Top 10 in Tech – August 2024

1- Has the AI bubble burst? Wall Street wonders if artificial intelligence will ever make money

Tech giants have heavily invested in AI, spending billions on infrastructure to support AI advancements like ChatGPT, but the returns have been underwhelming, causing investor concern. Despite promises of AI revolutionizing industries, the tangible revenue from AI products has been minimal, leading to doubts about the sustainability of such massive investments. Companies like Amazon, Microsoft, and Google continue to invest, betting on long-term returns, but the pressure to scale back spending is mounting as investors grow impatient with the lack of immediate monetization.

Source: CNN

2- Microsoft plans September cybersecurity event to discuss changes after CrowdStrike outage

Microsoft will host a cybersecurity summit on September 10 at its Redmond campus to address industry concerns following a faulty update from CrowdStrike that caused millions of Windows computers to crash in July. The event will bring together industry peers to explore alternatives to using the privileged kernel mode, which contributed to the widespread disruptions, and will discuss the adoption of safer technologies like eBPF and memory-safe programming languages such as Rust. The gathering aims to prevent similar issues in the future and includes participation from major cybersecurity firms, including CrowdStrike.

Source: CNBC

3- How Google makes custom chips used to train Apple AI models and its own chatbot, Gemini

Google’s Mountain View lab is focused on testing its custom Tensor Processing Units (TPUs), AI chips that have positioned the company as a significant player in the cloud AI market. These TPUs, crucial for training AI models like Google’s Gemini chatbot and used by companies like Apple, have helped boost Google’s cloud revenue. Despite challenges, including competition from Nvidia and geopolitical risks, Google remains committed to advancing its AI chip technology, recently unveiling its first general-purpose CPU, Axion, and continuing efforts to improve efficiency and sustainability in its data centers.

Source: CNBC

4- California allows Chinese robotaxi firm WeRide to test with passengers

China-based autonomous driving startup WeRide has received approval from California’s Public Utilities Commission to test its driverless vehicles with passengers in San Jose and surrounding areas. The three-year permit allows WeRide to operate with or without a safety driver but prohibits offering rides to the public or charging fares. This approval comes as WeRide seeks a $5 billion valuation for its New York IPO amid potential U.S. restrictions on vehicles with China-developed systems. WeRide has previously secured driverless permits in Singapore and the UAE.

Source: Reuters

5- Nvidia second-quarter sales likely to double, even a slight miss may hurt shares

Nvidia is expected to report a more than doubling of its second-quarter revenue, driven by strong demand for its AI chips, particularly from tech giants like Microsoft. However, concerns are rising among investors about whether Nvidia can sustain its rapid growth, especially with potential production delays and increasing costs. The company’s stock, which has surged over 150% this year, faces scrutiny as any miss on expectations could impact the broader AI market. Nvidia’s future growth could also be affected by geopolitical issues and antitrust investigations.

Source: Reuters

6- Crucial Lessons Learned For Cybersecurity Resilience

A global computer outage recently caused major disruptions in air travel, highlighting the vulnerability of interconnected systems. This incident, triggered by a faulty update from CrowdStrike, underscores the critical need for robust quality assurance and resilient system designs. While it wasn’t a cyberattack, the event serves as a wake-up call for organizations to improve their preparedness and resilience strategies. The focus should now be on balancing innovation with reliability and ensuring that systems can recover quickly from failures to prevent widespread disruptions.

Source: Forbes

7- What Is An AI-First Mindset?

As AI becomes increasingly integrated into daily life, people are adopting an “AI-first mindset,” using AI tools as their default for tasks like writing, searching for information, and problem-solving. This shift is changing behavior, with AI becoming a primary tool for decision-making and creativity. An AI-first mindset involves leveraging AI to enhance efficiency while recognizing its limitations. It also emphasizes the importance of soft skills, as they complement AI’s capabilities, and encourages a growth mindset focused on continuous learning and adaptation.

Source: Forbes

8- Apple is expected to debut the first generative AI iPhone at its September 9 event

Apple has announced its next major event for September 9, where the iPhone 16 is expected to be unveiled, featuring enhanced AI capabilities. The event, tagged “It’s Glowtime,” will be held at the Steve Jobs Theater and streamed online. The iPhone 16 is anticipated to integrate AI more deeply, building on features like a smarter Siri and personalized AI-generated content. This launch could significantly impact Apple’s trajectory, especially amid challenges like declining sales in China and competition in the AI space.

Source: CNN

9- The Vacuum of Space Will Decay Sooner Than Expected

Vacuum decay, a theoretical process that could end the universe, might occur 10,000 times sooner than previously thought, but still not for an unimaginably long time. This process involves the Higgs field, which controls the mass of particles, potentially tunneling to a lower-energy state, triggering a catastrophic transformation of the universe. Recent calculations suggest this could happen in 10^790 years instead of 10^794 years, but the difference is negligible given the immense timescale involved.

Source: Wired

10- Why AI Models Are Collapsing And What It Means For The Future Of Technology

Artificial intelligence’s progress in areas like customer service and content creation is at risk due to a phenomenon known as “model collapse,” where AI models, when trained on data that includes their own previous outputs, gradually produce less accurate and reliable content. This recursive training loop degrades the model’s performance and can lead to poorer decision-making, reduced content quality, and increased biases. To prevent this, it’s essential to use high-quality human-generated data, maintain transparency in training methods, and periodically refresh models with new, diverse data to preserve AI’s effectiveness and relevance.

Source: Forbes