Category: Artificial intelligence

February 27, 2025
Amazon’s Alexa+ Unveiled: Enhanced Autonomy for AI Assistant
Key Takeaways
1. Alexa+ is an upgraded version of Amazon’s voice assistant, featuring AI-driven enhancements for more natural interactions and better smart home control.
2. The subscription for Alexa+ costs $19.99 per month but is free for Amazon Prime members.
3. Alexa+ will launch in the U.S. in March 2025, initially available only on specific Echo Show devices starting at $184.
4. Future capabilities include understanding everyday language, providing follow-up questions, and dynamic interactions with smart home devices.
5. Alexa+ can independently manage tasks, such as arranging appointments and processing payments, but there are concerns about data privacy and limited device access.
Amazon has unveiled Alexa+, an upgraded version of its voice assistant that comes packed with AI-driven features. This new version promises to provide more natural interactions, better control of smart home devices, and the ability to perform tasks on its own. However, these enhancements do come with a cost. The subscription for Alexa+ is priced at $19.99 per month, although it’s free for those who have a Prime membership.
Launch Details
Alexa+ is scheduled to launch in the United States in March 2025. However, initial access will be restricted to specific Alexa devices, such as the Echo Show 8, 10, 15, and 21, which are available on Amazon starting at $184. There has been no announcement regarding when Alexa+ will be released in other countries.
Future Capabilities
In the near future, Alexa is expected to comprehend everyday language and incomplete sentences, prompting users with follow-up questions when something isn’t clear. This should make conversations feel much more fluid and less mechanical.
Alexa+ will also enhance interactions with smart home devices by being more dynamic and aware of context. Users can control settings using simple voice commands, like saying, “Alexa, dim the living room lights to 50% when the movie starts.” Additionally, it will learn from user behaviors and suggest handy actions, such as warming up the house before you arrive.
Independent Task Management
With Alexa+, the assistant will be capable of managing tasks online all by itself. It can arrange appointments, locate and book service providers, and even process payments.
Many users on Reddit see Alexa+ as a much-needed update, particularly due to its upgraded voice recognition and better smart home compatibility. Nonetheless, some criticism arises from the limited access to certain Echo Show models. There are also worries about data privacy, potential increases in advertising, and the monthly subscription fee, which has led to some skepticism.
Source:
Link
Tags: Alexa, Echo Show, Smart Home
February 27, 2025
Apple Dominates 2024 with Over Half of AI-Capable PCs Shipped
Key Takeaways
1. Apple leads the market with over 54% of all AI-capable PCs shipped in 2024.
2. The rise of AI in households is linked to the increasing availability of AI-supporting personal devices.
3. Apple’s success is attributed to its new M4 processor and the integration of Apple Intelligence in recent laptops.
4. Windows laptops are struggling to compete, particularly with Microsoft’s Copilot PCs underperforming in sales.
5. The AI PC market is growing, with an 18% increase from 2023 to 2024, indicating a promising future for AI technology in everyday devices.
With the rise of AI, this technology is slowly making its way into everyday households. A significant part of this increase is linked to the growing number of personal devices that support AI, like smartphones and computers. Apple seems to be at the forefront of this trend, and by a large margin.
Apple Dominates the Market
A recent study from Canalys, a market research firm, reveals that Apple accounted for over half of all “AI-capable PCs.” These PCs are defined as those with hardware specifically designed for AI tasks, including a neural processing unit (NPU). In 2024, 17% of shipped PCs fit this category, with Apple responsible for 54% of those shipments.
The tech giant from Cupertino stands out as the clear leader in this sector. Lenovo and HP hold the second and third positions, each contributing 12% of all AI-capable PCs in 2024.
Key Factors in Apple’s Success
The primary reason behind Apple’s impressive performance is likely its recently introduced M4 processor and the incorporation of Apple Intelligence into most of its latest laptops (for instance, the M4 Mac Mini, priced at $549 on Amazon). Apple has technically enabled AI capabilities with on-chip neural processors since the A11 Bionic SoC, which was used in the iPhone 8, iPhone 8 Plus, and iPhone X back in 2017. Every processor from Apple since then has included some form of neural engine, but the company’s recent AI initiatives have brought this technology to the forefront for everyday consumers.
Challenges for Windows Laptops
Windows laptops, however, have not kept pace in this arena. Microsoft’s Copilot PCs have not sold as well as anticipated, mainly due to issues surrounding Copilot (including its Recall feature, which would automatically take snapshots of a PC’s state for easy access).
Overall, the market for AI PCs is consistently expanding. This category experienced an 18% growth from 2023 to 2024. As AI becomes increasingly integrated into our personal devices and the necessary hardware becomes more affordable, we can expect ongoing growth in the AI PC sector.
Source:
Link
Tags: AI-capable PCs, Apple, M4 Mac mini
February 26, 2025
Tecno Launches Camon 40, Megabook S14, and AI Glasses at MWC 2025
Key Takeaways
1. Tecno is launching the Camon 40 Series smartphone, featuring zero-delay snap photography and AI-enhanced imaging for improved mobile photography.
2. The Megabook S14 is introduced as the world’s lightest 14-inch OLED laptop, equipped with AI tools to enhance productivity for on-the-go professionals.
3. The Tecno AI Glasses Series includes smart glasses that integrate advanced AI capabilities with a modern design for seamless connectivity and interactions.
4. Tecno aims to expand its AIoT ecosystem, creating a unified experience across its innovative devices.
5. These new products will be showcased at MWC Barcelona 2025, highlighting Tecno’s commitment to AI-driven mobile and computing technology.
Tecno is making a significant move in artificial intelligence at MWC Barcelona 2025, presenting a range of innovative devices aimed at incorporating AI throughout its product lineup.
Exciting New Devices
At the forefront of this event is the Camon 40 Series, a top-tier smartphone that promises exceptional mobile photography capabilities. Alongside it, the Megabook S14 is being introduced as the lightest 14-inch OLED laptop globally, along with a fascinating new entry into wearable technology, the Tecno AI Glasses Series.
Features and Innovations
According to the company, the upcoming Camon 40 series is designed with zero-delay snap photography and AI-enhanced imaging. This allows for fast and accurate captures through its one-tap capture button. It is powered by Tecno AI and the latest MediaTek Ultimate processors, providing advanced multi-skin tone imaging to create more natural and vibrant photos.
The device also includes AI enhancements for image generation, smart search, and intelligent calling, making it a versatile companion driven by AI. Tecno plans to reveal additional information during the official launch event.
Expanding AIoT Ecosystem
In its effort to grow the AIoT ecosystem, Tecno will also launch the Megabook S14, which is advertised as the lightest 14-inch OLED laptop in the world. It features AI-powered tools that aim to boost productivity and creativity for professionals who are always on the move.
Additionally, the Tecno AI Glasses Series, which comprises the Tecno AI Glasses Pro and standard Tecno AI Glasses, blends advanced AI capabilities with a stylish, modern aesthetic. These smart glasses provide seamless connectivity and intelligent interactions, creating a unified AI-driven product ecosystem that enhances digital experiences.
Tecno will present these groundbreaking developments at MWC 2025, Booth 6B11, Hall 6, Fira Gran Via, where visitors can get a glimpse of the future of AI-enhanced mobile and computing technology.
For the latest updates, please check our News Section.
Stay informed in technology! Connect with our Telegram community and subscribe to our daily newsletter for the top stories!
Source:

Tags: AI Glasses Series, Megabook S14, Tecno Camon 40
February 26, 2025
Apple announces $500 billion US investment plan amid tariff concerns
Key Takeaways
1. Apple plans to invest over $500 billion in the US over the next four years, including a new server manufacturing facility in Houston.
2. The company is doubling its US Advanced Manufacturing Fund to $10 billion and launching a skills academy in Michigan to boost manufacturing skills.
3. Apple will collaborate with thousands of suppliers nationwide, creating direct jobs and infrastructure for its Apple Intelligence systems and data centers.
4. The new Houston server plant will be crucial for Apple Intelligence and is expected to create thousands of local jobs by 2026.
5. Apple aims to add around 20,000 new positions focused on research, development, and advanced technologies over the next four years.
Apple has just revealed an unprecedented investment strategy, planning to pour over $500 billion into the US within the next four years. This initiative encompasses the establishment of a server manufacturing facility in Houston, aimed at powering its Apple Intelligence systems. Additionally, Apple is doubling its US Advanced Manufacturing Fund to $10 billion and will be initiating a skills academy in Michigan to enhance manufacturing skills.
Collaborations and Job Creation
In this ambitious plan, Apple intends to collaborate with thousands of suppliers across all 50 states. This effort will create direct employment opportunities, strengthen the infrastructure for Apple Intelligence, and develop data centers. Furthermore, it includes building corporate hubs and supporting Apple TV+ productions in 20 different states. Currently, Apple supports more than 2.9 million jobs in the US through direct employment, partnerships with American suppliers and manufacturers, and roles for iOS app developers.
New Server Plant in Houston
The new server facility in Houston, which will cover 250,000 square feet and is set to begin operations in 2026, is anticipated to generate thousands of jobs in the area. These servers, previously manufactured overseas, will now be essential for Apple Intelligence and will lay the groundwork for what’s being referred to as Private Cloud Compute, a system that integrates AI processing with high-level security for cloud services.
Investment in Advanced Manufacturing
As part of its enhancement to the US Advanced Manufacturing Fund, Apple is making a significant multibillion-dollar investment in advanced silicon production at TSMC’s Fab 21 in Arizona, where it stands as the largest client. Notably, mass production of Apple chips commenced just last month at this state-of-the-art facility, which already has over 2,000 workers on its payroll.
Future Hiring Plans
Looking forward, Apple aims to add approximately 20,000 new positions over the next four years, primarily in research and development, silicon engineering, software development, as well as AI and machine learning. The company has nearly doubled its investment in advanced R&D in the US over the past five years and is keen to maintain this growth.
All of this is happening amid potential tariff challenges, as the upcoming Trump administration signals a crackdown on Chinese imports. Recently, Apple CEO Tim Cook had a meeting with President-elect Trump, who suggested that Apple might withdraw from Mexico and increase its investments in the US. Meanwhile, the administration is considering a potential 25 percent tariff on chips and has already introduced a 10 percent tariff on goods from China.
Source:
Link

Tags: $500 billion investment, Apple, TSMC
February 25, 2025
Zhaoxin Launches DeepSeek-R1 Models for All CPU Types
Key Takeaways
1. Zhaoxin has launched DeepSeek-R1 Distill models across all its products, including desktops and servers, showcasing adaptability with multiple model versions up to 671B.
2. Their CPUs are compatible with Linux, Windows, and local Chinese operating systems, and support local GPU AI acceleration cards for diverse AI applications.
3. The KaiXian KX-7000 desktop processor runs the DeepSeek-R1-7B model, offering smart features like document management and coding support, integrated with popular applications.
4. The Lianhe Donghai XRS302 workstation, equipped with domestic GPU cards, effectively runs multiple DeepSeek-R1 models, optimized for low-latency and quiet operation.
5. The Kaisheng KH-40000/32 processor can independently run the DeepSeek-R1-671B model and supports up to 2TB of memory, showcasing powerful server-grade AI processing capabilities.
Zhaoxin has recently announced the successful launch of the DeepSeek-R1 Distill models across all its products, ranging from standard desktop computers to powerful servers.
The Chinese CPU manufacturer revealed that their processors can now handle multiple versions of the DeepSeek-R1 Distill model. This includes the 1.5B, 7B, 14B, 32B, 70B, and the massive 671B variant. This deployment covers all of their x86-compatible devices, showcasing the adaptability of their technology.
Broad Compatibility
In addition, Zhaoxin’s CPUs are designed to work seamlessly with Linux, Windows, and various indigenous Chinese operating systems. They are also fully compatible with local GPU AI acceleration cards. This combination creates a strong computing platform suitable for a wide range of AI applications.
Desktop AI Deployment
The KaiXian KX-7000 desktop processor has shown it can successfully run the DeepSeek-R1-7B model when used with domestically produced graphics and acceleration cards. By utilizing DeepSeek’s advanced language understanding, this system offers features such as smart document management, data analysis, and coding support, all integrated smoothly with popular applications like WPS Office and VSCode.
The KX-7000 belongs to Zhaoxin’s latest series of desktop processors. It uses the “Century Avenue” microarchitecture, which features innovative chiplet interconnections, includes 8 CPU cores with 32MB of shared L3 cache, and can achieve speeds of up to 3.7GHz.
Workstation Performance
The Lianhe Donghai XRS302 workstation, powered by the KH-40000/16 server processor, has successfully implemented the DeepSeek-R1-7B, 14B, and 32B models. Equipped with four high-performance domestic GPU inference acceleration cards, this workstation offers a budget-friendly yet powerful solution for AI inference tasks.
Zhaoxin has optimized its CPU-GPU interconnect system to enhance data flow while reducing latency. Additionally, the workstation features a low-power, quiet design, making it ideal for office environments that require AI inference without excessive noise and energy consumption.
Server-Grade AI Processing
On a more advanced level, the Kaisheng KH-40000/32 processor features a dual-socket configuration with a total of 64 cores and can support up to 2TB of memory. When paired with local accelerator cards, this CPU effectively deployed the DeepSeek-R1-70B model.
Moreover, the KH-40000/32 was able to run the entire DeepSeek-R1-671B model independently, without the need for external acceleration cards. This achievement highlights the impressive capabilities of Zhaoxin’s server CPUs to handle models with hundreds of billions of parameters with ease.
Although Zhaoxin’s claims are noteworthy, it is important to approach these performance outcomes skeptically. Like any benchmarks provided by vendors, actual performance can vary from controlled testing situations, and the company might showcase optimized scenarios that portray their products in the most favorable manner.
Source:
Link
Tags: DeepSeek-R1, KH-40000, Zhaoxin
February 25, 2025
Activision Confirms Use of Generative AI in Call of Duty
Key Takeaways
1. Fans were disappointed by the release of the “Necroclaus” loading screen in Call of Duty: Black Ops 6, featuring odd visuals like a zombified Santa with six fingers.
2. The “Gobblegum Mania” loading screen also showcased strange designs, including a hand missing a thumb, raising concerns among players.
3. Players are urging Activision Blizzard to clarify its use of generative artificial intelligence after previously dismissing real artists.
4. Steam introduced a new AI disclosure policy in December 2024, but its connection to Call of Duty remains unclear.
5. Developers may expand the use of AI in gaming, potentially creating entire game levels, following the launch of the Muse AI model.
When Activision Blizzard released the Season 01 Reloaded update for Call of Duty: Black Ops 6 (available on Amazon) in early December 2024, some fans were a bit disappointed by a new loading screen showcasing “Necroclaus.” This zombified Santa Claus has what appears to be six fingers on a rotting hand, along with presents that have oddly placed ribbons. Additionally, the “Gobblegum Mania” loading screen features another six-fingered hand in a glove grasping several gumballs, but it is missing a thumb. Fans also noticed strange visuals on a calling card from a set of paid downloadable content. Because of these examples, players have been urging the publisher to reveal its use of generative artificial intelligence, particularly after dismissing numerous real artists previously.
Uncertain Connections to Steam
It’s still unknown if Steam’s recent AI disclosure policy is connected to Call of Duty or if Valve was already planning to add it to its product pages. However, we do know it was introduced in December 2024. A new note has popped up on the Steam page for Call of Duty BO6, stating:
The developers explain how their game incorporates AI Generated Content: Our team uses generative AI tools to help develop some in-game assets.
The Future of AI in Gaming
While this statement confirms the hints and speculation about Call of Duty’s utilization of AI assets, it’s uncertain how the player community will respond. With the launch of the Muse AI model, developers might soon expand their use of artificial intelligence beyond just loading screens and in-game items, potentially creating entire game levels through AI technology.
Source:
Link

Tags: Activision Blizzard, Call of Duty Black Ops 6, Generative AI
February 25, 2025
Seemour Enhances AI Camera Notifications for Smarter Alerts
Key Takeaways
1. Seemour is an innovative platform that uses visual intelligence to enhance home security camera systems with smarter alerts.
2. The software can reduce unnecessary alerts by up to 70% by filtering through video feeds and highlighting important moments.
3. Seemour intelligently distinguishes between family members and strangers, providing tailored notifications for peace of mind.
4. Future updates will include voice command capabilities and custom visual alerts, addressing user needs without extensive setup.
5. The platform’s success will depend on its compatibility with existing security cameras and potential subscription fees, benefiting consumers in a competitive market.
AI is making home security smarter and easier. Prompt AI has launched Seemour, a new platform that uses visual intelligence to interact with camera feeds. This innovative software enhances traditional AI camera systems by providing smarter and more focused alerts. Homeowners can download the Seemour software directly from the App Store and start using it right away.
Growing AI Options in Home Security
Many companies, like Ring and Wyze, are adding AI detection features to their cameras. Seemour integrates seamlessly with existing home security systems, collecting video feeds and notifications through a user-friendly interface. This software can help homeowners save time by filtering through long hours of video and highlighting important moments. According to Prompt AI, Seemour can cut down unnecessary alerts by as much as 70 percent.
Smart Alerts for Peace of Mind
The intelligent notification system can distinguish between family members and strangers at the door. Seemour is designed to differentiate between potential risks and friendly visits, helping to eliminate unnecessary alarms. The app is smart, learning and remembering the names of users’ loved ones and pets to provide more tailored notifications.
Users can explore Seemour’s capabilities now, but there are plans for future enhancements. Prompt AI intends to enable voice commands for the app to answer questions. For example, parents can quickly check when their child leaves home or make simple requests to find lost items. Moreover, they are developing custom visual alerts that can recognize specific happenings without needing extensive setup. Privacy remains a vital issue with AI camera software, and Seemour will be updated to allow users to exclude certain individuals from video recordings.
Compatibility and Cost Considerations
The success of Seemour will largely depend on how well it works with popular security cameras. Additionally, any potential subscription fees are a factor to consider. However, the increasing competition in AI software services is likely to benefit consumers overall.
Tags: Prompt AI, Ring, Seemour
February 25, 2025
Anthropic Unveils Claude 3.7 Sonnet AI for Expert Pokémon Red Play
Key Takeaways
1. Enhanced Problem-Solving: Claude 3.7 Sonnet has improved problem-solving abilities, performing better on complex tasks and assessments, including PhD-level benchmarks.
2. Improved Gaming and Coding Skills: The AI excels in gaming, notably in Pokémon Red, and offers enhanced troubleshooting for coding tasks, aiding developers in managing complex code bases.
3. Safety Concerns: Claude 3.7 Sonnet showed a higher frequency of guideline violations during safety tests compared to its predecessor, although such occurrences were minimal.
4. Public Availability: Basic features of Claude 3.7 Sonnet are accessible for free, while advanced features require a paid subscription.
5. Increased Token Window: The model utilizes a larger 128K token window, allowing for deeper thinking and more efficient handling of complex prompts.
Anthropic has introduced its newest AI chatbot, Claude 3.7 Sonnet, which boasts enhanced coding capabilities and deep thinking skills. This allows it to tackle complex prompts and programming tasks more efficiently, utilizing a larger 128K token window.
Enhanced Problem-Solving Abilities
Like recent large language models from OpenAI and xAI, the extension of thinking time enables Claude’s latest version to spend more time resolving tough problems before providing solutions. This improvement has significantly boosted Claude’s performance, moving it from being a slow performer to ranking among the top AI systems on difficult assessments, including the PhD-level GPQA benchmark. However, it’s important to note that the 3.7 version isn’t the top AI globally; it balances being a leading model on certain benchmarks against other strong contenders.
Advancements in Gaming and Coding
Claude demonstrates substantial progress in gaming, particularly in titles like Pokémon Red, surpassing the capabilities of previous models. Programmers can also take advantage of its enhanced troubleshooting skills for real-world software problems and coding tasks. A limited preview of Claude Code offers access to an assistant that works alongside developers to edit, test, and maintain complex code bases on GitHub, which can save them considerable time.
Safety Concerns and Accessibility
The advancements in AI intelligence might pose some risks. During internal safety tests, Claude 3.7 Sonnet generated responses that went against Anthropic’s guidelines three times more frequently than Claude 3.5, although this occurred only 0.6% of the time. Additionally, the AI demonstrated the ability to compromise a test network and extract data using methods like code rewriting. However, the public version of Claude includes safeguards to mitigate such risks.
Users can access the basic features of Claude 3.7 Sonnet for free, while more advanced features, including extended thinking capabilities, require a paid subscription.
Anthropic Claude, Anthropic press release 1, Anthropic press release 2, Anthropic on YouTube, Anthropic Claude system card
Source:
Link
Tags: AI chatbot, Anthropic, Claude 3.7 Sonnet
February 25, 2025
AI Blood Test Predicts Diseases 15 Years Early
Key Takeaways
1. A new AI-powered blood test can predict disease risks up to 15 years in advance.
2. The “Human Health and Disease Proteome Map” was created by analyzing blood samples from over 50,000 individuals over 14 years.
3. Researchers focused on nearly 3,000 plasma proteins to forecast hundreds of diseases, including Alzheimer’s and heart disease.
4. The test aims to be affordable, costing about $5-10, making it accessible for regular health assessments, especially for older individuals.
5. AI is also enhancing drug discovery, potentially reducing research time from 15-20 years to just 3-5 years.
A revolutionary blood test powered by AI might be able to predict the chances of certain diseases as much as 15 years before any signs show up.
New Discoveries in Blood Testing
Scientists from China have introduced the first-ever “Human Health and Disease Proteome Map” after analyzing blood samples from over 50,000 individuals over a span of 14 years. This new technique utilizes artificial intelligence to identify early indicators in plasma, which could lead to diagnosing conditions like Alzheimer’s and heart disease much earlier than before.
Research Developments
Researchers from Fudan University started by investigating nearly 1,500 plasma proteins and focused on 11 that could indicate a future risk of dementia. By widening their study to examine almost 3,000 proteins in human blood and using AI methods, their system can now forecast the risk of hundreds of diseases well over a decade in advance.
“The main benefit of this technology is its capability for early detection at a lower cost,” mentioned Yu Jintai, who is the deputy director of the Fudan University Research Institute and also serves as the Deputy Director of Neurology at Huashan Hospital. The team is currently developing a quick test kit that aims to make this type of screening as easy and affordable as checking blood pressure or blood sugar levels.
Future Implications and Cost
When this test becomes available, it may only cost a few dozen yuan (approximately $5-10). This pricing makes it particularly beneficial for middle-aged and older individuals who require regular health assessments.
This advancement is part of a broader trend in AI-enhanced healthcare in China. For instance, Shen Ning and his team at Liangzhu Laboratory in Hangzhou have used AI algorithms to create a new treatment for children suffering from progeria, and the initial findings seem promising.
According to Shen, AI can significantly speed up the drug discovery process, reducing research time from the usual 15-20 years to just 3-5 years—while also boosting the chances of successful outcomes.
Source:
Link
Tags: AI Blood Test, Fudan University, Human Health and Disease Proteome Map
February 25, 2025
Apple Intelligence May Compete with Galaxy AI Through Gemini Integration
Key Takeaways
1. Apple’s AI features currently lag behind those of Samsung and other Android models.
2. Users may soon choose between Google’s Gemini AI and OpenAI’s ChatGPT in upcoming updates.
3. The integration of Google’s Gemini AI is expected in the iOS 18.4 update.
4. The update may introduce new features like prioritized notifications and an Image Playground mode.
5. Despite potential improvements, Apple’s cautious approach raises questions about significant advancements in AI capabilities.
While Apple has brought AI into its devices, it still falls behind the offerings found in Samsung and other Android top models. We’ve all witnessed the comparison between the AI-driven object removal feature on an iPhone and that of a Galaxy flagship. Several reports also point out major problems with the notification summarization feature. A recent leak, however, hints that improvements to Apple’s AI features may be on the horizon.
New AI Options for Users
Reports indicate that users will soon be able to choose between Google’s Gemini AI and OpenAI’s ChatGPT for Apple Intelligence features. Apple is said to be working on integrating Google’s Gemini AI into its upcoming iOS 18.4 update. This information was uncovered by developer Aaron Perris (@aaronp613) on X, and it points to Apple enhancing its AI capabilities with the addition of Gemini. For those who don’t know, Gemini is the backbone of many AI features seen in Android flagships like the Galaxy S25 series.
Increased Flexibility and Performance
The code snippet suggests that users will soon have the ability to pick either “Google” or “OpenAI” as their preferred AI model, which could lead to more flexibility and might enhance the overall performance of Apple Intelligence. The anticipated iOS 18.4 update is likely to bring various AI-driven upgrades, such as prioritized notifications and a new Image Playground mode. However, just because Apple is integrating Google Gemini doesn’t mean a big leap in AI features is guaranteed. Apple has been cautious from the start, and it seems their approach hasn’t shifted much yet. We remain hopeful that Apple Intelligence will become smarter, but we’ll need to wait for the official rollout to make further judgments.
Source:
Link
Tags: Apple, Google Gemini, iOS 18.4