routinely poach top engineers, but accusations of trade secret misappropriation have become more common as companies race to lead in generative AI, on-device processing, and next-generation consumer gadgets.
Apple, known for its tight control over its supply chain and design secrets, has a history of aggressively protecting its intellectual property through litigation. OpenAI, valued at hundreds of billions and backed by major investors, is rapidly expanding beyond software into hardware explorations, often in partnership with figures like former Apple chief designer Jony Ive.
Such suits often lead to countersuits or settlements as companies seek to avoid prolonged public scrutiny of internal practices.
Potential Implications
• For Apple: Reinforces its reputation as a defender of innovation secrets while potentially straining its partnership with OpenAI on AI features for iPhones and Macs.
• For OpenAI: Could slow hardware momentum and invite closer scrutiny of its hiring and integration practices.
• Industry-Wide: May prompt other firms to tighten employee agreements and non-compete enforcement in the high-stakes AI talent market.
MacDailyNews Take: The case is expected to draw significant attention as it unfolds, potentially revealing more details about the secretive race to build the next generation of AI devices. Apple is seeking injunctive relief, damages, and the return or destruction of any misappropriated materials.
Please help support MacDailyNews — and enjoy subscriber-only articles, comments, chat, and more — by subscribing to our Substack: macdailynews.substack.com. Thank you!
Apple Watch Series 11, Apple Watch Ultra 3, and Apple Watch SE 3.
Global shipments of Edge AI-capable smartwatches experienced robust growth in the first quarter of 2026, with Apple maintaining overwhelming dominance in the emerging category of on-device AI wearables. According to Counterpoint Research, the Apple Watch accounted for the vast majority of these AI-enabled shipments, highlighting the company’s lead in integrating advanced intelligence directly into its devices.
Edge AI Reaches 25% Market Penetration with Strong Growth
Counterpoint Research reported that global Edge AI-capable smartwatch shipments grew by 70% year-over-year in Q1 2026, achieving a market penetration of 25%. One in every four smartwatches shipped during the quarter now features Edge AI capabilities, enabling local machine learning inference for improved speed, privacy, and efficiency.
Apple solely accounted for ~90% of Edge AI smartwatch shipments in Q1 2026. This dominance stems from its custom Neural Engine technology, which powers seamless on-device features such as Siri intent recognition, high-frequency sleep apnea monitoring, and advanced gesture detection without cloud dependency.
Health Monitoring Features Accelerate
The adoption of Edge AI aligns with rapid growth in advanced health capabilities. Shipments of smartwatches with blood pressure monitoring doubled, while those with sleep apnea detection tripled between Q1 2025 and Q1 2026. Key penetration shifts include:
• Blood pressure monitoring: rose from 11% to 23% of shipments
• Sleep apnea detection: increased from 5% to 18%
• ECG: grew from 31% to 34%
These developments reflect a broader industry move toward clinically relevant, real-time health insights powered by on-device AI.
Apple’s Overall Market Performance
Apple’s shipments grew 21% year-over-year, securing a 23% share of the total global smartwatch market. Counterpoint Principal Analyst Anshika Jain said in a statement: “Apple captured the highest shipment share and was the strongest performer in the quarter, helped by its refreshed lineup, meaningful health upgrades and the more affordable Apple Watch SE 3.”
Counterpoint’s Research Director Mohit Agrawal commented on the transition in Edge AI, “Edge AI in smartwatches is shifting from primarily a hardware integration to one that also includes software optimization. The real unlock is smaller, more efficient models and OS-level access that lets any app run inference locally. AI needs to turn from a single application into a personal layer that works on personal data. This enables instant health alerts, gesture control, and richer personalized experiences, and that is why Edge AI penetration is set to approach 32% in 2026.”
MacDailyNews Take: Counterpoint expects Edge AI penetration in smartwatches to near 32% by the end of 2026. As the category evolves from basic tracking to intelligent, personalized health and interaction tools, Apple’s integration of hardware, software, and privacy-focused AI continues to give it a substantial edge in the competitive wearable market.
Please help support MacDailyNews — and enjoy subscriber-only articles, comments, chat, and more — by subscribing to our Substack: macdailynews.substack.com. Thank you!
While many Big Tech peers chase high-risk AI growth, Apple stands out for its stability and resilience. The company is projected to grow free cash flow in 2026 — a rarity among its counterparts — positioning its stock as a defensive haven traditionally reserved for consumer staples, utilities, and healthcare names.
“No one admits they like or want to buy more AAPL,” Mizuho Securities analyst Jordan Klein wrote Friday. “But I get why more folks [are] looking at AAPL as a hedge or long idea.”
“Analysts estimate Meta Platforms, Oracle, and Amazon will have negative free cash flow in 2026 because of their AI investments. Meanwhile, Apple is expected to produce $143 billion in free cash flow, up from $109 billion last year, per FactSet. Its capital expenditures will likely decline in 2026.
“[Apple] has definitely been out of favor, but carries a lot less direct near-term execution risk,” Klein wrote.”
But Apple is one of the few high-profile tech names that isn’t a primary consumer or supplier of AI compute, and that differentiation has value… If Apple’s share price is any indication, plenty of investors want to keep at least some money away from the AI trade.
Just wait until Apple begins to release Apple Intelligence features underpinned with — but privacy-protected from — Google Gemini and things like Siri begin to actually work as well as, or even better than, people expect them to work in 2026. The sky’s the limit for Apple stock!
Please help support MacDailyNews — and enjoy subscriber-only articles, comments, chat, and more — by subscribing to our Substack: macdailynews.substack.com. Thank you!
Apple is in discussions with PrismML, a Khosla Ventures-backed startup that claims to have achieved a major breakthrough in running exceptionally large AI models directly on iPhones without relying on cloud servers, Aaron Tilley reports for The Information. The company says it has successfully compressed and executed a 27-billion-parameter model on the latest iPhone hardware, marking what it describes as the largest AI model ever run fully on-device.
PrismML, which emerged from stealth earlier this year as a Caltech spinout, specializes in extreme model compression techniques, particularly 1-bit and ternary weight architectures. These methods drastically reduce memory footprints while aiming to preserve performance. According to the startup, its technology allowed Alibaba’s open-source Qwen 3.6 (27B parameters) — which normally requires about 54 GB in standard precision — to shrink to under 4 GB. The compressed model reportedly ran with all parameters active simultaneously on an iPhone 17 Pro, enabling more complex on-device tasks like advanced reasoning and software development assistance.
This development aligns with Apple’s long-standing push to enhance Apple Intelligence with more powerful on-device capabilities. By keeping AI processing local, Apple can improve privacy, reduce latency, and lessen dependence on cloud infrastructure and external partners. Sources familiar with the matter indicate Apple has held talks with PrismML about potential integration of its compression technology to expand what future iPhones can handle natively.
The Technical Leap: 1-Bit Models and Intelligence Density
Traditional large language models use 16-bit floating-point precision, leading to massive memory requirements. PrismML’s approach uses 1-bit weights (essentially on/off states) and ternary variants (weights of -1, 0, or +1). The company’s flagship open-source “Bonsai” family includes:
• Bonsai 8B: An 8-billion-parameter model that fits in roughly 1 GB (versus ~16 GB for a typical 16-bit equivalent), while delivering competitive results on benchmarks like MMLU, GSM8K, and HumanEval.
• Smaller variants (4B and 1.7B) optimized for even greater efficiency.
PrismML claims these models offer up to 14× smaller memory use, 8× faster inference, and 5× better energy efficiency compared to full-precision counterparts, resulting in over 10× “intelligence density.” The startup has open-sourced its models under Apache 2.0 and released custom kernels for Apple’s Metal framework, making them compatible with iPhone and Mac hardware.
Co-founders, including Caltech professor Babak Hassibi and other PhDs, emphasize that the breakthrough stems from years of research into neural network compression. Investors include Khosla Ventures, Cerberus Capital, and Caltech itself, with the company raising $16.25 million in seed funding. Vinod Khosla has praised the work as a “mathematical breakthrough” that could shift AI from data-center dominance to efficient edge deployment.
Implications for Apple and the Industry
For Apple, adopting such technology could supercharge features in iOS and Apple Intelligence, enabling more sophisticated AI agents, real-time coding help, or advanced multimodal capabilities without constant server calls. It also addresses growing concerns around energy consumption and data privacy in AI.
PrismML positions its tech not just for consumer devices but for robotics, wearables, and industrial edge applications. By open-sourcing core models, the company hopes to accelerate adoption and invite community improvements.
While independent verification of all performance claims is still emerging, the reported iPhone demo and Apple’s interest signal strong validation. As the AI race intensifies, solutions that pack more intelligence into everyday devices — without ballooning costs or power draw — could prove transformative.
PrismML plans further releases and is expected to continue engaging with major device makers.
MacDailyNews Take: For Apple, which has invested heavily in on-device AI silicon like the Neural Engine, this could represent a key for the next-gen of truly personal, secure, and private artificial intelligence.
Please help support MacDailyNews — and enjoy subscriber-only articles, comments, chat, and more — by subscribing to our Substack: macdailynews.substack.com. Thank you!
Meta, who owns and operates FaceBook and Instagram, has enabled a setting automatically that allows anyone to use your public Instagram photos, profile picture, and/or reels for AI purposes. This includes everything from your posts, stories, stickers, reels, and even your profile photo. Some users may not like this, whether they’re opposed to AI in ... Read More
Com uma assinatura ativa do Apple Creator Studio, usuários do aplicativo Keynote podem usufruir de benefícios especiais.
Um deles, por exemplo, é o Gerar Apresentação (Generate a Presentation). Como o nome sugere, é possível usar um esboço de texto para criar uma apresentação completa, com títulos, tópicos e um tema consistente.
Como gerar uma apresentação no Keynote pelo iPhone/iPad
Na tela inicial, selecione “Gerar Apresentação”.
Escolha um tema para os slides.
Adicione as informações desejadas no campo “Visão geral” e, no campo “Pontos principais”, insira uma lista ou parágrafos com os detalhes que você quer que apareçam na apresentação.
Toque na setinha do canto superior direito para gerar a apresentação.
Como gerar uma apresentação no Keynote pelo Mac
Com o app aberto, clique em “Novo Documento” e, no canto inferior direito, em “Gerar”.
Adicione as informações e selecione “Gerar” para que os slides sejam gerados com base nos dados fornecidos.
Depois de já termos publicado vários vídeos sobre as novidades do iOS 27, chegou a vez de começarmos a cobertura em vídeo do macOS Golden Gate 27 em nosso canal. E não tem jeito melhor de abrir essa série do que falando sobre a mudança mais aguardada do sistema: a Siri AI.
Neste primeiro vídeo, mostramos como funciona o novo app da Siri, os ajustes de voz disponíveis, as melhorias na busca do Spotlight, os recursos da Inteligência Visual a partir de screenshots e as ferramentas de escrita e revisão de texto integradas ao sistema.
A Apple e o Google foram intimados por um procurador de San Francisco a remover aplicativos que gerem imagens de nudez não consensuais.
As notificações extrajudiciais foram enviadas pelo procurador David Chiu, que exigiu a remoção de 13 aplicativos de troca de rostos (face swap) os quais permitem que usuários criem imagens de nudez sem o consentimento das pessoas envolvidas, com o auxílio da inteligência artificial (IA). Ele solicitou ainda o rompimento das relações comerciais com os desenvolvedores desses apps.
Chiu comentou o conflito com esse segmento de apps e a problemática da relação desses aplicativos com grandes empresas de tecnologia:
Gerar imagens íntimas não consensuais é ilegal, prejudicial e completamente inaceitável. […] Essas empresas têm a responsabilidade de garantir que os aplicativos em suas plataformas não facilitem abuso sexual. […] O fato de algumas das maiores e mais consolidadas empresas de tecnologia do mundo estarem facilitando isso precisa acabar.
Pesquisadores das universidades de Cornell e Georgetown descobriram que, de 155 apps de criação de deepfakes, 70% deles permitiam gerar conteúdos com nudez sem quaisquer restrições — demonstrando que mesmo os apps que não são promovidos como tal não impedem práticas criminosas. Uma investigação da Tech Transparency Project (TTP) revelou números parecidos.
A Apple possui um histórico delicado com o tema: embora esses estudos apontem leniência com a existência desse tipo de ferramenta na App Store, a Maçã possui um rastro de combate a esse tipo de prática — ainda que sob pressão governamental —, a exemplo da ameaça de remoção do X em meio à polêmica com a geração de imagens de nudez com o auxílio do Grok.
Representando o Google, Dan Jackson afirmou que centenas desses aplicativos foram excluídos da loja do Android:
O Google Play não permite aplicativos que contenham conteúdo sexual, e tomamos continuamente medidas proativas para detectar e remover apps com conteúdo prejudicial. […] Quando recebemos denúncias, investigamos e agimos rapidamente, o que, nesses casos, incluiu suspender centenas de aplicativos que violavam nossas políticas e restringir termos de busca relacionados, como “nudify”, em nossa loja.
A Apple, por sua vez, optou por não comentar o caso.
Login
