Apple has launched its first 2024 public betas for iOS 18, iPadOS 18, watchOS 11, and macOS Sequoia. These early builds offer a preview of the upcoming software updates set to officially released in the fall. Users interested in trying out these features can enroll their Apple ID in the Apple Beta Software Program to gain access.

The highlight of these updates is Apple’s new generative AI, called Apple Intelligence. This feature is integrated across all platforms, blending on-device and cloud processing, with optional ChatGPT integration. Apple Intelligence enhances Siri, provides Safari with new highlights and summaries, introduces advanced writing tools, and more.

iOS 18 offers significant customization for the Home Screen, allowing users to place apps and widgets freely and change app icon colors. The Control Center has been redesigned for easier access and greater customization. The Photos app receives its biggest update ever, and the Messages app introduces new formatting and effects.

iPadOS 18 incorporates many changes from iOS 18, including the addition of a native Calculator app with a new Math Notes feature for real-time problem-solving. The Smart Script tool in the Notes app can refine handwriting and convert recordings into searchable text. These updates enhance the functionality of the Apple Pencil and the overall note-taking experience.

macOS Sequoia introduces iPhone Mirroring for controlling an iPhone from a Mac. It also features a standalone Passwords app for managing credentials, window tiling for organizing open applications, and new video call options for customizing backgrounds. watchOS 11 brings the ability to pause Activity Rings, and adjust fitness goals, and includes a Vitals app for sleep tracking and a tool to monitor workout intensity over time.

To install the public betas, users need to register their Apple ID on the Apple Beta Software Program website. Once registered, they can navigate to Settings > General > Software Update on their devices, select the Beta Updates menu, choose the Public Beta option, and install the update.

On Monday, Elon Musk announced that Tesla’s robotaxi event, initially scheduled for August, has been postponed to October. Musk shared this update on X (formerly known as Twitter), a platform he owns, in response to a user’s inquiry.

He explained that the delay was due to his request for a significant design change to the vehicle’s front, and the additional time would also allow Tesla to showcase some other new features.

This announcement aligns with a recent Bloomberg report from July 11, which indicated that the event’s delay would provide Tesla with more time to develop additional prototypes.

The report mentioned that Tesla had communicated this delay internally, instructing the design team to rework specific elements of the car. This directive appears to be in line with Musk’s request to alter the front design of the robotaxi.

Originally, Musk had stated in early April that the Tesla Robotaxi would be revealed on August 8. This announcement coincided with a Reuters report revealing that Tesla was discontinuing plans for a $25,000 budget model, often referred to as the Model 2, to focus more on robotaxi development.

Musk responded to the Reuters report by accusing the publication of lying, though he did not specify which details he disputed.

In the weeks following these announcements, Tesla undertook significant cost-cutting measures, including reducing its workforce by over 10 percent from its more than 140,000 employees. Shortly thereafter, two senior executives were also let go as Musk emphasized the need for strict headcount and cost reduction.

Despite these efforts, Tesla reported a nearly seven percent drop in vehicle deliveries in the first half of 2024 and a nearly 15 percent decrease in production in the second quarter compared to the same period in 2023.

In 2024, brands have grown more comfortable in the gaming space, leading to longer and deeper sponsorships with creators. These partnerships are inherently integrated with creators’ content and communities, moving beyond one-off activations to more fully integrated programs.

This shift reflects brands’ recognition of the spending power of “cultural gamers” who engage with gaming through social media and digital content rather than playing games themselves.

Brands are increasingly reaching gamers outside of video games by extending the duration of their partnerships with creators. Instead of short-term campaigns with specific deliverables, brands are opting for open-ended campaigns lasting several months.

This allows creators the freedom to introduce brands to their audiences in a more natural and effective manner. Such long-term partnerships benefit both the creators and the brands by fostering a more authentic integration into the creators’ communities.

A notable example of this integrated sponsorship approach is Cody “Clix” Conrod, a YouTuber and professional Fortnite player. His partnership with apparel brand Hugo included not just social media promotions but also a sponsored trip to Berlin and a VIP fan event.

Conrod emphasizes the importance of brands committing long-term to add value to his community, reflecting a broader trend among influencers who prefer more flexible and meaningful collaborations with brands.

This sentiment is echoed by many creators, including those at Conrod’s esports organization, XSET. XSET CEO Greg Selkoe highlights the importance of allowing creators to be themselves and activate sponsorships authentically.

Brands are beginning to understand that rigid guidelines and prescriptive instructions can lead to inauthentic marketing efforts that fall flat with audiences. As a result, XSET prioritizes collaborations where creators genuinely like the brand or product.

As brands continue to deepen their presence within gaming platforms like Fortnite and Roblox, they rely on creators to promote these experiences and drive traffic. Long-term collaborations allow for ongoing feedback and improvements, making branded experiences more engaging and effective.

This approach is not a sudden strategic pivot but a natural evolution as brands work more closely with creators to reach gamers authentically. The unique relationship between creators and their audiences is key to successful marketing in the gaming space, as highlighted by Damon Lau of United Talent Agency.

The German Navy urgently needs to modernize its fleet of submarine frigates, particularly the outdated 8-inch floppy disk systems that have been in use since the 1990s. These disks are crucial for the operation of the Brandenburg class F123 frigates, managing essential functions such as airflow systems and power generation.

The reliance on such antiquated technology highlights the critical need for an upgrade to ensure the fleet’s efficiency and security.

Replacing the floppy disk system poses significant challenges. The disks play an integral role in controlling various ship operations, making the transition to modern systems complex and potentially risky.

This situation raises questions about the overall state of the technology used by the German Navy, humorously suggesting that other outdated gadgets like Palm Pilots or Tamagotchis are still in use for operational purposes.

Saab has been responsible for the maintenance of Germany’s F123 frigate fleet since 2021. Alongside the replacement of the floppy disk systems, the frigates are receiving upgrades to their weapons and weapon control systems, which are crucial for their role in submarine hunting. It is hoped that these upgrades will also include modern and user-friendly enhancements to the ship’s interface and control panels.

The use of floppy disks in government and military systems is not unique to Germany. Many institutions have relied on this outdated technology long after it became obsolete in the consumer market.

For instance, Japan’s Digital Agency recently eliminated floppy disks from its systems. The persistence of such old technologies raises concerns about the potential vulnerabilities in critical infrastructure, emphasizing the need for ongoing modernization efforts.

In 2024, digital publishers are experiencing a summer marked by optimism despite ongoing market uncertainties. Gallery Media Group and Apartment Therapy Media report robust performances, with 15% year-over-year revenue growth in total and direct ad revenue respectively. Despite this, caution prevails among some publishers, who acknowledge the volatile nature of the ad market, capable of abrupt shifts.

Apartment Therapy Media’s president, Riva Syrop, anticipates a strong 17% annual revenue increase, buoyed by a 20% surge in direct-sold advertising.

Syrop notes a promising outlook for Q3, citing high renewal rates and increased RFP volumes. Similarly, Gallery Media Group’s CRO, Chris Anthony, attributes their 15% revenue growth to enhanced event and branded content investments, bolstering average deal sizes by 10% year-over-year.

Vox Media’s CRO, Geoff Schiller, highlights market unpredictability, noting fluctuating revenue performance across months. Despite challenges, optimism persists, with indications of recovery and growth in sectors like tech, telco, and pharma, although media and entertainment face unique hurdles such as content scarcity from writer strikes.

Market dynamics remain influenced by broader economic factors. An anonymous publisher executive rates their optimism at six out of ten, citing interest rates and inflation as pivotal influences.

Syrop underscores the resurgence of previously impacted sectors like big-box retail and CPG, while Anthony notes unexpected growth in auto advertising for a lifestyle publisher.

Post-Cannes insights reveal publishers’ strategic shifts toward AI and sports content. While AI remains in its infancy as a transformative tool, sports content, especially featuring athletes like Megan Rapinoe and Sue Bird, garners significant advertiser interest.

Publishers are leveraging these insights to shape their pitches for Q3 and Q4, with some eyeing longer-term planning into 2025.

While the ad market shows resilience and potential, publishers remain cautious amid economic uncertainties. Strategic investments in events, branded content, and targeted client engagement strategies are proving pivotal in navigating the unpredictable terrain of 2024’s advertising landscape.

Yasmeen Ahmad, from Google Cloud, addressed the complex relationship between size and performance in large language models (LLMs) during a recent discussion. She emphasized that while larger models generally exhibit improved capabilities, this trend is not limitless.

Small, context-specific models can often outperform their larger counterparts, leveraging domain-specific knowledge that enhances their effectiveness. Ahmad highlighted the pivotal role of data in this equation, asserting that industry-specific information empowers models significantly.

Moreover, Ahmad stressed the transformative potential of LLMs for enterprises, enabling unprecedented creativity, efficiency, and inclusivity. By harnessing previously inaccessible data, organizations can achieve comprehensive insights and foster innovative engagement across all facets of their operations.

Ahmad noted that these advancements blur traditional boundaries between technology and creativity, ushering in a new era where AI’s capabilities redefine conventional norms.

The conversation shifted towards practical applications within enterprises, emphasizing two key techniques: fine-tuning and retrieval augmented generation (RAG). Fine-tuning tailors LLMs to grasp the nuances of business language, while RAG facilitates real-time data connectivity, crucial for dynamic fields like financial and risk analytics.

Ahmad underscored the importance of multimodal capabilities in LLMs, which enhance their ability to process diverse data formats such as text, images, and videos, thereby enriching customer experiences and operational insights.

Addressing the evolving role of AI in business interactions, Ahmad highlighted the necessity for LLMs to engage in coherent, contextual dialogues akin to human conversations.

This conversational capability, she argued, distinguishes advanced AI from mere chatbots, transforming them into invaluable ‘personal data sidekicks’ capable of nuanced interactions and decision-making.

Ahmad expressed optimism about AI’s evolution towards greater autonomy and strategic acumen, with implications that extend into every facet of modern enterprise operations.

Ahmad painted a picture of AI’s accelerating integration into business frameworks, driven by its expanding capacity for complex analysis, real-time responsiveness, and conversational engagement. This evolution, she suggested, marks only the beginning of what lies ahead in AI’s transformative journey across industries.

News U.K. recently concluded its application process for the head of the consent role, marking a significant step in response to evolving data privacy regulations and industry standards. The role itself is pivotal, tasked not merely with ticking compliance boxes but with shaping a proactive and forward-thinking consent strategy across the publisher’s brands.

The responsibilities of News U.K.’s head of consent are extensive and interdisciplinary. They are expected to collaborate closely with various departments including data protection, product development, ad tech, and commercial teams.

The goal is to establish robust compliance processes and utilize technology effectively for managing user consent—a critical aspect amidst tightening data privacy laws and heightened consumer awareness.

This role underscores a broader trend where publishers are grappling with the dual challenges of regulatory compliance and commercial imperatives. Balancing legal and ethical data practices with business objectives has become increasingly complex, necessitating specialized roles like head of consent to navigate these intricacies.

The emergence of AI further complicates this landscape, posing new challenges and opportunities in data privacy management. Publishers are now compelled to proactively address these issues rather than reactively responding—a strategic shift highlighted by the creation of roles such as head of consent.

Matthew Rance at Immediate Media serves as a parallel example, having pioneered a similar role focused on commercial data and analytics. His approach aligns closely with the goals News U.K. seeks to achieve, integrating consent management beyond mere compliance into a foundational aspect of business strategy.

At its core, the role of head of consent aims to empower users with transparency and control over their personal data—an approach that not only ensures regulatory compliance but also builds trust and enhances user experience.

While News U.K. has taken a pioneering step, the broader industry response remains varied. Some publishers may not yet see the immediate need for such a role, while others may struggle with resource constraints or competing priorities. However, the growing demand from advertisers for transparent data practices suggests that roles like the head of consent could become more commonplace in the future.

The trajectory of roles like the head of consent reflects a shift in how publishers view privacy—from a compliance burden to a potential competitive advantage. CEOs and industry leaders must embrace this shift, prioritizing privacy as a cornerstone of their organizational strategy to navigate future challenges effectively.

By doing so, they not only mitigate risks but also unlock opportunities for sustainable growth and consumer trust in an increasingly data-driven world.

Google has announced that its dark web monitoring service, known as “dark web report,” will be available to all Google users in 46 countries, including the US and UK, by the end of the month. This feature, which was previously exclusive to Google One members, aims to enhance online security by detecting if users’ personal information has been compromised on the dark web.

Users will be able to access the dark web report through the “Results about you” page on myactivity.google.com. The service scans for personal information such as names, addresses, phone numbers, and email accounts that may have been leaked.

The introduction of this feature is particularly beneficial for those who have experienced hacking, information leaks, or identity theft. For instance, the recent incident involving 7.6 million AT&T customers, whose information was leaked to the dark web, highlights the necessity of such protective measures.

While Google is expanding this service, it joins the ranks of other platforms like Proton Mail and LastPass, which already offer dark web monitoring to alert users if their information has been found. This move underscores the growing importance of digital security and the need for robust measures to protect personal data.

Microsoft unveiled its new MInference technology on the AI platform Hugging Face, showcasing a potential breakthrough in processing speed for large language models.

The demonstration, powered by Gradio, allows developers and researchers to test Microsoft’s latest advancement in handling lengthy text inputs for AI systems directly in their web browsers. This interactive demo highlights significant improvements in the efficiency and speed of processing long text inputs, which is a critical aspect of advancing AI technology.

MInference, which stands for “Million-Tokens Prompt Inference,” aims to significantly accelerate the pre-filling stage of language model processing, a common bottleneck for long text inputs.

According to Microsoft researchers, MInference can reduce processing time by up to 90% for inputs of one million tokens, equivalent to about 700 pages of text, while maintaining accuracy. This dramatic reduction in processing time addresses a significant challenge in deploying large language models widely.

The computational demands of large language model (LLM) inference are a substantial barrier, especially as prompt lengths increase. The attention to computation complexity often results in lengthy processing times.

For example, it takes 30 minutes for an 8 billion parameter LLM to process a prompt of one million tokens on a single Nvidia A100 GPU. MInference reduces this inference latency by up to 10 times for pre-filling on an A100 GPU, while maintaining accuracy, showing an 8.0x latency speedup for processing 776,000 tokens.

This innovative approach addresses the AI industry’s need for efficiently processing larger datasets and longer text inputs. As language models grow in size and capability, handling extensive context becomes crucial for applications ranging from document analysis to conversational AI.

The Gradio-powered demo enables developers to explore AI acceleration hands-on, potentially accelerating the refinement and adoption of MInference technology.

Beyond speed improvements, MInference’s ability to selectively process parts of long text inputs raises questions about information retention and potential biases. The technology’s approach to dynamic sparse attention might also reduce the energy consumption required for processing long texts, aligning with concerns about the carbon footprint of AI systems.

This could influence future research directions toward more sustainable AI technologies. Microsoft’s public demo of MInference intensifies competition among tech giants, potentially prompting rapid advancements in efficient AI processing techniques.

The coming months will likely see intense scrutiny and testing of MInference, providing valuable insights into its real-world performance and implications for the future of AI.

Meta AI researchers have introduced MobileLLM, a new way to make efficient language models for smartphones and other devices with limited resources. Published on June 27, 2024, this research challenges the idea that powerful AI models need to be very large.

The team, including experts from Meta Reality Labs, PyTorch, and Meta AI Research (FAIR), focused on creating models with fewer than 1 billion parameters, much smaller than models like GPT-4, which have over a trillion parameters.

The research team made significant innovations in MobileLLM by prioritizing model depth over width, implementing embedding sharing, grouped-query attention, and a novel block-wise weight-sharing technique.

These strategies allowed MobileLLM to outperform previous models of similar size by 2.7% to 4.3% on common benchmark tasks. Though the improvements seem minor, they are notable in the competitive field of language model development.

Yann LeCun, Meta’s Chief AI Scientist, highlighted these advancements on social media, underscoring the importance of the team’s work.

A remarkable achievement of MobileLLM is its 350 million parameter version, which demonstrated comparable accuracy to the much larger 7 billion parameter LLaMA-2 model in certain API calling tasks.

This finding suggests that more compact models could offer similar functionality to larger models while requiring significantly fewer computational resources. The ability to achieve high accuracy with smaller models is particularly valuable for applications on resource-constrained devices.

The development of MobileLLM reflects a broader trend towards more efficient AI models. As advancements in large language models begin to plateau, researchers are increasingly focusing on the potential of smaller, specialized designs.

MobileLLM’s efficiency and suitability for on-device deployment place it in the emerging category of Small Language Models (SLMs). Despite its name, MobileLLM aligns with this movement towards compact AI models that can perform effectively without massive computational power.

Although MobileLLM is not yet available for public use, Meta has open-sourced the pre-training code, enabling other researchers to build upon this work. This initiative may eventually lead to more advanced AI features on personal devices, though the timeline and specific capabilities are still uncertain.

The development of MobileLLM marks a significant step towards making advanced AI more accessible and sustainable, challenging the notion that only large-scale models can be effective and opening new possibilities for AI applications on personal devices.

Researchers at Osaka Metropolitan University have developed an artificial intelligence tool that can assess lung function by analyzing chest X-rays, a breakthrough that could revolutionize how we monitor respiratory health.

Traditionally, doctors use spirometry, which requires patients to blow into a device and follow specific instructions. However, this can be challenging for certain groups, like infants or those with cognitive impairments.

The AI model created by Associate Professor Daiju Ueda and Professor Yukio Miki was trained on a massive dataset of more than 140,000 chest X-rays collected over nearly two decades.

They compared the AI’s predictions with actual spirometry results and found a strong correlation, indicating its ability to accurately estimate lung function without the need for physical exertion from the patient.

This innovation addresses significant challenges in current healthcare practices by offering a non-invasive alternative for assessing pulmonary function. By relying solely on existing chest X-rays, the AI model simplifies the diagnostic process and potentially reduces medical costs associated with more complex tests.

Professor Ueda emphasized the collaborative effort involved in developing this AI model, which included contributions from healthcare professionals, researchers, and patients across various institutions.

He expressed optimism about the tool’s potential to alleviate patient burdens and improve healthcare efficiency by providing accessible and accurate lung function assessments.