❌

Reading view

O aviso de “armazenamento quase cheio” é o melhor vendedor da Apple — e a conta chega todo mês

“O armazenamento do iCloud está quase cheio”. Se você usa iPhone, aposto que leu essa frase nesta semana. Talvez hoje. Ela aparece nos Ajustes, aparece por notificação, aparece quando o backup falha de madrugada… e não desiste nunca. Não é um bug. É, possivelmente, o vendedor mais eficiente que a Apple já contratou.

Nas últimas semanas, escrevi aqui duas matérias medindo produtos da Apple numa moeda que todo mundo entende na pele: horas de trabalho. Vimos quanto custa um Mac depois dos aumentos e quanto custou cada geração de iPhone desde 2007. Mas hardware você paga uma vez e o assunto morre; assinatura, não. Assinatura é um pedágio silencioso: poucos reais por mês, para sempre. E ninguém de fato se dá ao trabalho de fazer essa conta — então eu resolvi fazer.

Os mesmos 5GB desde 2011

Antes dos números, um pouco de contexto. Quando o iCloud nasceu, em outubro de 2011, a Apple deu 5GB gratuitos para cada conta. Era a época do iPhone 4s, de câmera de 8 megapixels e fotos de uns 2,5MB cada. Dava para viver um bom tempo dentro daquele limite.

Quase 15 anos depois, os 5GB continuam exatamente os mesmos. Só que o mundo ao redor deles mudou de tamanho: o iPhone 17 tira fotos de 48MP, grava 4K a 60 quadros — cerca de 400MB por minuto — e sai de fábrica com 256GB. Ou seja, a Apple vende um aparelho com 256GB e dá uma nuvem que corresponde a 2% disso. Um backup básico, sozinho, já não cabe. A caixa ficou pequena de propósito ou só ficou pequena? Guarde essa pergunta.

Quem não aguenta o aviso assina o iCloud+. No Brasil, depois do reajuste de até 33% em 2025, as mensalidades saem por R$5,90 (50GB), R$19,90 (200GB), R$66,90 (2TB) e vão até R$399,90 (12TB).

A conta em horas, plano a plano

A régua é a mesma das matérias anteriores: salário mínimo por hora no Brasil (R$8,50), nos Estados Unidos (piso federal de US$7,25), no Canadá (C$17,60, em Ontário) e em Portugal (5,31€). Cada célula mostra quanto tempo de trabalho um mês de assinatura consome.

PlanoBrasilEUACanadáPortugal
iCloud+ de 50GB42min8min4min11min
iCloud+ de 200GB2h20min25min14min34min
iCloud+ de 2TB7h52min1h23min44min1h53min
iCloud+ de 12TB47h8h16min4h33min11h18min
Preços oficiais da Apple em julho de 2026, sempre no plano mensal.

À primeira vista, parece barato — e é justamente assim que uma assinatura funciona. O plano de 50GB custa 42 minutos de trabalho por mês. Quem reclamaria? Mas suba na tabela: o plano de 2TB, o mais comum para quem tem família ou anos de fotos acumuladas, consome quase um dia inteiro de trabalho por mês de um brasileiro de salário mínimo. O mesmo plano custa 44 minutos para um canadense. E o de 12TB beira o surreal: 47 horas mensais — mais de uma semana de expediente, todo mês, só para guardar arquivos.

Vale a comparação com o hardware: nas matérias anteriores, o brasileiro trabalhava de 8 a 12 vezes mais que o americano pelo mesmo iPhone ou iPad. Nos serviços, a distância é menor — cerca de 6 vezes no plano de 2TB —, mas segue aí, firme e forte. Contra o Canadá, a proporção volta às mesmas 10 vezes de sempre. E Portugal repete o papel de meio-termo que já apareceu na conta do hardware: 1h53min pelos mesmos 2TB, 4x menos esforço que o brasileiro — confortável perto de nós, longe do sossego canadense.

O pedágio que nunca acaba

Agora, o detalhe que muda tudo: essa conta não fecha nunca. O iPad de 706 horas dói uma vez; o iCloud cobra todo mês, até o fim dos tempos — ou das suas fotos, o que vier primeiro.

Um ano de iCloud+ de 2TB custa R$802,80 — 94 horas de trabalho, meio mês de expediente por ano. Em 10 anos, são mais de 900 horas: território de iPhone 17, que calculei em 1.006 horas na matéria anterior. Ou seja, guardar as fotos por uma década custa quase o mesmo esforço que comprar o telefone que as tirou.

E tem o degrau seguinte, claro. O Apple One Individual (R$42,90) leva 5 horas mensais de trabalho; o Premium (R$99,90), quase 12 horas — cerca de 140 horas por ano, ¾ de um mês de trabalho anual em assinatura. No fim das contas, o brasileiro de salário mínimo que assina o pacote completo da Apple dedica praticamente um mês do ano só para manter os serviços da Maçã rodando.

Por que a caixa não cresce

Aqui entra o olhar de produto, área em que trabalho há mais de 20 anos. Aquele aviso de armazenamento não é um acidente de percurso: é um funil de conversão dos bons! Ele aparece no momento de maior dor (backup falhou, foto não salvou), oferece a solução a um toque de distância e custa “só” R$5,90. Nenhum vendedor de loja chega perto dessa taxa de conversão.

E o incentivo para manter os 5GB congelados é gigante. Os serviços renderam à Apple US$109,2 bilhões no ano fiscal de 2025, com margem bruta de 75% — o dobro da margem do hardware. Cada pessoa que cede ao aviso vira receita recorrente, previsível e baratíssima de servir. Se os 5GB gratuitos virassem 50GB amanhã, quantos milhões de assinaturas de entrada evaporariam? A Apple sabe a resposta melhor do que ninguém. Tanto que Google (15GB) e Microsoft (5GB, mas com Office no pacote) jogam jogos parecidos — a Apple só joga com a caixa mais apertada do mercado.

Sejamos justos num ponto: nuvem custa dinheiro de verdade (data center, energia, redundância) e o iCloud+ entrega junto recursos como o Retransmissão Privada (Private Relay) e o Ocultar Meu E-mail (Hide My Email). Não é um produto ruim; eu mesmo assino e não pretendo cancelar. O problema não é o plano pago existir, é a versão gratuita ter sido desenhada para não ser suficiente — em 2011, ela era.

No fim, é uma escolha de modelo de negócio — e das mais bem executadas. Só acho saudável que a gente a chame pelo nome. Quando o aviso aparecer de novo aí no seu iPhone — e ele vai aparecer —, lembre que aqueles 42 minutos de trabalho por mês não foram um acaso: foram desenhados, testados e otimizados para esse exato momento. A Apple não vende mais só o telefone. Vende o lugar onde a sua vida digital mora — e cobra o aluguel em dia.

  •  

Intel, Google Cloud deepen AI partnership

Intel and Google Cloud broadened their multi-year strategic collaboration with plans to deploy Gemini Enterprise and Google Cloud infrastructure across the chipmaker’s global operations as part of a wider AI-driven transformation.

Under the expanded partnership, Intel will roll out Gemini-powered generative AI tools across engineering, supply chain and corporate functions, aiming to scale workforce capabilities through agentic workflows.

The hyperscaler’s infrastructure will also support Intel’s semiconductor development environment, with custom agentic workflows intended to speed up the chip design lifecycle and improve cross-functional execution.

Intel CIO Cindy Stoddard stated the collaboration gives employees a central hub to build and deploy agents through Gemini Enterprise, alongside elastic cloud infrastructure to scale silicon development, as part of the company’s push to move faster and more efficiently.

Karthik Narain, chief product and business officer at Google Cloud, described the partnership as an effort to pair Intel’s engineering expertise with agentic AI tools to create what he called an autonomous foundation for how the chipmaker designs, operates and scales its business.

The companies stated the deployment moves Intel beyond isolated AI pilot programmes. Through the Gemini Enterprise Agent Platform, Intel’s business units will be able to build and run tailored line-of-business agents aimed at automating core processes and speeding up data-driven decision-making.

Gemini’s reasoning capabilities will also power dedicated coding assistance and engineering automation tools designed to streamline software development pipelines and automate multi-step workflows.

Intel is separately piloting Google Cloud-powered tools for marketing and communications, including agents that can identify relevant subject matter experts, draft executive-ready messaging, and generate supporting content across communications channels.

On the infrastructure side, Intel is extending its on-premises compute capacity into Google Cloud’s C4 and N4 instances to support silicon development simulations and core engineering workloads, letting engineering teams run high-performance computing simulations concurrently to shorten chip development cycles.

The tie-up builds on a partnership which has deepened steadily this year. In April, Google committed to using multiple generations of Intel’s Xeon processors, including the 18A-manufactured Xeon 6, to support AI training and inference in its data centres, alongside a plan to co-develop custom infrastructure processing units.

The post Intel, Google Cloud deepen AI partnership appeared first on Mobile World Live.

  •  

President Xi makes pitch for China to lead global AI race

China President Xi Jinping (pictured) set out his vision for the country to assume a leadership role in AI, calling for increased international cooperation and promotion of open-source technology while taking aim at the US over attempts to govern the sector.

Xi made the comments in a speech at the World AI Conference in Shanghai, where it was announced China, Brazil, Russia and Indonesia are among 30 countries to have joined the World Artificial Intelligence Cooperation Organisation.

Chinese state media describe the new group as an intergovernmental organisation headquartered in Shanghai to promote global AI governance.

Of the new group, Xi said AI should not be a “solo performance by one country, but rather a symphony of global cooperation”.

He said the initiative is a response to the call of the Global South to unite the international community in promoting development and governance of AI, “and it will become a major milestone in the history of artificial intelligence development”.

To further support this, Xi said China will provide 5,000 training opportunities in AI-related fields for developing countries, in addition to establishing AI application cooperation centres.  

In his speech, Xi noted that as AI continues to develop at “staggering speed”, there was a greater need for positive direction.

“We must more precisely grasp the scale of oversight and governance, and more promptly refine measures to forestall loss of control.”

Overstretch
While not naming the US directly, Xi said countries should oppose “overstretching the national security concept in the field of AI, or placing one country’s security over that of others”.

The US has notably placed several restrictions on China, impacting companies in the country from accessing high-end technology.

XI continued to state he was committed to promoting cooperation, bridging the digital divide, promoting sustainable development and avoiding “creating new historical injustices in the field of artificial intelligence”.

Finally, he called on the countries to “seize this rare historical opportunity, encourage open source and open collaboration”, in turn upgrading traditional industries and ensure AI can “empower all sectors”.

The post President Xi makes pitch for China to lead global AI race appeared first on Mobile World Live.

  •  

NTT, TBS target child AI education

NTT and Japanese media company TBS identified AI as a potential danger to children’s decision-making abilities and outlined an intention to form a company to tackle the issue using digital games and smartphone apps.

The pair plan to commence operations of their e6 joint venture on 8 August to help children form clear ideas in a world where AI offers myriad possible answers to questions. The business is to be used to commercialise a namesake project NTT and TBS launched in February to provide an experience-based educational approach using various technologies.

NTT stated the companies created Emotional Knight, a story-based “original IP” emotive education game employing AI and extended reality to teach decision-making.

NTT President and CEO Akira Shimada said the companies believe huge volumes of information are not necessarily the best thing for children.

Instead, they need “the ability to think for themselves and make their own decisions while interacting with a wealth of information and AI”.

The companies intend to expand the line-up of educational games, develop relevant IP and add more “entertainment experiences” which foster children’s “ability to make independent decisions”.

NTT is contributing group resources including its innovative optical and wireless network, tsuzumi 2 LLM, R&D and social implementation capabilities. TBS brings content creation, IP development and distribution elements.

The Japanese conglomerate stated feedback from an event held in March showed a high level of satisfaction.

The post NTT, TBS target child AI education appeared first on Mobile World Live.

  •  

Micron strikes top-tier vehicle supply chain deals

Micron Technology secured strategic supply agreements with major automotive partners including Qualcomm and Harman to guarantee long-term chip access for software-defined vehicles.

The deals come during an aggressive industry-wide race to expand manufacturing capacity to meet a massive surge in memory chip demand driven by rapid AI adoption.

As the only US-based manufacturer of high-bandwidth memory chips used in Nvidia’s AI processors, Micron occupies a unique market position alongside South Korean rivals Samsung and SK Hynix, allowing it to command premium prices while solidifying long-term commitments during ongoing hardware shortages.

Micron also struck deals with Visteon, Joynext, Denso, Astemo, and Hyundai Mobis. It stated the agreements would improve visibility and help optimise production planning. It expects greater stability in pricing and supply chains for increasingly sophisticated AI-enabled vehicle platforms.

Demand for high-performance DRAM and NAND flash memory has grown as the automotive sector transitions toward highly connected, autonomous and intelligent designs.

Modern vehicle architectures require robust computing hardware to support advanced driver assistance systems, infotainment services and high-speed connectivity.

Sanjay Mehrotra, chair, president and CEO of Micron, emphasised the next phase of automotive innovation hinges on the strength of the ecosystem behind it. He noted “memory and storage serve are critical enablers” for the sophisticated digital experiences consumers expect.

Qualcomm CEO Cristiano Amon highlighted the collaborative effort required to deliver advanced digital cockpit and driver assistance technologies. He stated working with Micron provides a strong hardware foundation as vehicles become more intelligent and connected.

The automotive sector is a key area for increasing Qualcomm’s revenue beyond mobile chips. In its April Q2 earnings, it reported Automotive sales rose 38 per cent to $1.3 billion.

The post Micron strikes top-tier vehicle supply chain deals appeared first on Mobile World Live.

  •  

Aina scores $5.5M for AI interface work

Consumer hardware company Aina raised millions of dollars in seed funding to finance development of an interface it argues is required to fully unlock the potential of AI on smartphones and PCs.

The company believes current device hardware lags the progress of software and AI, resulting in a disconnect between expectation and reality in terms of the interface with apps and services.

It raised $5.5 million in the recent funding round to advance development of experimental interfaces which adapt to the app being used.

Aina stated it already shipped hundreds of experimental keypads called Dune to early adopters, a product for Mac computers it describes as “context-aware” which primarily focuses on work tools.

It is using the feedback from Dune to develop a general-purpose interface for smartphones and computers.

Aina founder and CEO Apoorv Shankar (pictured) said current smartphones and PCs “are still primarily designed for browsing” and use manual inputs requiring users to “figure out what you need”.

He argued this is laborious when users are “doing hundreds of tasks” per day.

The company believes there is room within the broader AI assistant space for a contextualisation system which makes it easier for people to interact with the technology. Shankar explained the general-purpose interface on the drawing board is being designed to “capture human approval”.

Shankar was previously VP of hardware at smart ring company Ultrahuman.

The seed funding round was led by Redstart Labs (Infoedge) and 360 ONE Asset, with two other investment companies and a host of angel investors.

Along with launching its interface, Aina intends to use the backing to grow its team in the US and India.

The post Aina scores $5.5M for AI interface work appeared first on Mobile World Live.

  •  
❌