AISЦена AISwap

The digital labor market is witnessing a reversal as AI agents, equipped with their own crypto wallets, begin to hire humans for tasks requiring biological nuance—such as CAPTCHA solving, emotional reasoning, or content creation. This shift marks the emergence of a Machine-to-Human (M2H) economy, positioning platforms like SUBBD Token ($SUBBD) as key infrastructure for AI-to-human payroll systems. Targeting the $85B creator economy, SUBBD offers Ethereum-based smart contracts to reduce intermediary fees and integrates AI tools like voice cloning and digital influencer creation, enabling creators to scale their output while retaining ownership. The project has raised over $1.4M in its presale, reflecting institutional interest in hybrid AI-human workflows. A staking mechanism offering 20% APY aims to incentivize long-term holding. Regulatory challenges around AI-generated content remain, but the platform uses blockchain for transparency and governance.

Anthropic's announcement of Claude Mythos, an AI tool claiming to discover thousands of zero-day vulnerabilities—including a 27-year-old bug in OpenBSD—triggered alarm on Wall Street and prompted emergency meetings among financial regulators fearing systemic cyberattacks. However, independent tests revealed significant exaggerations in these claims. Researchers found that many reported vulnerabilities were in obsolete software or were impractical to exploit, and the findings relied on only 198 manual reviews. Furthermore, multiple smaller, open-source AI models (some with as few as 3 billion parameters) successfully identified the same critical flaws at a fraction of the cost, demonstrating that AI cybersecurity capability does not linearly scale with model size. Meanwhile, users reported severe performance degradation in Claude Opus 4.6, with reduced reasoning depth and increased API costs. Critics, including prominent hacker George Hotz, accused Anthropic of overstating risks for marketing purposes, creating a "wolf cry" scenario where hype overshadows reality.

ChatGPT has introduced a major new "Dreaming" memory system, making its AI assistant more human-like by automatically learning and updating user preferences over time. Unlike the older "Saved memories" feature, which required explicit user commands to store information, the new system actively synthesizes useful details from past conversations—such as travel plans, work projects, or personal interests—and intelligently determines which facts remain relevant and which have become outdated. OpenAI reports significant performance improvements: accuracy in recalling facts rose from 41.5% with the old system to 82.8% with Dreaming V3, while adherence to user preferences increased from 31.4% to 71.3%. A key enhancement is the system's ability to keep information current over time, with accuracy jumping from 9.4% to 75.1%. Users can now view and edit a "Memory Summary" page, see sources that influenced responses, and control memory settings. While the feature initially rolls out to US Plus and Pro users, it will later expand globally and to free-tier users. This evolution marks a shift for ChatGPT from a general conversational model toward a persistent, personalized assistant that builds a long-term understanding of each user—raising both practical possibilities and deeper questions about privacy and the nature of AI-human relationships.

The 2026 World Cup quarter-finals are set, with AI models unanimously predicting France, Spain, England, and Argentina to advance to the semi-finals, though they differ on match details. All six AI models (ChatGPT, Claude, Gemini, DeepSeek, Qwen, and Grok) forecast France will defeat Morocco in regulation time, with predicted scores of 2-0 or 2-1. Spain is also favored to beat Belgium, with most AIs predicting a 2-0 or 2-1 win in regular time, though Gemini suggests a possible 1-1 draw leading to a Spanish victory on penalties. The Norway vs. England match is considered the most unpredictable due to Erling Haaland's threat. While all AIs ultimately pick England to advance, predictions range from a 2-1 regulation win to a 2-2 draw requiring extra time. For Argentina vs. Switzerland, AIs predict an Argentine victory, with scores varying from 2-0 to 2-1 in regulation, though Claude and ChatGPT warn the disciplined Swiss could force extra time or a penalty shootout. The core consensus is a final four of France, Spain, England, and Argentina, with the primary disagreements centering on match scores and whether games will require extra time or penalties.

OpenAI announced that its AI model, GPT-5.6 Sol Ultra, has successfully proved the 50-year-old Cycle Double Cover (CDC) conjecture in graph theory in under an hour. This long-standing problem, posed independently by several prominent mathematicians, states that every bridgeless finite undirected graph contains a set of cycles where each edge is covered exactly twice. The breakthrough was achieved using a novel "parallel test-time computation" (TTC) approach. Instead of a single AI working sequentially, the system deployed 64 concurrent AI agents, each exploring distinct proof strategies—from algebraic perspectives to structural induction. The process included strict protocols to avoid common research pitfalls: initial exploration of fundamentally different paths, preventing herd mentality by not revealing the most promising direction, and employing a "critic squad" of agents to rigorously attack and verify every proposed proof step. The system forbade vague assertions, demanding concrete lemmas and constructions. The resulting proof, generated by GPT-5.6 and formatted with Codex, employed a sophisticated multi-step strategy. It first reduced the general case to cubic graphs, then leveraged Tutte's group-flow theorem to establish the existence of a nowhere-zero 8-flow on the graph. A key inventive step was introducing a "two-element set" labeling scheme (Lemma 2.1), which, if satisfied, guarantees a cycle double cover. The AI then transformed this combinatorial condition into a large system of linear equations (Lemma 2.2), using linear algebra over finite fields to conclusively demonstrate that a solution always exists. Researchers highlighted that parallel TTC dramatically compressed the reasoning time, making deep, extended AI problem-solving practically feasible. While some observers marveled at the implications for mathematics and science, others questioned whether parallel breadth can fully substitute for deep, continuous logical chains. Nonetheless, this achievement marks a significant advance in AI's autonomous capacity for high-level abstract reasoning and complex proof generation.