Apple’s September 9 “Surprise and Shine” Event: iPhone 18 Pro, First Foldable iPhone & New Apple Watch Era
Nazima 12:05 pm September 4, 2026 Apple’s September 9 “Surprise and Shine” Event: iPhone 18 Pro, First Foldable iPhone and a New Apple Watch Era Apple is set to hold its next major product event on Tuesday, September 9, 2026, at 10 a.m. PT (1 p.m. ET) at the Steve Jobs Theater in Apple Park. The event, reportedly titled “Surprise and Shine,” is expected to introduce the next generation of Apple hardware, led by the iPhone 18 Pro and iPhone 18 Pro Max, alongside Apple’s long-rumored first foldable iPhone and new Apple Watch models. The event could also mark an important leadership moment for the company. John Ternus is expected to lead his first major Apple product launch as CEO, following Tim Cook’s transition to executive chairman on September 1, 2026. That gives this year’s September event significance beyond the products themselves: it will offer an early glimpse at the direction of Apple under its new leadership. iPhone 18 Pro and Pro Max: Focused Upgrades, Higher Prices The iPhone 18 Pro lineup is widely expected to build on the existing Pro design rather than introduce a complete visual overhaul. Instead, Apple is reportedly focusing on improvements in performance, power efficiency, battery life and camera capabilities. At the center of the upgrade could be Apple’s next-generation A20 Pro chip, reportedly manufactured using a 2nm process. A smaller manufacturing process could allow Apple to improve both performance and efficiency, potentially delivering faster processing while reducing power consumption. For users, that could mean smoother performance, stronger battery endurance and better handling of demanding tasks such as gaming, photography and on-device artificial intelligence. Pricing Could Become a Major Story One of the biggest questions surrounding the iPhone 18 Pro is its price. Recent reports have pointed to higher component costs, particularly for memory and other advanced components, as Apple faces increasing pressure across the supply chain. Earlier speculation suggested that the Pro lineup could see significantly higher starting prices, while newer reports have pointed to a possible $100 increase, potentially bringing the entry-level iPhone 18 Pro to around $1,199, compared with $1,099 for the previous generation. Nothing about the final pricing should be considered official until Apple announces it. Still, even a $100 increase would make the Pro lineup more expensive at a time when consumers are already facing higher smartphone prices. The broader strategy appears straightforward: Apple may be prioritizing meaningful internal improvements while using the Pro lineup to support higher margins rather than relying on a dramatic redesign. Apple’s First Foldable iPhone Could Steal the Show The most significant announcement of the event could be Apple’s long-awaited entry into the foldable smartphone market. The device is widely rumored to carry the iPhone Ultra name, although Apple has not officially confirmed either the product or its branding. Reports suggest that the phone could use a book-style folding design, combining a conventional outer display with a much larger internal screen. One widely circulated configuration includes a 5.3-inch external display and a 7.8-inch internal display when unfolded. The device is also rumored to support MagSafe, use the A20 Pro chip, include 12GB of RAM, and feature a dual-camera system rather than the triple-camera arrangement found on current Pro models. Another notable rumor is the possible return of Touch ID for the foldable device instead of Face ID. However, this remains unconfirmed and should be treated strictly as part of the pre-launch speculation surrounding the product. A Premium Device With a Premium Price Apple’s foldable iPhone is expected to sit firmly at the top of the smartphone market. Several reports have suggested a starting price of more than $2,000, potentially making it one of Apple’s most expensive mainstream consumer devices. That pricing would place the foldable iPhone well into premium territory and distinguish it from conventional flagship smartphones. Rather than competing primarily on affordability, Apple appears likely to position the device around design, durability, software integration and the wider Apple ecosystem. Apple is entering a market where competitors such as Samsung have already spent years refining foldable hardware. Its late arrival could therefore be less about being first and more about producing a device that feels sufficiently mature to persuade existing iPhone users to make the switch. For Apple, the foldable iPhone could become more than another product category. It may represent the company’s biggest form-factor change to the iPhone in years. Apple Watch Series 12 and Ultra 4 The iPhone lineup is expected to be joined by new Apple Watch hardware, reportedly including the Apple Watch Series 12 and Apple Watch Ultra 4. Details remain limited, but the new models are expected to follow Apple’s established strategy of improving performance while refining health, fitness and battery-related features. Faster processors, improved sensors and incremental improvements to battery efficiency are among the areas that could see attention. Unlike the rumored foldable iPhone, however, the Apple Watch updates are expected to be evolutionary rather than transformational. The focus is likely to remain on making the existing experience faster, smarter and more useful rather than dramatically changing the form factor. Why Apple’s September Event Matters This year’s event could be particularly important because it brings together three major storylines for Apple: a new iPhone generation, the company’s possible entry into foldable phones and the beginning of a new CEO era. For consumers, the decision may ultimately come down to priorities. Users upgrading from an older iPhone could find the iPhone 18 Pro appealing for its expected performance, efficiency and camera improvements. Those looking for a fundamentally different smartphone experience will likely be far more interested in Apple’s rumored foldable device. The bigger question is whether Apple can justify the premium that appears to accompany both products. The iPhone 18 Pro may represent refinement, while the rumored iPhone Ultra could represent experimentation on a much larger scale. With John Ternus taking center stage for his first major product launch as CEO, Apple’s September 9 event has the potential to signal not only what the company is selling next, but
More Than 100 Tech Companies Warn of Growing AI-Powered Cyber Threats
Nazima 7:52 am September 1, 2026 More Than 100 Tech Companies Warn of Growing AI-Powered Cyber Threats More than 100 major technology companies, AI labs, and cybersecurity firms have signed an urgent open letter warning that artificial intelligence is making sophisticated cyberattacks easier, faster, and more accessible. The coalition says critical infrastructure—including hospitals, water utilities, and essential digital networks—could face increasingly automated attacks as AI systems become more capable. The companies are calling for governments, businesses, and technology providers to strengthen defenses before attackers gain an even greater advantage. A United Industry Warning The letter brings together some of the biggest names in technology and cybersecurity, including OpenAI, Google, Microsoft, Anthropic, Amazon Web Services, CrowdStrike, Fortinet, and Okta. According to the signatories, advanced AI models can help attackers automate activities such as finding software vulnerabilities, developing malicious code, and creating highly targeted phishing campaigns. As AI systems improve at coding, reasoning, and automation, cyber capabilities that once required highly skilled teams or significant resources could become accessible to smaller and less sophisticated criminal groups. This is what makes the threat different: AI can potentially increase the speed, scale, and efficiency of cyberattacks. Critical Infrastructure Faces Greater Risk The warning places particular attention on essential services such as healthcare, water systems, and public infrastructure. Many of these organizations still depend on older technology, limited security budgets, and small IT teams. These challenges can make it difficult to keep systems fully updated and protected. Key weaknesses include: Legacy technology: Older systems may contain known vulnerabilities and can be difficult to replace.Security misconfigurations: Poorly configured software and excessive administrative access can create additional attack opportunities.Limited budgets: Smaller public institutions may not have enough funding for advanced cybersecurity tools.Staff shortages: There is a continuing shortage of skilled cybersecurity and IT professionals. Critical infrastructure also presents a unique challenge because systems cannot always be taken offline for long periods to install updates or fix vulnerabilities. A cyberattack against a hospital, for example, could interfere with essential medical systems. An attack on a water utility could potentially disrupt operations that communities depend on every day. What the Industry Is Calling For The companies argue that traditional cybersecurity measures alone may not be enough to keep pace with increasingly automated threats. They are calling for stronger cooperation and faster investment in defensive technology. Priority What It MeansDefensive AI: Use AI to detect unusual activity, identify threats, and speed up security responses and patching. Threat Intelligence Sharing: Improve real-time information sharing between technology companies, security teams, and government agencies. Public-Private Investment: Provide financial support to hospitals, utilities, and other organizations that cannot easily afford advanced security infrastructure. Reduce Technical Debt: Replace outdated systems, close known security gaps, and strengthen basic protections such as multi-factor authentication. The goal is not simply to respond to attacks after they happen, but to improve defenses before automated threats become widespread. The AI Cybersecurity Challenge The growing concern comes from a fundamental reality of artificial intelligence: the same technologies that can help defend computer systems can also be used to attack them. AI can help security teams identify suspicious activity, analyze large volumes of data, and respond to threats more quickly. At the same time, attackers may use similar capabilities to search for weaknesses, automate malicious activity, and scale their operations. This creates a race between offensive and defensive AI. The companies behind the open letter believe the world still has an opportunity to strengthen its defenses, but that the window is narrowing. Governments and technology providers will need to ensure that advanced cybersecurity tools reach vulnerable organizations—including hospitals, municipalities, and public utilities—before attackers gain the upper hand. Why It Matters The warning is more than a prediction about future cyberattacks. It reflects a growing concern across the technology industry that AI is changing the economics of cybercrime. As advanced AI tools become more capable and widely available, organizations with limited security resources could face increasingly sophisticated threats. The challenge now is to make sure defensive innovation moves faster than offensive automation. Recent Posts
Judge Blocks Pentagon’s Anthropic Blacklist, Calls It Unlawful
Nazima 4:39 am August 29, 2026 Judge Blocks Pentagon’s Blacklist of Anthropic, Calls It Unlawful A U.S. federal judge has blocked the Pentagon from blacklisting Anthropic, the AI company behind Claude. The court found that the government’s decision to label Anthropic a “supply-chain risk” was unlawful and violated key legal protections. The ruling was issued by U.S. District Judge Rita F. Lin in the Northern District of California. Why Did the Pentagon Target Anthropic? The dispute began over how the U.S. military wanted to use Anthropic’s AI technology. Anthropic has restrictions on certain high-risk uses of Claude, including some applications involving autonomous weapons and mass surveillance. The company did not agree to give the Pentagon unrestricted access for all military uses. The disagreement became more serious when Defense Secretary Pete Hegseth designated Anthropic as a “supply-chain risk.” The designation led to restrictions on Anthropic’s work with the Department of Defense and affected companies working with the government that used Anthropic’s technology. Why Did the Court Reject the Government’s Action? The judge identified three major legal problems. First, First Amendment:The court found that the government’s actions were connected to Anthropic’s public position on AI safety and its policies for how its technology could be used. Second, Due Process:The court found that Anthropic was not given the required legal process before facing such serious restrictions. Third, Administrative Law:The court found that the government had not provided a sufficient legal and factual basis for calling Anthropic a supply-chain security risk because of a disagreement over AI-use policies. What Happens to Anthropic Now? The ruling blocks the challenged restrictions against Anthropic. The company can continue working with federal agencies and businesses that have government contracts, subject to normal procurement rules. However, the ruling does not mean that the Pentagon must buy Anthropic’s AI products. The government can still decide which AI systems it wants to use. The key point is that the government cannot use a broad supply-chain-risk designation as a way to punish a company without a proper legal basis. Why Is This Important for the AI Industry? The case highlights a growing conflict between AI companies that want limits on how their technology is used and governments that want greater access to advanced AI for defense and national security. It also raises an important question about how much control AI companies should have over the use of their technology. For the wider AI industry, the ruling shows that companies can establish safety and usage policies, while government agencies must still operate within constitutional and legal limits. Bottom Line The Pentagon wanted greater access to Anthropic’s AI technology, while Anthropic wanted to keep restrictions on certain military uses. The dispute eventually led to the government labeling Anthropic a supply-chain risk. The court ruled that the government’s response went too far. The decision does not give Anthropic control over government procurement. Instead, it reinforces the principle that the government must have a proper legal and factual basis when restricting a technology company. As artificial intelligence becomes more important to defense and national security, the Anthropic case could play an important role in future debates over AI safety, government power, and the rights of technology companies. Recent Posts
AI’s Macro Impact: Inflation, Rising Debt, and Investor Fatigue
Nazima 6:24 am August 27, 2026 AI’s Macro Impact: Inflation Pressure and Rising Debt Fatigue Artificial intelligence is rapidly moving beyond the technology sector and becoming an increasingly important macroeconomic force. Its impact is emerging through two closely connected channels: the investment boom required to build AI infrastructure is creating near-term inflationary pressure, while the growing reliance on debt to finance that expansion is beginning to test investor capacity and appetite. This creates a complicated economic transition. AI has the potential to raise productivity, expand productive capacity, and reduce costs over the long term. Yet, during the build-out phase, the surge in investment, demand for scarce inputs, and rapid growth in corporate borrowing can generate inflationary pressures and tighter financial conditions. The following analysis brings together recent observations from central banks, the Bank for International Settlements (BIS), market participants, and financial analysts as of August 2026. 1. Swiss National Bank: AI Could Push Inflation Higher in the Short Term What the SNB Says Petra Tschudin, a Governing Board member of the Swiss National Bank (SNB), said in a late-August 2026 interview that artificial intelligence could push inflation higher in the short to medium term. The long-term effect, however, remains uncertain. A key reason is that investment is being redirected toward AI-related activities. As capital, labor, and other resources move toward the AI sector, parts of the broader economy may face adjustment difficulties and capacity constraints. One of the clearest examples is the supply of semiconductors. If demand for chips and other critical AI inputs grows faster than supply, shortages can emerge and push prices higher. Similar constraints can arise in infrastructure, energy, data-center capacity, and specialized labor. Over the longer term, AI could have the opposite effect. Greater productivity and lower production costs could create disinflationary pressure. However, higher productivity does not automatically result in sustained deflation. For inflation to turn structurally negative, prices would need to fall repeatedly and persistently rather than simply increase at a slower rate. The SNB’s latest forecast continues to place Swiss inflation within its 0–2% target range through the first quarter of 2029. Importantly, this is a conditional forecast based on unchanged policy rates, rather than a commitment to keep rates unchanged for the entire period. Tschudin also emphasized that the SNB would respond to new inflation information and adjust monetary policy when necessary. The central bank does not publish a fixed interest-rate path. Why This Matters The SNB’s assessment reflects a broader concern among central banks: AI is no longer simply a technology-sector story. It is becoming a macroeconomic shock that can influence demand and supply simultaneously. That combination makes the economic cycle harder to interpret. Strong AI investment can increase demand today, while productivity gains may expand supply later. For central banks, distinguishing between temporary inflationary pressure and longer-lasting changes becomes considerably more difficult. 2. BIS and Other Central Banks: AI Is Blurring Inflation Signals The Bank for International Settlements and other policymakers have highlighted similar risks. According to a July 2026 BIS Bulletin, the AI boom is driving a large and increasingly debt-financed investment surge. This expansion is affecting trade, equity markets, wealth effects, and terms of trade, with the consequences varying across countries. The central challenge is that AI can influence both sides of the economy at the same time. On the demand side, AI investment increases spending on infrastructure, equipment, energy, technology, and services. Rising equity valuations can also generate wealth effects that encourage additional spending. On the supply side, successful AI adoption could increase productivity and expand the economy’s productive capacity. This creates a difficult policy environment. If demand rises faster than supply during the early stages of the AI build-out, inflationary pressures can increase. Later, once productivity gains become more widespread, the additional supply generated by AI could help contain inflation. ECB Executive Board member Philip Lane has identified another important transmission channel. If households and businesses quickly interpret AI as a permanent productivity improvement, they may bring forward spending in anticipation of higher future income. That shift in current demand can itself create upward pressure on inflation during the transition. IMF research cited through central-bank channels has similarly suggested that higher productivity from AI does not necessarily translate into lower inflation in a simple or immediate way. In other words, AI represents a dual-edged macroeconomic force: potentially inflationary during the investment and adjustment phase, but potentially disinflationary as productivity gains become more established. 3. US Corporate AI Debt: Rapid Growth and Emerging Investor Fatigue The inflation story is only one side of the equation. The other is how the AI investment boom is being financed. A Rapid Increase in Debt US technology “hyperscalers” — including Microsoft, Amazon, Alphabet, Meta, and Oracle — have significantly increased borrowing to finance AI-related capital expenditure. According to the estimates cited in the underlying sources, AI-related debt issuance by hyperscalers reached approximately $220 billion in 2026 by early to mid-August. That compares with much smaller issuance levels in 2024 and 2025, although estimates vary depending on how AI-linked borrowing is defined. Another estimate placed AI-linked gross issuance at approximately $98 billion in 2025 and around $200 billion in 2026, suggesting that borrowing could roughly double within a year as infrastructure investment accelerates. A separate, broader estimate has placed “hidden debt” associated with technology companies at approximately $1.65 trillion, representing an eightfold increase over four years. This figure should not be directly compared with conventional bond issuance because it refers to a broader category of off-balance-sheet and structured financing. The underlying purpose of this borrowing is largely the same: funding the enormous capital requirements of the AI build-out, including data centers, GPUs and other AI chips, electricity infrastructure, networking equipment, and related facilities. Investors Are Becoming More Selective The challenge is no longer simply whether investors are willing to fund large technology companies. The issue is how much additional debt the market can absorb without demanding significantly higher compensation. Reports cited in the draft indicate signs of growing investor fatigue as
Selective Activation Sparsity: How Smarter AI Computation Improves Efficiency
Nazima 4:49 am August 22, 2026 Selective Activation Sparsity in AI: How Smarter Computation Could MakeModels More Efficient What Is Selective Activation Sparsity? Artificial intelligence has traditionally improved by making models larger, adding more parameters, and increasing the amount of computation available to them. While scaling remains important, researchers are increasingly exploring another question: Does an AI model really need to use everything it has learned for every task? Selective activation sparsity is based on the idea that it does not.Instead of activating a large portion of a model for every input, a sparse system attempts to activate only the neurons, pathways, layers, or subnetworks that are most relevant to the task at hand. Less relevant components remain inactive. Think of it like having a large team of specialists. A company may employ hundreds of people, but a marketing project does not require every employee to attend every meeting. The goal is to bring in the people with the knowledge needed for that specific project. AI systems can follow a similar principle: keep the model’s capabilities available, but use only the computation that is necessary. Activation sparsity is already an active area of research, with studies examining how reducing weakly contributing activations can improve the efficiency of large language models. How Does Selective Activation Sparsity Work? The basic concept is relatively straightforward.A conventional dense model may process an input through a broad set of computational pathways. A model using selective activation tries to determine which parts are useful for the current input and suppress the rest. For example, imagine an AI system receiving two different requests. For a mathematics problem, the model may benefit more from pathways that have learned patterns related to mathematical reasoning. For a translation request, language-related pathways may be more useful. The model does not necessarily need to make every component equally active for both tasks. The process can be understood through three main ideas:1. The Model Learns What MattersDuring training, the model can learn which internal features are useful for different types of inputs. Over time, certain neurons or computational pathways may become strongly associated with particular patterns, tasks, or features. 2. Relevant Computation Is SelectedWhen the model receives an input, it can determine which parts of its internal computation are most relevant. Rather than treating every component as equally important, it selectively activates a smaller subset. 3. Unnecessary Computation Is SuppressedComponents that are not needed for the current operation remain inactive or contribute less to the computation. This creates a sparse activation pattern rather than a fully active one. The practical objective is simple: perform useful computation without spending resources on computation that contributes little to the current task. Why Does Activation Sparsity Matter?Large AI models can require substantial computational resources. Training and running them may involve powerful accelerators, large amounts of memory, high bandwidth, and significant energy consumption. That makes efficiency increasingly important.Activation sparsity offers one possible way to reduce the amount of computation performed during inference. Research has specifically investigated whether sparse activations can reduce computation and memory movement while preserving model performance. The potential benefits include:Lower Inference CostsIf a model performs less computation for each request, the cost of serving AI applications could potentially decrease.This could matter particularly for organizations running large numbers of AI requests. Faster AI SystemsReducing unnecessary computation may improve inference speed, although the actual improvement depends heavily on how sparsity is implemented and whether the underlying hardware can efficiently skip inactive values. Lower Energy ConsumptionLess computation can potentially translate into lower energy requirements, which is particularly relevant as AI systems become more widely deployed. More Capable Edge DevicesEfficient computation could make advanced AI more practical on devices with tighter computational and power constraints, such as smartphones and laptops. This does not mean that sparsity automatically makes every model faster. Efficient software, sparse kernels, memory behavior, and hardware support all matter. Research on activation sparsification has demonstrated practical speedups in some settings, but results vary by method and model. Selective Activation Sparsity vs. Mixture of ExpertsSelective activation sparsity is closely related to the broader idea of Mixture of Experts (MoE). A Mixture-of-Experts model contains multiple specialized components, often called experts, and a routing mechanism determines which experts should process a particular input. The connection is intuitive: both approaches are concerned with avoiding unnecessary computation by activating only a relevant subset of a larger system. However, they are not identical.Mixture-of-Experts typically focuses on selecting among expert modules, while activation sparsity can operate at a finer level by reducing the number of active neurons, channels, or other internal computations. Recent research continues to explore the relationship between activation sparsity and expert-based architectures, including approaches that combine sparse activation with expert routing. What Are the Main Challenges? The promise of selective activation sparsity is significant, but several important problems remain. Choosing the Wrong Path Can Hurt AccuracyA sparse system must decide which computations are important.If it suppresses a pathway that turns out to be necessary, the model may produce a weaker answer—not because the model lacks the required knowledge, but because the relevant computation was never activated.This makes routing and selection critical.Sparsity Does Not Automatically Mean Speed One of the biggest misconceptions about sparsity is that removing computations from a model automatically makes the model faster. In practice, hardware needs to be capable of efficiently skipping inactive values.If the processor still performs much of the same underlying work, the theoretical sparsity may not translate into a meaningful real-world speed improvement. Research on activation sparsification explicitly points to hardware and implementation as important factors in achieving practical inference gains. Routing Adds Additional WorkA selective system needs a mechanism to determine which components should be activated. That decision itself requires computation.If routing becomes too complicated or expensive, some of the efficiency gained from sparsity can be reduced. Errors Can Be Harder to DiagnoseIn a dense model, many pathways may contribute to an output.In a sparse model, a poor result could occur because the model selected an inappropriate subnetwork rather than because the