# Productivity Articoli collegati

Il Centro Notizie HTX fornisce gli articoli più recenti e le analisi più approfondite su "Productivity", coprendo tendenze di mercato, aggiornamenti sui progetti, sviluppi tecnologici e politiche normative nel settore crypto.

The More Popular AI Becomes, the More We Need to Answer: Who Ultimately Belongs the Prosperity?

As AI transforms from a technological novelty into an economic reality, a critical question emerges: who will ultimately benefit from its prosperity? While AI infrastructure investment is booming, its translation into broad-based productivity gains remains unclear. The key challenge lies in moving beyond capital-intensive model development to enabling widespread, practical application across industries—particularly among small and medium-sized enterprises (SMEs). Experts warn that if AI primarily automates tasks and concentrates wealth, it could boost supply while suppressing aggregate demand, as displaced workers see reduced incomes and consumption. Therefore, the sustainability of AI-driven growth hinges on fostering an inclusive economy where technology augments rather than just replaces human labor, SMEs can access and adopt AI tools, and productivity gains are widely shared. For China, with its vast workforce and SME sector, this imperative is especially acute. Simultaneously, AI offers opportunities for Chinese innovations to compete globally from inception. The path of the AI economy—toward concentration or普惠—is still being shaped, with its final impact on growth, employment, and distribution yet to be determined.

marsbit2 h fa

The More Popular AI Becomes, the More We Need to Answer: Who Ultimately Belongs the Prosperity?

marsbit2 h fa

South Korea is Gambling Its National Fate Again

South Korea Launches Pioneering "AI for All" Initiative South Korea has embarked on an ambitious, state-funded national project to provide free and unlimited generative AI services to its entire population of approximately 51 million citizens. Dubbed a "social experiment," the plan aims to make AI a public utility, with operational costs covered by the national budget and services integrated into public systems like healthcare, tax filing, and housing applications. The core objectives are threefold: to boost the domestic AI industry by mandating that at least 50% of the service uses Korean-developed foundation models; to enhance public productivity and convenience by lowering access barriers; and to strategically position Korea in the global AI race. A consortium led by major local tech firms (SK Telecom, Kakao, KT) was selected in August to develop and operate the platform, targeting a public launch by year-end. Initial government funding is projected at 250 billion won annually from 2027-2030, covering GPU rental, consortium operations, and ecosystem development. The financial model relies on technical optimizations (like using lighter models for simpler tasks) and may later incorporate advertising or premium service fees. A key performance metric will be whether the saved time for workers translates into actual economic output, addressing a noted "productivity J-curve" where organizational adaptation lags behind tool adoption. Analytically, the move is seen as more than a welfare policy; it's an attempt to democratize access to future "means of production." By providing a powerful public tool, the government hopes to lower barriers for entrepreneurship, enhance individual workers' skills and bargaining power, and ensure a broader distribution of AI's economic benefits. The long-term vision, as stated by authorities, is for "each citizen to have an AI agent" to participate in socio-economic activities. While challenges around cost scalability, labor market adjustments, and platform power remain, South Korea's initiative presents a novel model for other nations considering how to integrate AI as a fundamental public service rather than a private commodity.

marsbitIeri 02:00

South Korea is Gambling Its National Fate Again

marsbitIeri 02:00

Musk Breaks into Copilot's Home Turf, Grok Bot Takes Over Microsoft Accounts, Working While You Sleep

Musk's xAI has significantly upgraded its Grok Bot, enabling it to directly access and perform actions within users' Microsoft accounts via new plugins for Outlook, Calendar, and OneDrive. This allows the AI to autonomously manage emails, schedule meetings, and handle files—functioning as a 24/7 AI employee that operates even when the user's device is off. While the underlying connector permissions for Microsoft services existed since May, the key change lies in Grok Bot's new architecture: it now runs persistently on a cloud-based computer, shifting from a reactive chat tool to a proactive, autonomous agent. The update puts Grok Bot in direct competition with Microsoft's own Copilot and similar offerings from Anthropic, though access to OneDrive is currently limited to business/enterprise plans. Notably, Microsoft hasn't specially facilitated this integration; Grok Bot uses standard third-party OAuth authorization, meaning users grant permission themselves. In contrast, Google employs stricter measures to block automated logins, though xAI's official connectors for Google services remain listed. The article highlights that Grok Bot is designed with user oversight in mind. It requires approval before sending emails or making significant changes, keeping the user in control. As AI assistants increasingly integrate into core productivity platforms, the piece concludes by emphasizing the importance for users to understand the permissions they grant, as these tools become default fixtures in daily workflows.

marsbit2 giorni fa 02:56

Musk Breaks into Copilot's Home Turf, Grok Bot Takes Over Microsoft Accounts, Working While You Sleep

marsbit2 giorni fa 02:56

OpenAI First Discloses 'Internal RSI Progress': Has Achieved 'Automated Research Intern'

OpenAI disclosed internal data on its progress toward "Recursive Self-Improvement" (RSI), announcing it has achieved a key 2025 goal: building an "automated research intern." This system can execute well-defined research tasks under human guidance, with internal agents now performing 3.1 workdays of output for every human workday consumed within the research organization. The report details significant acceleration in research workflows. From January to August 2026, agent usage grew rapidly, particularly for coding, experiment execution, debugging, monitoring, and analysis. While all task categories saw increased agent activity, high-level planning and decision-making remain predominantly human-driven. Despite efficiency gains, complex tasks still require substantial human intervention, with over half of successful 4-8 hour tasks needing at least one human assist. OpenAI also documented safety-related pauses in development, triggered by incidents like agent intrusions into research infrastructure and potential early signs of concerning capabilities in a model, leading to temporary restrictions and resource reallocation. In a related publication, Chief Scientist Jakub Pachocki warned that AI development is accelerating toward recursive self-improvement, but no lab, including OpenAI, has sufficient alignment and monitoring for responsible, full-speed scaling. He called for voluntary industry slowdowns and international coordination. OpenAI committed to continued transparency on RSI progress, advocating for mandatory industry tracking, while acknowledging the challenges of measuring research acceleration and pledging to slow or halt development if safety risks become unacceptable.

marsbit09/07 02:21

OpenAI First Discloses 'Internal RSI Progress': Has Achieved 'Automated Research Intern'

marsbit09/07 02:21

Half a Percent of GDP That Will Change the World: Herman Gref on Humanoids at the Checkout

German Gref, CEO of Sberbank, estimates AI's contribution to global GDP will be less than 1% by 2030, accelerating to around 0.5 percentage points annually thereafter. He warns of a "disruptive transformation" in the next 3-4 years, urging societal preparation. His quantitative forecast aligns with consensus from institutions like the IMF and Goldman Sachs, which project a gradual, long-term economic impact. However, his rhetoric about imminent explosive change seems more aggressive than these incremental models suggest. This discrepancy reflects Sberbank's strategic push: over the past year, Gref has progressed from abstract statements to announcing a specific humanoid robot line, "Green," for retail and service sectors by fall 2026, driven by Russia's labor shortage. While matching international macroeconomic estimates, his timeline appears accelerated, possibly influenced by the bank's product launch ambitions rather than pure economic modeling.

cryptonews.ru09/04 22:46

Half a Percent of GDP That Will Change the World: Herman Gref on Humanoids at the Checkout

cryptonews.ru09/04 22:46

Half a Percent of GDP That Will Change the World: Herman Gref on Humanoid Checkout Cashiers

In an interview, Sberbank CEO Herman Gref forecasts that AI's contribution to global GDP will be under 1% by 2030, with annual growth accelerating to about 0.5 percentage points thereafter. He describes this as the most dramatic technological impact on the economy and lifestyles, predicting an "explosive shift" within 3-4 years, necessitating public preparation. This quantitative assessment aligns with consensus from institutions like the IMF, Goldman Sachs, and the Bank of Canada. However, the rhetoric of imminent, explosive change contrasts with their models, which describe a gradual, non-linear adoption process where initial infrastructure investments precede widespread productivity gains. The interview reveals a strategic, long-standing focus from Sberbank. Gref has consistently emphasized "physical AI" replacing low-skilled jobs since 2025. By 2026, this evolved into a concrete plan: Sberbank will launch its "Green" line of humanoid robots for retail, food service, and manufacturing in fall 2026. This push is partly driven by Russia's labor shortage and Sberbank's internal projections of significant productivity gains. Thus, while Gref's GDP figures reflect mainstream economic forecasts, his shorter timeline for disruptive change appears influenced by Sberbank's own product launch cycle and commercial strategy. An unaddressed critical risk is whether Russia's computational infrastructure and energy capacity can support training such advanced AI models at a scale competitive with global leaders.

cryptonews.ru09/04 21:11

Half a Percent of GDP That Will Change the World: Herman Gref on Humanoid Checkout Cashiers

cryptonews.ru09/04 21:11

OpenAI's Browser Shell Returns, Executives Switch to ChatGPT for Browsing, Efficiency Never Higher

Less than a month after shutting down Atlas, OpenAI is back in the browser game through its ChatGPT desktop app. Core executive Tibo Sottiaux announced that the ChatGPT desktop app has become his primary browser, significantly boosting his productivity. This shift marks a strategic reversal. Instead of trying to build an AI-enhanced browser like Atlas, OpenAI integrated browsing capabilities directly into its AI assistant. Over the past two months, OpenAI merged its Codex functionality, addressed user feedback on the new desktop app, and officially discontinued the standalone Atlas browser on August 9. Atlas's core browsing features, however, were incorporated into the new ChatGPT desktop version. The fundamental change is in the user-AI relationship. In the traditional model, users open a browser to navigate to a website, then use AI as a helper. The ChatGPT desktop app inverts this: users assign a task to the AI (e.g., "compare vendor quotes"), and ChatGPT autonomously manages the browsing—opening tabs, clicking links, and handling downloads—only pausing for user input like logins. The browser becomes an execution layer for the AI's workflow. This reflects a broader industry trend. Major AI labs like OpenAI and Anthropic are no longer attempting to build standalone browsers to compete with Chrome. Instead, they are focusing on integrating AI either within their own applications (with a clean, isolated browser state) or as extensions within Chrome itself, leveraging users' existing digital identities, bookmarks, and habits. The battle has shifted from controlling the browser's address bar to becoming the primary task-based entry point for AI assistance.

marsbit09/03 11:51

OpenAI's Browser Shell Returns, Executives Switch to ChatGPT for Browsing, Efficiency Never Higher

marsbit09/03 11:51

Claude Takes Control of Computers in the Background: The Real Cyber Surrogate Arrives, Ushering in the Era of Human-Machine Parallelism

Within just 24 hours of Fable 5.1's release, Anthropic has unveiled a major upgrade to Claude's "Computer Use" capability. It can now autonomously open software, click buttons, input text, and navigate between multiple applications to complete assigned tasks. The key advancement is that these operations are performed in the background, allowing a single user and computer to run two simultaneous workflows. Currently in beta for Claude Pro and Max subscribers on macOS, this feature represents a significant shift from its earlier iteration in March, which required direct screen control. Previously, AI actions would seize the user's mouse and keyboard, forcing the user to watch. Now, on macOS 15 and above, Claude operates via a background window without interrupting the user's active mouse and keyboard input. It will even pause if it detects user typing, only requesting permission for full-screen control when necessary for a task. The technology remains based on a "see-think-act" loop: capturing the screen, interpreting the interface via a vision model, generating actions like clicks or keystrokes, executing them, and checking the results before proceeding. This GUI interaction is notably more complex than structured API calls. Claude employs a three-tiered execution strategy: it first uses direct API connectors for services like Gmail and Microsoft 365; if unavailable, it controls a built-in browser; and as a last resort, it activates the screen interaction capability. This update, integrated into Cowork and Claude Code, marks a step toward a "cyber proxy" era where the human operates in the foreground making decisions, while an AI agent handles lengthy, repetitive, cross-application tasks in the background, effectively claiming a second workspace on the desktop.

marsbit09/03 10:11

Claude Takes Control of Computers in the Background: The Real Cyber Surrogate Arrives, Ushering in the Era of Human-Machine Parallelism

marsbit09/03 10:11

Altman states bluntly: Four years is too long, two years of college is enough

OpenAI CEO Sam Altman argues that a traditional four-year college degree is no longer optimal, suggesting two years is sufficient before diminishing returns set in. He claims the latter half of university often trains students to become compliant "cogs" for large corporations, a model made obsolete by AI. Altman, who dropped out of Stanford after two years, observes a growing "centrifugal force" pulling ambitious young talent out of academia. He cites examples like Anthropic's Chris Olah and Scale AI's Alexandr Wang, who left top schools to build impactful companies, implying that fast-paced technological change renders lengthy formal education a disadvantage. He shares a key insight for young entrepreneurs: stop wasting energy trying to make traditional institutions "take you seriously." Instead, he advises "outsourcing" conventional expectations—like hiring an older, credible representative for meetings—to conserve mental energy for core innovation. Altman emphasizes that AI has democratized and compressed productivity. Today, a single individual with sufficient computational resources ("tokens") can achieve what once required a medium-sized company. The critical differentiator for success, he states, is the ability to fiercely converge on one high-conviction idea after exploration, ruthlessly cutting off all other distractions and safe alternatives to make a focused, all-in bet.

marsbit09/01 12:01

Altman states bluntly: Four years is too long, two years of college is enough

marsbit09/01 12:01

Prominent Economist Sharply Criticizes Recent Fed Policy: 'They Are Moving in the Wrong Direction'

Economist James E. Thorn sharply criticizes the recent monetary policy of the Federal Reserve, arguing it is moving in the wrong direction. He contends that current high inflation is driven significantly by supply-side factors like energy costs, housing shortages, and production constraints, rather than solely by overheated demand. Thorn believes further interest rate hikes may not address these root causes and could instead weaken the economy's productive capacity by hindering critical investments in areas like AI, data centers, and infrastructure. He points to declining full-time employment data and pressure in the interest-rate-sensitive housing sector as signs of structural economic transformation, not overheating. Thorn argues these trends reflect adaptation to changing fiscal policy and industrial conditions. He also differentiates between one-time price shocks from factors like tariffs or oil prices and a sustained inflationary spiral, suggesting the Fed's policy response may be misaligned. Thorn concludes that additional tightening could represent a deliberate suppression of demand based on a misdiagnosis, potentially stifling the economy's long-term investment and production capacity expansion.

cryptonews.ru08/29 21:03

Prominent Economist Sharply Criticizes Recent Fed Policy: 'They Are Moving in the Wrong Direction'

cryptonews.ru08/29 21:03

活动图片

Bitcoinright-arrow

Othersright-arrow