O Gemini, do Google, ultrapassou oficialmente a marca de 1 bilhão de usuários mensais ativos — tornando-se, segundo a própria gigante de Mountain View, o produto de crescimento mais rápido na história da empresa.

Para colocar esse marco em perspectiva, o Gmail levou cerca de 12 anos para conquistar uma base de usuários desse tamanho. Isso também coloca o Gemini no mesmo patamar do ChatGPT (da OpenAI), que atingiu essa marca no início de junho.

O CEO ¹Chief executive officer, ou diretor executivo. do Google, Sundar Pichai, utilizou o X para destacar esse marco.

1B+ people are now using @Geminiapp every month to spark new ideas and get things done. It’s our fastest growing product ever, and our 14th to hit the 1B-user mark.

Kudos to @JoshWoodward & the entire Gemini team, and thank you to everyone who has been on this journey with us -… pic.twitter.com/sSWrULX4Qn

— Sundar Pichai (@sundarpichai) August 11, 2026

Além do número de usuários, o Google divulgou detalhes sobre como as pessoas interagem com o chatbot no dia a dia. Entre eles, a empresa aponta que os usuários dos dispositivos da Apple estão adotando a ferramenta em grande escala, com mais de 100 milhões de usuários ativos no iOS. Além disso, usuários avançados do macOS acionam o Gemini cerca de duas vezes mais frequentemente do que em outras plataformas.

Eles apontam ainda que interação por voz tornou-se a forma preferida de usar o aplicativo, com 63% dos usuários optando por falar em vez de digitar. Ademais, uma em cada cinco sessões do Gemini Live vai além do áudio simples, envolvendo o compartilhamento de imagens da câmera/tela em tempo real.

Estudantes representam outro grupo importante de usuários: 38% das solicitações acadêmicas incluem o envio de arquivos de tarefas, anotações ou PDFs. A geração de imagens também gera um tráfego enorme, sendo responsável pela criação de mais de 150 milhões de imagens diariamente.

Notas de rodapé

• 1
  Chief executive officer, ou diretor executivo.

O 9.º Tribunal de Recurso dos EUA recusou ontem o recurso das empresas tecnológicas, concluindo que a Secção 230 oferece uma defesa de responsabilidade, e não imunidade total. O julgamento que envolve 29 procuradores-gerais estatais começa esta quarta-feira em Oakland.

The post Tribunal dos EUA obriga Meta, TikTok, Google e Snapchat a enfrentar mais de 3.000 processos judiciais por dependência das redes sociais appeared first on Tek Notícias.

A new analysis of three skull fragments unearthed at the Gran Dolina cave site in Spain indicates that Homo antecessor had bone thickness closer to that of Neanderthals and modern people than to more archaic hominins.

The post Homo antecessor Had Surprisingly Modern Skull, New Fossils Suggest appeared first on Sci.News: Breaking Science News.

O Bluesky anunciou um novo recurso que permite ocultar as repostagens de um determinado usuário no feed. Aos moldes da ferramenta equivalente do X, ele permite que a sua página inicial não seja inundada com compartilhamentos de posts de terceiros feitos por pessoas que você segue.

Para aplicar a ação a um determinado usuário, basta abrir as opções do perfil (os três pontos ao lado do botão “Seguindo”) e selecionar “Hide reposts in feeds”.

v1.130 is live!Want to follow someone for their posts but not their reposts? Now you can — turn on "Hide reposts in feeds" for any account you follow.

— Bluesky (@bsky.app) 2026-08-10T18:23:59.962Z

A opção foi liberada na atualização mais recente (como mencionada, a versão 1.130.0) do Bluesky na App Store e na versão web da rede social.

Ao The Verge, os desenvolvedores anunciaram que estão lançando um recurso beta que exibe automaticamente uma contagem de postagens para threads (seu equivalente da Meta, o Threads, também conta com essa opção). Ele pode ser habilitado nas configurações do app.

Apesar de contar com um aplicativo dedicado para Macs, o WhatsApp oferece uma experiência acessível por meio de qualquer navegador.

Batizado de WhatsApp Web, ele pode ser utilizado para trocar mensagens, fazer chamadas de áudio ou de vídeo e até conversar com a Meta AI, por exemplo.

Recentemente, o site trouxe um novo recurso: agora, é possível usar a integração com o Adobe Acrobat para abrir arquivos PDF diretamente por lá, sem precisar baixar todo o software. Você também pode fazer algumas edições simples, como destacar ou adicionar comentários, por exemplo.

Veja como usar isso! 👨‍💻

No seu navegador, abra o site do mensageiro.

Selecione, então, a conversa onde o documento está localizado e clique em cima dele. Na parte superior direita, selecione o menu suspenso “Editar PDF” e escolha “Editar no Acrobat”.

Confirme a ação para ser levado a uma nova aba.

Após fazer as edições, vá até “Enviar para WhatsApp” (no canto superior direito) para compartilhar o documento com alguém no mensageiro.

The 2016 discovery of a large mass grave in the ancient cemetery of Phaleron, on the coast of Athens, shocked the archaeological community and the general public. There lay the remains of 79 adult men, many still with metal shackles, arranged in three parallel rows. Since then, researchers have debated who these individuals were: pirates, […]

A Anthropic anunciou nos últimos dias duas novidades bem interessantes para o Claude Code, sua plataforma agêntica de codificação com inteligência artificial (IA) generativa: as conversas entre diferentes sessões e o modo automático como padrão para alguns de seus planos.

Começando pela primeira novidade supracitada, ela permite que o Claude envie informações de uma sessão para outra num mesmo computador ou responda a sessões em outras máquinas — algo que pode ser bastante útil quando mais de uma sessão está trabalhando num mesmo projeto.

O recurso pode ser usado, por exemplo, quando há uma mudança que quebra o sistema como um todo e cuja resolução depende de algo executado em outra sessão — ou simplesmente quando alguma delas precisa saber de uma decisão de design ou de um resultado de testes.

Por ora, no entanto, apenas texto simples é suportado, de modo que nenhum arquivo ou o histórico da conversa é transferido entre as diferentes sessões. Uma sessão também não pode aprovar comandos originados de outras sozinha — sempre é requerida a autorização do usuário para tal.

New in Claude Code: your sessions can now message each other.

Instead of having to re-explain yourself in another session, you can now tell Claude to do it. It sends a summary (not your history or files), and the other session picks it up mid-task. pic.twitter.com/PtNsfXeQXP

— ClaudeDevs (@ClaudeDevs) August 7, 2026

O Claude pode decidir enviar uma mensagem por conta própria se julgar necessário, ou o usuário pode solicitar explicitamente que ele informe algo a outra sessão — inclusive, é possível definir se elas devem ser aceitas ou recusadas automaticamente ou se devem requerer aprovação.

Modo automático como padrão

Um dos recursos mais interessantes do Claude Code é o modo automático, que permite à plataforma usar uma espécie de “classificador de segurança” para decidir se prossegue ou não com determinada tarefa, sem precisar a todo instante solicitar a permissão do usuário.

Com o modo automático, que antes era opcional, o Claude pode trabalhar sozinho por períodos muito mais longos e sem qualquer interferência humana — o que permite que seus modelos realizem tarefas complexas durante horas e entreguem um trabalho mais completo ao fim da sessão.

Pesquisas feitas pela Anthropic indicaram que o modo automático também tende a deixar o desenvolvimento mais seguro, já que desenvolvedores costumam aprovar 97% dos pedidos de permissão por reflexo (sem revisão), e bloqueiam bem menos prompts perigosos que a ferramenta (acima).

Com isso, esse modo agora passará a ser o modo padrão do Claude Code nos planos Pro, Max e Team — a partir de 14 de agosto. Como consequência dessa mudança, a Anthropic também deixará de cobrar dos usuários desses planos pelo custo computacional desse classificador.

Juli Clover (release notes, security, no enterprise, no developer, full installer, IPSW):

macOS Tahoe 26.6.1 fixes a vulnerability that could allow an attacker to authenticate to Screen Sharing without valid credentials.

See also: Adam Engst and Mr. Macintosh.

Previously:

• macOS 26.6

Update (2026-08-14): Bill Toulas:

In an update to the initial advisory, the Dutch agency said it received a report indicating that the vulnerability is being exploited in the wild in attacks where port 5900 is exposed to the internet.

According to the NCSC, the attacker obtained root access to the system and deployed a Monero cryptocurrency miner.

Update (2026-08-20): Edovia (tweet, MacRumors):

The vulnerability, identified as CVE-2026-65400, could allow an attacker on the network to authenticate to Screen Sharing without valid credentials. Reports now indicate that the vulnerability has been actively exploited on Macs where Screen Sharing was accessible from the Internet.

[…]

[We] strongly recommend that all Screens users install the latest macOS updates as soon as possible.

[…]

For users who want to further restrict access to Screen Sharing, Screens can connect to a Mac through an SSH tunnel instead of exposing the Screen Sharing service directly.

Talk about homework: a large class of undergraduates at an unnamed university in Utah became the unfortunate subject of a severe public health response after an extraordinary bungle in a science lab.

In September of last year, the group of lab students were tasked with identifying a bacterium known as P. shigelloides, a relatively harmless pathogen known to cause diarrhea when ingested. But as lab results came in, students were alarmed to find the specimens handed out in class were actually the bacteria Neisseria meningitidis, a potentially deadly pathogen that causes meningitis.

Unlike P. shigelloides, the bacterium N. meningitidis calls for high-security practices, like double gloves, disposable clothes, and fit-tested N95 masks, as the bare minimum. (In one unfortunate case from 2012, a 25-year old microbiologist whose lab was ill equipped to handle N. meningitidis died mere hours after exposure to the aerosolized microbe.)

According to People, Utah public health authorities soon raided the Utah stockroom, finding samples labeled “N. meningitidis.” That discovery prompted a recommendation that all 33 of the exposed students begin a cautionary antibiotic regimen, especially given that the lab participants were only wearing gloves and lab coats at the time of exposure.

On top of the antibiotics dose, at least two people sought medical attention due to possible exposure symptoms, which included a painful and invasive spinal tap.

It was only after the frenzy that further professional lab tests identified the offending pathogen as Neisseria sicca, a bacterium “commonly found in the upper respiratory tract” that is “nonpathogenic in immunocompetent persons,” according to a recently-filed Center for Disease Control study of the Utah lab incident.

Student misidentification notwithstanding, the event underscores the importance of accurately labeling microbial cultures — lest aspiring biologists become victims of their own studies.

More on pathogens: Meta’s AI Data Center Caught Leaking Deadly Bacteria Into Water Town Uses for Irrigation

The post 33 Students Treated for Deadly Bacteria Exposure When Classroom Experiment Goes Horribly Wrong appeared first on Futurism.

Neanderthals vanished roughly 42,000 years ago, but a piece of their biology appears to be quietly at work in some people today, according to a new study led by scientists from the Max Planck Institute for Evolutionary Anthropology.

The post Neanderthal Gene Still Shapes Muscle and Jaws in Some Living Humans appeared first on Sci.News: Breaking Science News.

O sistema operativo Android conta com ferramentas que ajudam a criar uma espécie de “modo de férias” no smartphone, permitindo silenciar o que é menos importante neste momento e definir as notificações ou chamadas que realmente quer receber nas férias. Veja como configurar.

The post Quer ter um verão mais descansado? Veja como colocar o smartphone em “modo de férias” appeared first on Tek Notícias.

Even notorious anti-vaxxer and US health secretary Robert F Kennedy Jr had to urge families to vaccinate their children against measles last week.

Despite being an entirely preventable disease thanks to a highly effective vaccine that has been widely available since the 1960s, cases of measles have skyrocketed in recent years, facilitated by a White House that’s declared war on scientific research, made a mess of childhood vaccine mandates, and furthered dangerous misinformation.

As of last week, the US had recorded 2,465 confirmed infections so far this year, already more than the 2,289 total cases in 2025. The last two years saw more total infections than the rest of the 21st century combined, as The Guardian reports, with 2026’s tally beating a 35-year record.

Now, in the midst of all that crisis crisis, a character has stumbled into the narrative who perfectly illustrates how it’s gotten so bad. As the LA Times reports, a measles-infected individual spent a whopping 11 hours at the Universal Studios Hollywood theme park last month, potentially exposing countless other visitors.

The Los Angeles County Public Health Department was forced to issue a stark warning: anybody who visited the amusement park on July 26 during teh hours between 10am and 9pm may have been exposed and should monitor for symptoms including fever, cough, or rash — and stay vigilant until August 16.

The virus is contagious up to four days before the characteristic rash appears on the skin, and up to four days after that.

Measles can easily spread through the air through just via talking or breathing, and can persist on surfaces for hours at a time.

The vaccine against it, available as part of the measles, mumps, and rubella combination (MMR) shot, is extremely worthwhile. Two doses are 97 percent effective in preventing a measles infection.

The department also urged “everyone to confirm their immunity and get the vaccine if needed, especially before traveling and attending crowded events.”

California has been hit particularly hard, recording at least three localized measles outbreaks this year so far. However, according to the state’s health department, all three outbreaks have ended.

According to the Centers for Disease Control and Prevention, there have been 38 new outbreaks in 2026 across the country so far this year, with several thousand cases spanning 47 states.

The danger to one’s health is substantial — especially to the unvaccinated. According to the CDC, roughly one in five unvaccinated people in the US who get measles are hospitalized. One out of 20 children with measles develop pneumonia. and one to three of every 1,000 children who become infected die from respiratory and neurologic complications.

More on infectious diseases: This Measles Outbreak at a Detention Center Perfectly Encapsulates America Right Now

The post Authorities Alarmed When Person With Measles Spends Entire Day Frolicking Around Universal Studios appeared first on Futurism.

In case you weren’t already feeling paranoid about how AI is shaping our future, scientists for the first time have used an AI model to create new and viable viruses that have never been seen in nature.

Before you panic, the work, detailed in a study published in the journal Science, involved creating viruses that only target bacteria and are harmless to humans. Bacteriophages, as they’re called, are an important medical intervention for targeting drug-resistant pathogens.

“This is an important milestone,” Patrick Cai, a synthetic biologist at the University of Manchester who was not involved in the study, told The New York Times.

Still, it’s an alarming testament to AI’s power and trajectory. The researchers at Stanford University and the Arc Institute trained an AI model to recognize patterns of genetic structure in nature to create new DNA molecules. These molecules were inserted into bacteria that, over time, turned into wholly new viruses, at the expense of killing their hosts.

In their research, which first gained attention last year but has now been peer-reviewed, the scientists used an AI model called Evo. Unlike a general purpose large language model, Evo is trained on billions of lines of genetic sequences, and for this work in particular, millions of bacteriophage genomes. But it still functions like an LLM, predicting what DNA building blocks — nucleotides — should follow each other like a chatbot predicts what words to spit out to finish a sentence.

The researchers chose to focus on a E. coli-killing virus called Phi X-174, a variant that’s so well studied that it serves as the equivalent of a lab rat in virology. Viruses in general are simple compared to complex cellular life, making them an ideal platform to test AI’s capabilities.

They used Evo to create new viruses based on Phi X-174, and it generated 700,000 candidates. They then narrowed it down to just 300 that they would recreate in the lab. Of those 300, 16 proved to be viable, infecting and killing their bacterial hosts.

There’s clear medical merit to using AI to create bacteria-slaying viruses, but there’s already fears that the technology could be abused to crank out pathogens that are deadly to humans.

“Although this is promising for life sciences applications, it also raises urgent biosafety and biosecurity questions. The ability to compose viral genomes using generative AI now exists; the governance to safely steer it does not,” warned Tom Inglesby and Moritz Hanke at the Center for Health Security at Johns Hopkins University in Baltimore in a perspective article.

Tom Ellis, a professor of synthetic genome engineering at Imperial College London who was not involved in the research, acknowledged this threat, but told The Guardian that it would be much harder to create more complex genomes and that controlling access to genetic data could stymy the threat.

And if someone were to deliberately engineer deadly viruses, he added, doing it by hand would probably be more practical.

“But honestly, the threat from full AI design and writing of a genome of a virus or bacteria is very overblown when we consider that just taking existing pathogens and making gain-of-function changes to their genomes is so much easier and much more likely to be a real pathogenic threat,” he said.

More on pathogens: America Getting Lesson in Consequences of Trump’s War on Public Health as Rare Disease Outbreaks Ravage the Country

The post AI Model Invents Completely New Viruses Unknown to Nature appeared first on Futurism.

Chaque année, un traitement essentiel sauve des millions de jeunes vies. Mais une vaste étude internationale révèle qu’un phénomène inquiétant progresse à un rythme bien plus rapide que prévu, avec des conséquences potentiellement dramatiques d’ici une dizaine d’années. 

À medida que o debate sobre a utilização das redes sociais pelas crianças continua a ganhar força em vários países, juntando-se à aplicação de novas regras para a IA em escolas, um estudo reforça a ideia de que o acesso precoce a estas plataformas pode ter consequências no percurso escolar.

The post Uso precoce de redes sociais pode afetar o desempenho escolar. Estudo aponta para impacto negativo nas notas appeared first on Tek Notícias.

A team of researchers from the University of Sydney and the University of Glasgow argues that dietary sugars, not animal foods, may have played the decisive role in the expansion of the human brain over the past 4 million years, as hominin brain size grew nearly fivefold, from about 300 grams in early relatives like […]

The post Study Points to Sugar, Not Just Meat, as Fuel for Human Brain Evolution appeared first on Sci.News: Breaking Science News.

As imagens captadas pela sonda Danuri permitiram aos investigadores da Administração Aeroespacial da Coreia do Sul confirmar que a colisão do estágio superior do foguetão da SpaceX causou alterações na superfície lunar. Além da Danuri, a Lunar Reconnaissance Orbiter da NASA também vai fotografar o local do impacto.

The post Foguetão da SpaceX deixou uma nova cratera na Lua? Sonda sul-coreana capta o antes e depois do impacto appeared first on Tek Notícias.

For a president who upholds “Make America Healthy Again” as a load-bearing narrative pillar, Donald Trump has gone to remarkable lengths to make the US a breeding ground for a widening range of viruses, bacterium, and parasites — and no, we’re not talking about his cabinet.

Early into his second tenure as commander-in-chief, Trump and his cronies set to rapidly dismantling offices doing critical work on public health. His appointments were a one-two punch on longstanding institutions safeguarding the US from the ravages of disease: Elon Musk’s DOGE cut the government’s capacity to identify and track public health problems, while Robert F Kennedy Jr’s Department of Health and Human Services limited what the government was willing to do about them.

Case in point: the Centre for Disease Control has lost nearly a quarter of its workforce since January 2025, while those who remain have watched in horror as the CDC’s acting director cut the number of universally recommended childhood vaccines from 17 down to 11.

Examples of those kinds of policy faceplants abound, and the cumulative consequences will surely be felt for decades to come. And nowhere are the consequences of the Trump administration’s war on public health more obvious than in the diseases we’d previously had under control, but which are now surging back after a year and a half of MAHA.

There’s the cyclospora outbreak, of course, spreading from tainted lettuce to over 18,000 victims across 45 states. And those are just the ones we know about — the true extent may be far worse, as reporting delays create a major lag in fresh data. And while cyclospora flare ups come and go, it’s hard to overlook the fact that Trump officials made reporting of the parasite optional for the CDC in 2025, right before the largest known outbreak in modern history.

Separately, a nation-wide salmonella outbreak prompted officials to recall millions of cartons of eggs as at least 100 cases were reported throughout the country. Though salmonella reporting is still mandatory for the CDC, the Food and Drug Administration under Trump has delayed the implementation of a food traceability rule, designed to speed up the tracking of food borne illnesses back to their source.

The reason for the delay? Budget cuts.

Beyond foodborne illnesses, there are even more nasty bugs spreading around the country under Trump’s watch.

The formerly-eradicated measles is back, with over 2,300 American infected in 2026 alone, the highest number since 1992, Vox reported. Flesh-eating screwworms are ravaging livestock, as staffing cuts and mandatory budget reviews at the US Department of Agriculture slowed government containment efforts. Even the nightmarish botulism toxin has made it into the headlines, entering the food supply via contaminated infant formula after the Trump administration disbanded the National Advisory Committee on Microbiological Criteria for Foods — coincidentally right as the group was about to release a report on infant formula.

We could go on. While some will surely shoot back that these diseases are always hanging over us regardless of who sits in the Oval Office, that’s exactly the point.

Public health isn’t a game of checkers, where enemy pieces leave the board once they’re conquered. It’s more like a game of whack-a-mole, demanding quick reflexes, a mallet, and someone to wield them with unending focus. Under Trump, the mallet has been sold off to the highest bidder, while the player has wandered off, convinced the moles are a hoax.

More on MAHA: RFK Jr Wants to Waste $5 Billion in the Worst Way Imaginable

The post America Getting Lesson in Consequences of Trump’s War on Public Health as Rare Disease Outbreaks Ravage the Country appeared first on Futurism.

Lançado originalmente em 2022, o aplicativo Sunbird possui como premissa integrar a experiência do iMessage em dispositivos Android. No entanto, algum tempo depois, o serviço foi encerrado e o app, removido do Google Play devido a preocupações envolvendo privacidade e segurança.

Depois de já ter lançado, em 2024, uma versão beta do seu (então) novo aplicativo, a equipe por trás do serviço anunciou o seu retorno definitivo, afirmando que a nova configuração soluciona essas questões e que “qualquer mensagem é protegida em trânsito e descartada assim que é entregue”, enquanto as mídias enviadas ou recebidas são excluídas “normalmente em até 48 horas, e nunca além de 72 horas”.

Muitos devem estar se perguntando, porém, o motivo que leva alguém a optar pelo Sunbird considerando que a Apple já adotou o suporte ao padrão RCS ¹Rich communications service, ou serviço de comunicação rico. no iOS, facilitando a comunicação com dispositivos Android pelo app Mensagens (Message). Essa é a explicação do serviço:

Em 2024, a Apple passou a oferecer suporte ao RCS nos iPhones, o que melhorou alguns aspectos básicos, como a qualidade das fotos e os indicadores de digitação. No entanto, os resultados são inconsistentes, e a mudança não permitiu que usuários de Android se tornassem membros plenos de grupos do iMessage. Reações, respostas e recursos de grupo do iMessage frequentemente falham quando há um celular Android no chat, e as mensagens enviadas por Androids continuam aparecendo em balões verdes.

Segundo Danny Mizrahi, CEO ²Chief executive officer, ou diretor executivo. da Sunbird Messaging, o aplicativo vai além de simplesmente oferecer balões azuis aos usuários de Android. O executivo afirma que recursos do iMessage (como reações, respostas, vídeos de alta qualidade e gerenciamento de grupos) funcionam perfeitamente no Sunbird para Android.

O Sunbird também introduz novos recursos, como uma nova caixa de entrada prioritária, com abas “Principal” e “Secundária”, de modo que as conversas individuais e os grupos com quem o usuário mais interage permanecem em destaque na aba “Principal”, enquanto mensagens rotineiras — como códigos de verificação e contatos de números desconhecidos — são direcionadas à aba “Secundária”.

Futuramente, o app ganhará ainda a Sunbird Intelligence, um conjunto de recursos de inteligência artificial que operam localmente. Entre as funções está o “Catch Me Up”, que oferecerá aos usuários um resumo consolidado de tudo o que perderam em suas conversas. Ele também terá um assistente integrado que poderá realizar ações como redigir uma resposta ou mover uma conversa entre categorias.

Ainda não se trata, porém, da versão final do Sunbird, e sim da versão “beta aberto”; no entanto, não há mais um processo de inscrição nem lista de espera. Ademais, agora o aplicativo conta com opções de assinatura interna (US$3 mensais ou US$25 anuais).

O Sunbird já está disponível no Google Play.

via AppleInsider

Notas de rodapé

• 1
  Rich communications service, ou serviço de comunicação rico.
• 2
  Chief executive officer, ou diretor executivo.

What has always stirred me with a deep set of conflicting emotions is his reference to the “elephant in the room.” 

Source