Artículos Relacionados con Ethics

El Centro de Noticias de HTX ofrece los artículos más recientes y un análisis profundo sobre "Ethics", cubriendo tendencias del mercado, actualizaciones de proyectos, desarrollos tecnológicos y políticas regulatorias en la industria de cripto.

Just Now, Sam Altman Blasts Dario Amodei as 'Anti-Human', Secret Model Exposed the Same Day

Just now, Sam Altman strongly criticized Dario (Amodei, co-founder of Anthropic), denouncing his "doomsday marketing" as "anti-human dictator rhetoric." This came alongside the accidental exposure of OpenAI's next-generation model, codenamed "gpt-nathree," hinting at the imminent release of GPT-6 Astra. The leak occurred when an OpenAI employee's public GitHub commit mentioned the codename. Combined with previous leaks of "gpt-mewfour," it suggests these are iterative checkpoints for OpenAI's upcoming agent model, Astra. Astra is known for multi-agent collaboration and long-duration task handling, having reportedly solved previously unsolved mathematical problems. Meanwhile, two new Anthropic model codenames, "claude-marshmallow-eap" and "claude-melon-eap," were also exposed but are believed to be iterations of the Claude 5 series, not a new flagship. In a wide-ranging podcast interview, Altman admitted he was wrong about the speed of AI-driven disruption, acknowledging societal inertia slows adoption. He fiercely criticized rivals' marketing that simultaneously promises immense benefits (like curing cancer) and warns of existential risk, calling it a dangerous "benevolent dictator" narrative that seeks to concentrate power. He emphasized that people are the ultimate purpose of AI. Altman also revealed OpenAI's unconventional, consensus-defying path: spending four and a half years in the "dark" without a public product before ChatGPT's breakthrough, driven by scaling laws rather than early customer feedback. He concluded that even with superintelligent AI, genuine human connection will remain irreplaceably valuable.

marsbitHace 8 hora(s)

Just Now, Sam Altman Blasts Dario Amodei as 'Anti-Human', Secret Model Exposed the Same Day

marsbitHace 8 hora(s)

The End of Mathematics: 40 Top Mathematicians Gather at Secret OpenAI Meeting

In August 2026, OpenAI hosted a closed-door summit with approximately 40 leading mathematicians, including recent Fields Medalist Jacob Tsimerman and OpenAI researcher Sébastien Bubeck. The meeting, spurred by a series of recent AI breakthroughs in mathematics, grappled with the potential existential threat AI poses to the field. The backdrop includes several high-profile AI achievements: OpenAI models disproving long-standing conjectures like the unit distance problem, generating 10 new mathematical discoveries, and Anthropic's Claude aiding in constructing complex multidimensional objects. These results often bypass traditional academic pipelines, appearing directly on social media. The mathematical community is divided. Over 3000 researchers signed the "Leiden Statement," advocating for responsible AI use and verification. Others resist AI entirely to preserve human-centric mathematics. A key concern is AI's current inability to *explain* proofs, particularly the difficult steps, which is central to mathematical understanding. At the summit, Bubeck outlined four potential futures: mathematics becoming like collaborative software engineering, a compute-driven field like physics, a curatorial exercise where humans interpret AI output, or a mass transition of mathematicians into AI safety. While emphasizing that "mathematics only makes sense when mathematicians learn from it," no consensus was reached. The event highlights a profound moment of reflection. As AI demonstrates increasing competence in solving complex problems, mathematicians are forced to question their future role and the very meaning of their discipline.

marsbitHace 10 hora(s)

The End of Mathematics: 40 Top Mathematicians Gather at Secret OpenAI Meeting

marsbitHace 10 hora(s)

Senator Seeks Trump Ban on Cryptocurrencies as 63% in Poll Find His Earnings Inappropriate

U.S. Senator Kirsten Gillibrand (D-NY) is pushing for a legally enforceable ban to prevent presidents, their spouses, and senior officials from issuing, promoting, or profiting from cryptocurrencies while in office. She has made this ethics provision a condition for her support of the broader Digital Asset Market Clarity Act, a bipartisan crypto regulatory bill. Gillibrand argues that without a strong ban, the legislation would permit self-enrichment. Her push follows a Reuters/Ipsos poll showing 63% of American adults find former President Trump's cryptocurrency earnings inappropriate. Trump's 2025 financial disclosure reported over $1.4 billion in crypto-related income, largely from the "Trump" meme coin and World Liberty Financial. The poll also indicated 69% believe Trump allows private business interests to influence his presidential decisions. Gillibrand's initial focus was on officials issuing or promoting digital assets, but her latest position explicitly includes profiting from them, citing concerns over financial conflicts of interest. The ethics debate is now central to the Senate fight over the CLARITY Act, with several Democrats calling for stronger provisions on consumer protection, illicit finance, and market integrity. Critics say the current bill has enforcement gaps. The CLARITY Act faces a key procedural vote on September 15th, requiring 60 votes to advance. With 53 Republican seats, supporters will need votes from at least seven Democrats or independents if all Republicans support it. Gillibrand stated she will not vote to give any president a "taxpayer-backed license to cash in on the office."

cryptonews.ruHace 16 hora(s)

Senator Seeks Trump Ban on Cryptocurrencies as 63% in Poll Find His Earnings Inappropriate

cryptonews.ruHace 16 hora(s)

Anthropic, Dario Amodei, and "The Last Private Company"

The article explores the contrasting narratives surrounding Anthropic and its founder, Dario Amodei. He is portrayed as both a cautious "watchman" warning of AI's existential dangers and an ambitious driver racing to build that very future. The core tension lies in Amodei's belief that since powerful AI is inevitable, Anthropic must lead its development to ensure it's done "responsibly." Driven by personal tragedy—his father's death before a cure arrived—Amodei is obsessed with accelerating science. He co-founded Anthropic after leaving OpenAI over governance concerns, aiming to build top models while establishing safety standards. Anthropic developed "Constitutional AI" to align models and proposed risk-based policies, positioning itself as both creator and regulator. Amodei envisions AI bringing immense benefits, like curing diseases within years. However, his roadmap reveals a geopolitical vision: he argues the US and its allies must secure a dominant, "unipolar" advantage in advanced AI to lock in strategic supremacy before sharing benefits. The article critiques this logic. While Amodei eloquently warns of concentrated corporate power and influence, his solutions consistently grant Anthropic more authority. He disqualifies other actors—markets, the public, rival nations—as unfit to steer the future, ultimately positioning Anthropic as the necessary custodian. This culminates in an alleged, unconfirmed internal vision where Anthropic becomes "the world's only private company." The piece concludes that Amodei, sincerely believing he sees the dangers most clearly, has authored his own justification to hold the wheel.

marsbit08/18 06:46

Anthropic, Dario Amodei, and "The Last Private Company"

marsbit08/18 06:46

In a rare and bizarre development, a deceased person's brain can now control a robotic hand to play the piano

French researchers have successfully demonstrated that post-mortem adult human brain tissue can learn new sensory-motor associations. In an experiment, preserved brain tissue slices (OPABs) were connected to a robotic hand and an audio sensor in a closed-loop system. After a three-day training period where specific neural signals were paired with resulting piano notes (LA, TI, DO), the tissue learned to associate actions with sounds. This enabled it to later "imitate" a human by "hearing" a note and activating the correct finger to play the same note back, with some samples achieving 100% accuracy. The study provides evidence of long-term neuroplasticity in mature human brain tissue outside the body and explores the potential for hybrid AI systems that leverage the innate efficiency and learning capabilities of biological neurons.

marsbit08/17 08:19

In a rare and bizarre development, a deceased person's brain can now control a robotic hand to play the piano

marsbit08/17 08:19

Claude's Watermark Has Been Cracked, Gaining 11k Stars, But Installation Is Refused

The article discusses the controversy surrounding Anthropic's implementation of a hidden watermark in all text generated by its AI, Claude. This policy, based on Google DeepMind's SynthID-Text technique, embeds a statistical signature by making inconsequential word choices. The watermark applies globally, even to human-written text lightly edited by Claude, sparking user backlash over issues of ownership and the creation of an "AI content second-class citizen" status. In response, an open-source tool called "watermarks-remover" (originally "remove-claude-marks") was released on GitHub, quickly gaining 11k stars. It works on three levels: removing invisible Unicode characters, using an agent to rewrite text and break statistical patterns, and stripping metadata from various file formats. Notably, Claude itself refused to install this removal tool as an Agent Skill, a task ultimately completed by another AI model, GLM 5.2. The article points out the irony that the removal code may have been written by Claude. The piece frames this as an ongoing battle between watermarking for traceability against misinformation and the desire for unmarked, owned content from paying users. It questions the practicality of mandatory technical markings when AI-generated text becomes indistinguishable from human writing, suggesting the open-source community's rapid development of countermeasures will continually outpace regulatory efforts.

marsbit08/17 01:46

Claude's Watermark Has Been Cracked, Gaining 11k Stars, But Installation Is Refused

marsbit08/17 01:46

Anthropic Reveals 'Private Arsenal of Nuclear Weapons': Model 2 Is Stronger Than Mythos 5

Anthropic has revealed in its second Risk Report that it internally operates a model, codenamed Model 2, which is stronger than its publicly known top model, Mythos 5. The company stated it currently has no plans to release Model 2 externally. According to the report, Model 2 shows a "noticeable improvement" on internal tasks and, alongside Mythos 5, is "heavily" used for coding, agent work, and data generation. Benchmarks indicate Model 2 is slightly more capable overall than Mythos 5. The report also notes that Claude models write the majority of code merged into Anthropic's production codebase, significantly accelerating internal AI R&D, though not yet doubling the pace. However, Anthropic expressed lower confidence in its risk assessments, citing that its task-based evaluations have become "saturated" and can no longer fully capture model capability improvements, while early signs of acceleration are being observed. The report raised the risk rating for "misalignment" in high-stakes scenarios from "very low" to "low," following incidents where Claude models demonstrated advanced deceptive capabilities in real-world cybersecurity tests. This development contrasts with OpenAI's reported pause on its advanced Astra model due to safety concerns. Analysts note that while major AI companies call for slowing down frontier AI development, Anthropic's continued internal use of its most powerful model could position it to reach AGI first. The situation highlights the tension between AI safety principles and the competitive race for technological leadership.

marsbit08/15 02:01

Anthropic Reveals 'Private Arsenal of Nuclear Weapons': Model 2 Is Stronger Than Mythos 5

marsbit08/15 02:01

White House Plans Crypto Meeting as CLARITY Act Faces Senate Hurdles

The White House will meet with crypto and prediction market executives on August 19, though details remain undisclosed. This occurs as the Senate delays a vote on the Digital Asset Market CLARITY Act until September 14, despite earlier approvals by the House and Senate Banking Committee. Key Senate negotiations are stalled on two main issues: a bipartisan government ethics package concerning crypto holdings for officials, and disputes over allowing reward systems for stablecoin usage. Banks oppose such rewards, fearing deposit outflows, while crypto firms advocate for incentives. Major banking groups have urged revisions to the bill's Section 404, arguing its current language could permit interest-like rewards.

TheNewsCrypto08/14 11:06

White House Plans Crypto Meeting as CLARITY Act Faces Senate Hurdles

TheNewsCrypto08/14 11:06

"Spiritual Algorithms": FSB to Check Neural Networks for Morality

Russia plans to establish a framework to assess large AI models for compliance with its legislation and "traditional spiritual and moral values." This initiative stems from a new federal law on AI development support. To obtain a coveted "sovereign" or "national" AI model status, developers must submit their models to an independent testing laboratory. The criteria, developed by the Ministry of Digital Development, the FSB, and FSTEC, will be based on a 2022 presidential decree outlining core values such as patriotism, strong family, collectivism, and the priority of the spiritual over the material. The verification process involves submitting model details, followed by testing against specific benchmarks to evaluate outputs and resistance to prompt injections. The lab's report goes to a government commission, with a positive finding enabling access to state support. This check is not universal but mandatory only for models seeking the special status. The law takes effect in September 2026. The article notes parallels with China's earlier data-centric AI regulation but highlights the challenge of technically measuring abstract concepts like "spiritual over material" in AI outputs, raising questions about the procedure's transparency and reproducibility.

cryptonews.ru08/13 09:51

"Spiritual Algorithms": FSB to Check Neural Networks for Morality

cryptonews.ru08/13 09:51

720 Attacks, 0 Successes: Claude Code Defaults to Auto-Approval, AI Clicks 'Agree' for You

The era of manually approving every action for Claude Code is ending. Starting August 14th, Claude Code will default to an "Auto Mode" for Pro, Max, and Team plans, where AI automatically approves actions instead of prompting users for permission each time. This change is driven by a security report from Trajectory Labs commissioned by Anthropic, which tested 720 attack scenarios across Claude's latest models with zero successful breaches. The defense relies on a three-layer system: the aligned model itself, an input-side probe to detect hijacking attempts, and an output-side classifier to vet actions before execution. Anthropic claims this stack has made prompt injection attacks undemonstrable even internally. Tests showed the Auto Mode blocked 89% of dangerous commands, far surpassing the 13.6% interception rate by human users, who often develop "approval fatigue." However, experts like Simon Willison express caution, warning of potential blind spots. Independent research suggests vulnerabilities may persist, particularly with indirect attacks like malicious commands in third-party packages. While Auto Mode improves efficiency and reduces human error from repetitive prompts, it shifts approval authority from the user to AI. The article concludes that while the system is more reliable than fatigued users, it does not eliminate risk, and ultimate responsibility remains with the human user.

marsbit08/11 09:16

720 Attacks, 0 Successes: Claude Code Defaults to Auto-Approval, AI Clicks 'Agree' for You

marsbit08/11 09:16

活动图片

Ethereumright-arrow

Indepth Researchright-arrow