The Formula Remains the Same, but the Mathematics Changes: Physicists Show Quantum World Can Be Described Without Imaginary Numbers
Quantum mechanics has relied for more than a century on numbers that cannot be placed on the ordinary number line. Now physicists have shown that it may be possible to dispense with these so-called imaginary numbers without altering any of the theory’s predictions.
Specialists at Heinrich Heine University Düsseldorf and the German Aerospace Center have re-examined one of the central mathematical pillars of quantum mechanics. Quantum mechanics describes the behavior of matter and energy at the scale of atoms and subatomic particles. The theory explains why particles can exhibit wave-like properties, tunnel through barriers, and maintain non-local correlations with one another—phenomena that underpin emerging technologies such as quantum computing and quantum communication.
Conventionally, quantum states are represented using complex numbers, each composed of a real part and an imaginary part. The real component encodes the magnitude of the state, while the imaginary component captures its phase. For decades, physicists regarded this mathematical structure as indispensable.
In 2021, another research group concluded that imaginary numbers cannot be eliminated from quantum mechanics, a claim supported by experimental results at the time. However, the Düsseldorf team led by Professor Dagmar Bruß and doctoral student Pedro Barrios Hita decided to scrutinize the assumptions underlying that earlier work.
The researchers discovered that one of the conditions imposed in the 2021 study was stricter than required by physical reality. By replacing it with a different mathematical requirement that still correctly describes composite quantum systems, they obtained a broad family of theories formulated exclusively with real numbers.
When expressed in this revised form, the new framework yields exactly the same observable results as standard quantum mechanics for every possible experiment. Consequently, the two mathematical models cannot be distinguished from each other through measurement.
The authors emphasize that they are not proposing to rewrite conventional quantum mechanics. Complex numbers remain a highly convenient and efficient calculational tool. Their study simply shows that the imaginary component does not necessarily reflect an intrinsic feature of nature; it is merely one of several mathematically valid ways to describe the quantum world.
Related articles
Anthropic Releases Eight Claude Code Updates in August Focused on Multi-Agent Workflows
Between August 13 and 21, Anthropic shipped eight consecutive Claude Code releases from version 2.1.232 to 2.1.239. The updates center on enabling multiple long-running agents that can share context, communicate across sessions, and continue work automatically after hitting usage limits. Key additions include subagent forking that inherits prompt cache and conversation history, cross-session messaging via @mentions, and an automatic mode that uses a classifier model to approve actions. Additional improvements cover GitLab merge request integration, memory management fixes for extended sessions, and support for native add-ons in musl-based Alpine environments. The changes significantly expand the scale of tasks that can be delegated to Claude Code without constant human oversight.
Selectel Revenue Exceeds 10 Billion Rubles as Cloud Services Drive Growth Amid Rising Costs
Selectel reported revenue of 10.2 billion rubles for the first half of 2026, marking a 14 percent increase compared with the same period a year earlier. Cloud infrastructure services remained the primary growth engine, contributing 8.9 billion rubles or 87 percent of total revenue. The customer base expanded by 14,400 clients over twelve months to reach 44,500, with smaller customers fueling much of the increase through new VDS configurations. Professional services showed the fastest consumption growth at 1.4 times, while financial and IT sectors raised infrastructure spending by 1.3 times. Adjusted EBITDA rose only 3 percent to 5.4 billion rubles, lowering its margin from 59 percent to 53 percent, and net profit stayed nearly flat at 1.9 billion rubles with margin declining from 21 percent to 18 percent. Operating expenses jumped 23 percent to 4.7 billion rubles, driven mainly by payroll costs that accounted for 65 percent of the total. The company invested 6.4 billion rubles in development, including 4.4 billion rubles on server equipment, a 1.8-fold increase, while expanding GPU purchases for AI workloads despite higher component prices, resulting in negative free cash flow of 3.4 billion rubles and a net debt to EBITDA ratio of 2.1.
Russian Backup Market Hits 10.6 Billion Rubles as Domestic Solutions Reach 70% Share
The Russian market for backup solutions grew 18.5% in 2025 to reach 10.6 billion rubles, outpacing the overall infrastructure software segment which expanded 16%. Domestic vendors increased their share from 19% in 2021 to 70% in 2025, with analysts from Strategy Partners forecasting further growth to 92% by 2030 and a market size of 24 billion rubles. Cyberprotect maintains the leading position with its Cyber Backup product, while Astra Group’s RuBackup is noted as another significant offering. The company itself estimates the market at approximately 12 billion rubles and claims a 56% share. Competition is shifting from feature lists toward compatibility with Russian virtualization platforms, databases, Kubernetes environments, storage systems, and monitoring tools. The broader infrastructure software market reached 158 billion rubles in 2025, with Russian developers holding 68% compared to just 8% four years earlier.
Google Issues Clear 2026 Guidance: No Separate SEO for AI Overviews or AI Mode
Google has published dedicated documentation clarifying that optimization for generative search features remains standard SEO. The company stresses that pages must be crawlable, indexable, relevant to user intent, and contain reliable information that the system can extract and synthesize. While classic ranking still applies, generative systems now evaluate pages across multiple stages including retrieval, fact extraction, cross-referencing, and citation decisions. Google explicitly rejects the need for llms.txt files, special AI-oriented Schema markup, or content rewritten specifically for language models. Instead, the focus is on non-commodity content that provides unique data, concrete metrics, and genuine user value. The guidance also warns against scaled content abuse and doorway-style pages created solely to capture long-tail AI-generated queries.