Archive
All Reports
Every published POPR report, in reverse-chronological order. Each originates from a source-backed Verified Research Record.
Uber's Robotaxi Strategy Is Bigger Than Self-Driving Cars: Inside the $10 Billion Plan to Become the Operating System for Autonomous Mobility
In February 2026, Uber formally launched Uber Autonomous Solutions and committed more than $10 billion to autonomous-vehicle expansion while working with more than 20 AV partners across multiple continents. The strategy is not about owning self-driving technology. It is about owning what comes after.
OpenAI IPO Economics in 2026: $40 Billion Revenue Run Rate, $750 Billion Compute Plan and the Race to Become Profitable
OpenAI is running at roughly $40 billion annualized while planning $750 billion in compute spending through 2030. Improving gross margins and a confidential S-1 filing don't settle the central question: can unit economics beat infrastructure economics?
Apple Negotiates Pay-Per-Use News Deals for Siri. Nine-Figure Plan Could Change How Publishers Get Paid for AI Answers.
Apple is discussing a nine-figure budget to pay publishers when Siri uses current news. The commercial mechanics are unresolved, the attribution problem remains unsolved, and the bigger question is not how to get an Apple deal.
Neota Logic Launches Multi-Model AI Orchestration for Legal Teams: Why Governing OpenAI, Claude, Google and Bedrock May Become the Next Enterprise AI Control Layer
The two-direction architecture is the finding that survives both hostile audits cleanest: AI can operate inside Neota's deterministic workflows, or an external AI assistant can call Neota when it needs a governed answer. Governance becomes a service the model calls, not a replacement for the interface the user prefers.
SpaceX Is Building Chips for Orbit: Inside the $16.8 Billion Terafab Plan and the Silicon-to-Satellite Race
SpaceX's own SEC filing connects Terafab to a space-optimized chip class for orbital compute. CesiumAstro just acquired RF-chip designer Jariet Technologies. Amazon built Prometheus three years ago. Three data points, one macro pattern: space companies are pulling semiconductor design closer to the mission.
AI Agents Are Getting Their Own Budget Layer: Why TokenOps, Run-Scoped Spending and Resource Governance Matter for Autonomous AI
TokenOps gives every agent run a shared ledger, a resource envelope and a deterministic enforcement layer. The larger story is that Cloudflare, Portkey, Microsoft, OpenTelemetry, OWASP and the Linux Foundation are independently assembling the same economic control plane.
Claude Now Watermarks AI Text in 2026, but It Does Not Prove Authorship: What Anthropic's New System Means for Publishers, Amazon KDP and Copyright
A detected Claude watermark is evidence of possible AI processing, not proof of authorship, accuracy or copyright. For publishers, Amazon KDP's AI-assisted versus AI-generated distinction is the more consequential rule.
Pokémon Turns 30, and Its Past Is Moving: FireRed and LeafGreen Return as Pokémon Bank Nears Its End
FireRed and LeafGreen are back on Nintendo Switch, old event-gated encounters have been reopened, Pokémon HOME is expanding, and Pokémon Bank now has a confirmed 2027 shutdown date. Together, these are not separate product announcements. They are a generational handoff.
Ray Dalio's 2026 AI Bubble Warning: Why AI Can Win While Investors Lose, and What the $1 Trillion Buildout Really Means
AI can be genuinely revolutionary and still create a financial bubble. BIS data confirms the $1 trillion buildout, rising debt financing, circular capital and growing private-credit risk, while Federal Reserve officials say a 2008-style crisis has not been established.
Publishing in 2026: AI Slop, Google Zero, Author Fraud and the Verification Economy Reshaping Books and News
AI-generated books, Google traffic loss, answer engines, unreliable AI detectors, author fraud and newsroom failures are converging on one scarce resource: trustworthy verification and provenance.
AI Agents Are Starting to Act Before We Know What They Understand: The Hardest Lesson From 2026's Cybersecurity Tests
Frontier AI systems have taken consequential actions against real infrastructure and real people during controlled cybersecurity evaluations. The evidence does not establish that those systems always understood they were operating in the real world. That distinction may be one of the most important AI safety problems of 2026.
AI Didn't Escape the Sandbox: What Four Real 2026 Security Incidents Reveal About Frontier AI Safety
Four separate AI security incidents were disclosed between July and August 2026 involving OpenAI and Anthropic models in cybersecurity evaluations. Each had a different cause, a different mechanism and a different severity. None of them is accurately described by a single headline about AI escaping containment.
The U.S. and Europe Are Regulating Different AI Risks First: What the 2026 Regulatory Split Actually Means
The United States and European Union are not simply choosing between light and strict artificial-intelligence regulation. The verified 2026 record shows something more specific: each jurisdiction is prioritizing a different layer of AI risk, using a different legal mechanism, on a different timeline.
Why Math and Physics Solvers Should Show Every Step in 2026: Projectile Motion, Vectors, Units and Verification
A math or physics solver can return the correct numerical answer while hiding nearly everything a student needs to understand how that answer was produced. Verified physics mathematics shows why the intermediate steps matter: a physical problem becomes variables, variables become components, components become equations, and those calculations still require assumptions, units, precision and verification before the final number has meaning.
Why Students Get Math and Physics Answers Wrong in 2026: What Misconceptions, Equations and STEM Research Reveal
Students can reach the same score on a physics test while holding materially different misconceptions about the underlying concepts. Research across physics, mathematics and chemistry education increasingly shows why this matters: the space between the question and the answer contains some of the most important information about whether learning actually occurred.
What Is Candida auris? Why the Drug-Resistant Fungus Can Hide in Hair Follicles and Keep Spreading Through Hospitals
Candida auris is a drug-resistant fungus that spreads primarily through healthcare facilities, often by colonizing people who have no symptoms. A peer-reviewed study published in Science in August 2026 identifies one mechanism that may help explain why the organism can remain on skin for extended periods: in mice, C. auris persisted inside hair follicles by suppressing the immune defenses normally used to clear fungi.
Marvel's Wolverine Is Not a Marvel-PlayStation Merger. It Is the Next Major Expansion of Insomniac's Marvel Universe on PS5
Marvel's Wolverine launches September 15, 2026, exclusively on PlayStation 5, developed by Insomniac Games and published by Sony Interactive Entertainment in collaboration with Marvel Games. The game is not evidence of a corporate merger. What it establishes is a continuing creative relationship expanding Insomniac's Marvel continuity beyond Spider-Man into a separate standalone Wolverine story.
Why the August 12, 2026 Total Solar Eclipse Is Unusual: Spain's 121-Year Wait, a Near-Perigee Moon and What Comes Next
The August 12, 2026 total solar eclipse is Spain's first since 1905, arrives approximately 2.2 days after lunar perigee and lasts roughly two minutes — followed in less than a year by a much longer eclipse reaching 6 minutes 23 seconds of totality.
What NASA Wants You to Watch During the August 12 Total Solar Eclipse: Not Just the Sky, but the Animals
NASA is using the August 12, 2026 total solar eclipse for more than astronomy. Alongside tracking the eclipse itself, the agency is running a citizen-science effort asking the public to observe how animals respond when daylight suddenly collapses into temporary darkness.
What Is Personal Superintelligence? Inside Mark Zuckerberg's New Political Philosophy for AI
Mark Zuckerberg is arguing that the most powerful artificial intelligence should be distributed broadly to individuals rather than concentrated inside a handful of companies or governments. The unresolved question is whether distributing models actually distributes power when Meta still controls enormous platforms, infrastructure and distribution networks.
Why Marvel's Wolverine Matters to PS5: How PlayStation Is Expanding Its Marvel Game World Beyond Spider-Man
Marvel's Wolverine is one of PlayStation 5's biggest confirmed 2026 exclusives, but its significance reaches beyond a single release. The game gives Sony Interactive Entertainment another major Marvel property inside the same Insomniac continuity already established through Spider-Man, while preserving a different structure, cast and identity around Wolverine.
What Is Meta Muse Glimmer? Why the New Local AI Model Signals Meta's Return to Open-Weight AI
Meta released Muse Glimmer on August 10, 2026, as a 30-billion-parameter open-weight AI model designed for agentic work on consumer hardware. The more significant story is not that Meta has made its most powerful AI open, because it has not. Meta appears to be building a two-track strategy: distributing smaller capable models broadly while retaining tight control over its most expensive frontier development.
What Is Terafab? Inside the $16.8 Billion Tesla-SpaceX Chip Project and the AI Infrastructure Empire Forming Around It
Terafab is not simply Elon Musk's attempt to build an enormous semiconductor factory. The planned Texas complex connects a system stretching from AI models and data centers to rockets, satellites, autonomous vehicles, robots and the chips underneath them. The more important question it raises is what happens when AI competition moves beyond models toward control of the infrastructure required to run them.
Does Siri Offer a Clearly Identifiable Black American Voice on iPhone 17 Pro Max and iOS 26? What Our 2026 Audit Found
Apple introduced Siri voices in 2021 that many Black users perceived as Black. A 2026 POPR Newsroom audit of an iPhone 17 Pro Max running iOS 26 found no explicitly labeled Black American voice and no option the consumer personally perceived as clearly representing Black American speech.
What Is Generative Engine Optimization? What GEO Actually Does and What the Research Really Shows
GEO is a real and developing field. Academic research has measured meaningful effects, production evidence shows it can work at scale, and 2026 research also shows that some popular techniques actively reduce visibility when tested against a realistic retrieval pipeline.
Can Alcohol Brain Damage Be Reversed? What Brain Imaging Shows After You Stop Drinking
Human brain imaging research documents measurable recovery in gray matter and white matter during sustained abstinence, sometimes beginning within weeks and continuing for months. The evidence does not show that every injury disappears — but it does show that stopping changes the biological trajectory.
Rhonda Patrick Says Creatine Is More Than a Muscle Supplement. What the Brain Research Actually Shows
Dr. Rhonda Patrick has spent years drawing public attention to creatine's possible effects on brain energy, cognition, sleep deprivation and mental health. Human trials now support several parts of that argument, but the evidence does not show that creatine replaces sleep, works as a universal nootropic or should be used at research-level doses without medical context.
Apple's First Foldable iPhone Looks Increasingly Real: Everything We Know Before the 2026 Launch
Apple has not announced a foldable iPhone, but the evidence surrounding one is becoming unusually difficult to dismiss. Supplier reports point to 2026 production, iOS 27 contains folding and multi-display code, and current leaks converge around a book-style design with a roughly 7.8-inch inner screen, Touch ID, a dual-cell battery and a premium price approaching $2,000 or more.
Apple Didn't Replace Siri With Gemini: Inside the New AI Architecture Behind iOS 27
Apple's new Siri is powered by Apple Foundation Models, not Gemini. Apple collaborated with Google on model development and runs its most capable server model on NVIDIA GPUs in Google Cloud. The architecture is considerably more complicated than the headline version of this story.
Michael Saylor Says He Used ChatGPT to Help Build a Financial Product. The SEC Record Shows What Happened Next
Michael Saylor says ChatGPT helped him develop STRK, a real preferred security issued by Strategy. The financial record confirms the security and the capital it raised, but it does not establish that ChatGPT was indispensable or that STRK itself raised $15 billion. The more consequential story may be how AI is entering financial-product design.
Who Really Pays Tariffs? What the 2026 Evidence Says About Your Money
Tariffs are collected from American importers, but the cost does not necessarily stop there. Federal Reserve research shows how tariffs move through businesses, prices and household spending, while the latest 2026 estimates put the average household impact at about $1,100 a year under current law.
Does Ozempic Cause Muscle Loss? What the 2026 Research on GLP-1 Drugs and Lean Mass Actually Shows
Semaglutide, tirzepatide and other incretin-based weight-loss treatments can reduce lean mass along with fat. However, lean mass is not the same thing as skeletal muscle, most of the weight lost is still fat, and new 2026 research shows that strength and muscle health can behave differently from a DXA lean-mass measurement.
Can Ozempic or Semaglutide Reduce Alcohol Cravings? What the New 2026 GLP-1 Research Actually Shows
A major 2026 clinical trial found that semaglutide reduced heavy drinking and alcohol craving in adults with alcohol use disorder and obesity. The VA is now running a Phase 3 trial. The evidence is real, but the drug is not approved for this use and major questions remain.
Plasmalogens, APOE4 and Alzheimer's: What the New 2026 Brain Lipid Research Actually Shows
Plasmalogens are important membrane lipids that are depleted in Alzheimer's-affected brain tissue, and new 2026 research has linked lower plasmalogen carriage directly to the APOE4 genetic risk variant. However, evidence that oral supplements can rebuild brain white matter, prevent Alzheimer's disease, or reverse neurodegeneration in humans has not yet been established.
Can Scientists Really Rejuvenate Zombie Cells? What the AP2A1 and IU1 Anti-Aging Studies Actually Found
Scientists have discovered new ways to change features associated with cellular aging and improve cellular cleanup systems, but no study has shown that humans can currently rejuvenate zombie cells or dramatically extend lifespan. Two discoveries involving AP2A1 and IU1 are real. The viral story created by combining them is not.
Why Does Claude Code Use So Much of Your Usage Limit? What Actually Burns Usage and How to Fix It
Claude Code usage can rise quickly because of reasoning effort, long sessions, tool activity and context consumption. Anthropic also confirmed that a resolved caching bug temporarily caused some users' limits to drain faster than expected in spring 2026. Here is how to separate that past incident from the usage controls that matter today.
How to Get Your Website Into ChatGPT and Google AI Answers in 2026
Answer Engine Optimization is becoming a major concern for website owners, but Google and OpenAI's own documentation shows that appearing in AI-generated answers depends far more on crawlability, indexing, useful content and the right crawler permissions than on a new collection of AI-specific SEO tricks.
The Replit Google Gap: Why a Deployed App Can Still Be Invisible to Google
A verified POPR investigation into Replit deployment, JavaScript rendering, URL discovery, and why getting a web application online is not the same thing as getting it discovered by Google.
Room-Temperature Quantum Computing Is Becoming Real. The Hard Part Is What Comes Next
New diamond quantum computers, room-temperature quantum materials and photonic communication devices are weakening one of quantum technology's oldest infrastructure constraints. But removing cryogenic cooling does not solve the harder problems of scaling, entanglement, fidelity, error correction and useful quantum advantage.