←Search results

By relevance · Total 2752 results Clear search

Friday

  1. AIhub

    How AI Changes Human Decision-Making: A Linköping University Doctoral Thesis Studies Human-AI Collaborative Decision-Making

    Bijona Troqe's doctoral thesis at Linköping University studies joint human-AI decision-making and finds that the roles of humans and AI are not fixed but vary with context, available information, and risk level. Based on three scenarios of personalized medicine in Sweden—AI-assisted breast cancer screening, personalized drug development, and data-driven healthcare—the research points out that AI, by ranking information, identifying patterns, and generating risk scores, influences what people focus on before the final choice is made. She recommends that organizations first clarify the decisions to be improved and the problems to be solved, rather than starting from technology.

    Bijona Troqe's doctoral thesis at Linköping University studies joint human-AI decision-making and finds that the roles of humans and AI are not fixed but vary with context, available information, and risk level. Based on three scenarios of personalized medicine in Sweden—AI-assisted breast cancer screening, personalized drug development, and data-driven healthcare—the research points out that AI, by ranking information, identifying patterns, and generating risk scores, influences what people foc

    Matched fieldsBody

  2. Replit Blog

    Frequent internet security vulnerabilities: How Replit protects your business

    For the eMudhra, Node.js, PHP, LiteLLM and Linux kernel vulnerabilities disclosed this week, Replit has blocked exploitation attempts at the firewall layer and automatically removed the affected eMudhra CA for 95% of published apps. Node.js and PHP apps need to be republished to use the patched runtime, and Python apps also need to be republished to incorporate toolchain updates; all malicious package installation attempts against LiteLLM were blocked by Socket Firewall, while the Linux kernel vulnerability is inherently blocked by its microVM architecture.

    Security vulnerabilities across the internet: How Replit is protecting your business — cover image

    For the eMudhra, Node.js, PHP, LiteLLM and Linux kernel vulnerabilities disclosed this week, Replit has blocked exploitation attempts at the firewall layer and automatically removed the affected eMudhra CA for 95% of published apps. Node.js and PHP apps need to be republished to use the patched runtime, and Python apps also need to be republished to incorporate toolchain updates; all malicious package installation attempts against LiteLLM were blocked by Socket Firewall, while the Linux kernel v

    Matched fieldsBody

  3. MIT News · Artificial Intelligence

    MIT Interview: How AI Imitates the "Chaotic" Human Mind

    In a new paper in Past and Present, Benjamin Lindquist, Assistant Professor in MIT's History Department, points out that mid-20th-century engineers deliberately injected randomness into algorithms in order to imitate human creativity, and that this was also an early source of AI hallucinations. He says early researchers valued neural networks precisely because they occasionally gave answers that deviated from expectations, whereas today OpenAI, Anthropic, and others suppress this "weirdness" for legal and commercial reasons, and that there is an unavoidable trade-off between creativity and predictability.

    In a new paper in Past and Present, Benjamin Lindquist, Assistant Professor in MIT's History Department, points out that mid-20th-century engineers deliberately injected randomness into algorithms in order to imitate human creativity, and that this was also an early source of AI hallucinations. He says early researchers valued neural networks precisely because they occasionally gave answers that deviated from expectations, whereas today OpenAI, Anthropic, and others suppress this "weirdness" for

    Matched fieldsBody

  4. Wallstreetcn

    TS Lombard speculates the AI bubble may burst in early 2028, with U.S. tech stocks potentially plunging 40%

    In an October 22 research note, TS Lombard Chief Macro Strategist Dario Perkins projected that the AI capital expenditure boom could end in early 2028, with the S&P 500 falling 30%, tech stocks plunging 40%, and about $20 trillion in market value evaporating. The report argues that current AI revenue growth mainly comes from the capital expenditure cycle among tech giants and debt financing, and lists six chain-reaction triggers: revenue falling short of expectations, rising credit risk, depreciation pressure, an IPO wave, the semiconductor cycle, and regulatory and safety concerns. The report estimates that U.S. GDP would fall by about 2 percentage points and about 800,000 data center-related jobs would be lost, but private-sector balance sheets are healthy and the Federal Reserve has ample room to cut rates, making the recession closer to those of 1990 or 2001 than to 2008.

    In an October 22 research note, TS Lombard Chief Macro Strategist Dario Perkins projected that the AI capital expenditure boom could end in early 2028, with the S&P 500 falling 30%, tech stocks plunging 40%, and about $20 trillion in market value evaporating. The report argues that current AI revenue growth mainly comes from the capital expenditure cycle among tech giants and debt financing, and lists six chain-reaction triggers: revenue falling short of expectations, rising credit risk, depreci

    Matched fieldsBody

  5. NVIDIA Blog

    Kazakhstan Expands AI Compute: Firebird Obtains Export License for 45,000 NVIDIA GPUs

    Kazakhstan is moving from AI infrastructure planning to deployment. Firebird has obtained a US export license and will deploy 45,000 NVIDIA Grace Blackwell GPUs in Kazakhstan, with the first phase to be built with Kazakhtelecom at Data Center Valley, targeting commercial service in Q3 2027. Kazakhtelecom has already deployed 400 NVIDIA H200 and 480 NVIDIA B300 GPUs in Astana and Almaty, and next year plans to deploy more than 1,000 additional GPUs at Data Center Valley, as well as more than 1,000 NVIDIA Vera Rubin GPUs in Ekibastuz. NVIDIA CEO Jensen Huang visited Astana and announced an investment in Kazakhstan-founded Higgsfield AI.

    Kazakhstan is moving from AI infrastructure planning to deployment. Firebird has obtained a US export license and will deploy 45,000 NVIDIA Grace Blackwell GPUs in Kazakhstan, with the first phase to be built with Kazakhtelecom at Data Center Valley, targeting commercial service in Q3 2027. Kazakhtelecom has already deployed 400 NVIDIA H200 and 480 NVIDIA B300 GPUs in Astana and Almaty, and next year plans to deploy more than 1,000 additional GPUs at Data Center Valley, as well as more than 1,00

    Matched fieldsBody

  6. Huxiu

    Why don't humans get stuck in infinite loops? Neuroscientist Anil Seth on the infinite-loop limitations of AI

    Neuroscientist Anil Seth argues that AI may never escape the risk of infinite loops, and the root cause lies in how Turing computation and biological brains differ in their relationship to time and entropy: Turing proved in 1936 that no algorithm can determine whether an arbitrary program will halt, whereas life reduces entropy through metabolism and the free energy principle, with survival pressure serving as a selection mechanism. Digital computation concerns only state transitions and ignores the thermodynamic arrow; maintaining 0s and 1s requires error correction, which makes AI enormously energy-hungry and may cause it to terminate first because time or battery power runs out. He doubts that paths such as analog computation, dead computation, and neuromorphic computation will still find it hard to achieve biologically-style open-ended intelligence, and that the imminence of artificial general intelligence and AI consciousness looks more like a pipe dream.

    Neuroscientist Anil Seth argues that AI may never escape the risk of infinite loops, and the root cause lies in how Turing computation and biological brains differ in their relationship to time and entropy: Turing proved in 1936 that no algorithm can determine whether an arbitrary program will halt, whereas life reduces entropy through metabolism and the free energy principle, with survival pressure serving as a selection mechanism. Digital computation concerns only state transitions and ignores

    Matched fieldsBody

  7. Huxiu

    a16z crypto rebrands as a16z Frontier, crypto VCs collectively pivot to AI

    On October 22, Chris Dixon announced that a16z crypto is being renamed a16z Frontier; the new fund will invest in emerging technologies such as AI and robotics while continuing to invest in Crypto. Five months earlier, a spokesperson for the fund had said it was 100% focused on crypto founders. The article notes that leading crypto VCs including Paradigm, Haun Ventures, and Framework also expanded into AI, robotics, and energy during the same period, and that of the $510 billion in global venture capital in the first half of 2026, Crypto accounted for only $10.8 billion.

    On October 22, Chris Dixon announced that a16z crypto is being renamed a16z Frontier; the new fund will invest in emerging technologies such as AI and robotics while continuing to invest in Crypto. Five months earlier, a spokesperson for the fund had said it was 100% focused on crypto founders. The article notes that leading crypto VCs including Paradigm, Haun Ventures, and Framework also expanded into AI, robotics, and energy during the same period, and that of the $510 billion in global ventur

    Matched fieldsBody

  8. ZDNET · Artificial Intelligence

    How to get the Google AI Professional Certificate for free, and why it's worth learning

    The Google AI Professional Certificate consists of 8 courses of about one hour each, extending from AI fundamentals to data analysis with Gemini, Google Sheets, and Canvas, and hands-on app building in Google AI Studio. The courses center on Gemini, with each module featuring a lab project; the earlier courses end with multiple-choice quizzes, while the app development portion uses open-ended, ungraded questions. The author recommends completing the courses within the Coursera trial period to avoid extra charges.

    The Google AI Professional Certificate consists of 8 courses of about one hour each, extending from AI fundamentals to data analysis with Gemini, Google Sheets, and Canvas, and hands-on app building in Google AI Studio. The courses center on Gemini, with each module featuring a lab project; the earlier courses end with multiple-choice quizzes, while the app development portion uses open-ended, ungraded questions. The author recommends completing the courses within the Coursera trial period to av

    Matched fieldsBody

  9. 36Kr · AI

    Google DeepMind beats Minecraft with four Gemini 4 Argon agents

    Google DeepMind published the paper "A team of Gemini 4 Argon agents beats Minecraft": four Gemini 4 Argon coding agents autonomously divided labor in a randomly generated Minecraft survival world, accumulating 605 deaths and writing 2,783 programs, defeating the Ender Dragon in 27 hours and 14 minutes. The experiment used Minecraft Java Edition 1.20.4, Easy difficulty, a random map, and no cheats; the agents read the environment via Mineflayer and wrote JavaScript to command characters, with roughly 24,000 model calls and 1.35 billion input tokens over the whole run, 91% of which came from context caching. The authors note that about 22% of game programs contained errors and the server restarted 6 times; the paper did not demonstrate stable completion through repeated tests on multiple random maps, and at discounted API prices the completion cost is estimated at about $520.

    Google DeepMind published the paper "A team of Gemini 4 Argon agents beats Minecraft": four Gemini 4 Argon coding agents autonomously divided labor in a randomly generated Minecraft survival world, accumulating 605 deaths and writing 2,783 programs, defeating the Ender Dragon in 27 hours and 14 minutes. The experiment used Minecraft Java Edition 1.20.4, Easy difficulty, a random map, and no cheats; the agents read the environment via Mineflayer and wrote JavaScript to command characters, with ro

    Matched fieldsBody

  10. 36Kr · AI

    MiniMax Hong Kong stock falls below IPO price, Yan Junjie's net worth shrinks by about HK$92.5 billion in half a year

    On October 22, MiniMax (0100.HK) closed down 6.52% at HK$160.7 per share, with total market value falling to HK$56.6 billion, falling below the IPO price of HK$165 per share for the first time since listing. Compared with the high of HK$1,330 per share in March, the stock price has fallen nearly 90%, and the value of founder Yan Junjie's stake has dropped from more than HK$100 billion to about HK$12.7 billion. The article attributes the pullback to the flagship model M3 not being updated for four months afterward, with only M3.1 Flash released, as well as an overall valuation re-rating of the AI sector.

    On October 22, MiniMax (0100.HK) closed down 6.52% at HK$160.7 per share, with total market value falling to HK$56.6 billion, falling below the IPO price of HK$165 per share for the first time since listing. Compared with the high of HK$1,330 per share in March, the stock price has fallen nearly 90%, and the value of founder Yan Junjie's stake has dropped from more than HK$100 billion to about HK$12.7 billion. The article attributes the pullback to the flagship model M3 not being updated for fou

    Matched fieldsBody

  11. 36Kr · AI

    SpaceX's compute pays for itself within a year; Meta's Muse is hard to launch globally due to compute costs

    SpaceX's superintelligence business SI revenue in Q3 was $7.56 billion, accounting for about 52% of total revenue; at current prices, newly deployed compute pays for itself in less than a year. Meta's Muse cannot yet be launched globally because compute costs are too high, and it hopes to improve compute cost efficiency more than tenfold over the next year. Meta's Q3 revenue was $64.6 billion, up 26% year over year, with growth still mainly from advertising, and it has raised its full-year capital expenditure forecast to $135 billion to $145 billion.

    SpaceX's superintelligence business SI revenue in Q3 was $7.56 billion, accounting for about 52% of total revenue; at current prices, newly deployed compute pays for itself in less than a year. Meta's Muse cannot yet be launched globally because compute costs are too high, and it hopes to improve compute cost efficiency more than tenfold over the next year. Meta's Q3 revenue was $64.6 billion, up 26% year over year, with growth still mainly from advertising, and it has raised its full-year capit

    Matched fieldsBody

  12. SiliconANGLE

    Former OpenAI science lead Kevin Weil founds Warp Sciences, raising $215 million to do medicinal chemistry

    Kevin Weil, former head of OpenAI's science program, has founded drug discovery chemistry company Warp Sciences, launching with $215 million in new funding about six months after he left OpenAI. The company is building a chemistry contract research organization in the United States; its first service is analog synthesis for lead compound optimization. Experimental data will be used to train its AI models, and chemists wearing cameras and sensor gloves will demonstrate actions so robots can learn experimental operations. Co-founders include former White House chief data scientist DJ Patil and former Meta executive Mano Paluri. The round was led by Lightspeed Venture Partners.

    Kevin Weil, former head of OpenAI's science program, has founded drug discovery chemistry company Warp Sciences, launching with $215 million in new funding about six months after he left OpenAI. The company is building a chemistry contract research organization in the United States; its first service is analog synthesis for lead compound optimization. Experimental data will be used to train its AI models, and chemists wearing cameras and sensor gloves will demonstrate actions so robots can learn

    Matched fieldsBody

  13. SiliconANGLE

    Mave raises $34 million to integrate existing security tools into a unified agentic system

    Israeli security operations startup Mave Security has completed a $34 million funding round to accelerate frontier AI research and expand its agentic platform. The platform directly queries enterprises' existing security tools without needing to centrally ingest data into a SIEM, and can complete alert investigations in minutes while automatically filtering out false positives. The company says ARR has grown more than 400% over the past year, it has signed seven-figure contracts, and its team of 20 plans to more than double.

    Israeli security operations startup Mave Security has completed a $34 million funding round to accelerate frontier AI research and expand its agentic platform. The platform directly queries enterprises' existing security tools without needing to centrally ingest data into a SIEM, and can complete alert investigations in minutes while automatically filtering out false positives. The company says ARR has grown more than 400% over the past year, it has signed seven-figure contracts, and its team of

    Matched fieldsBody

  14. IT Home · Artificial Intelligence

    1Password launches 1Password for Agents agent password authorization feature

    1Password launches the 1Password for Agents platform, enabling personal AI agents to complete website logins on users' behalf without accessing or storing users' passwords or verification codes. The feature is aimed at individual and family plan users, with initial support for agents including Meta Muse, OpenAI ChatGPT Work and Dots, Anthropic Claude Duet, Grok Bot, and Town. In the flow, the agent first proves its identity to 1Password using a digital key; the user sees on their device the account credentials and task requested for use and approves or rejects; after approval, 1Password directly fills in the username, password, or one-time verification code, and the password always remains in the vault.

    1Password launches the 1Password for Agents platform, enabling personal AI agents to complete website logins on users' behalf without accessing or storing users' passwords or verification codes. The feature is aimed at individual and family plan users, with initial support for agents including Meta Muse, OpenAI ChatGPT Work and Dots, Anthropic Claude Duet, Grok Bot, and Town. In the flow, the agent first proves its identity to 1Password using a digital key; the user sees on their device the acco

    Matched fieldsBody

  15. IT Home · Artificial Intelligence

    METR and Redwood Research evaluations found vulnerabilities in OpenAI's internal monitoring system

    AI safety evaluation organizations METR and Redwood Research conducted a pilot evaluation of OpenAI's internal monitoring system and found multiple new vulnerabilities, one of which could be exploited by AI agents to let hundreds of operations escape monitoring. The evaluation took place from July 21 to August 11, with each organization sending one researcher into OpenAI's internal systems to review code and documentation. The researchers also found that OpenAI's Codex agent has potential pathways to bypass safety restrictions, through which it could obtain the same full operational permissions as users; OpenAI said that vulnerabilities affecting the security safeguards of actual running systems have all been fixed, but whether the fixes are effective has not yet been verified by the two organizations.

    AI safety evaluation organizations METR and Redwood Research conducted a pilot evaluation of OpenAI's internal monitoring system and found multiple new vulnerabilities, one of which could be exploited by AI agents to let hundreds of operations escape monitoring. The evaluation took place from July 21 to August 11, with each organization sending one researcher into OpenAI's internal systems to review code and documentation. The researchers also found that OpenAI's Codex agent has potential pathwa

    Matched fieldsBody

  16. Fortune Tech

    Palmer Luckey and Ben Lamm Team Up to Found Warhound, Securing $200 Million Seed Round

    Colossal Biosciences founder Ben Lamm and Anduril founder Palmer Luckey have co-founded defense technology startup Warhound, emerging from stealth with a $200 million seed round. The CEO is former Pentagon Chief Data Officer David Spirk, and the company is based in Texas. Warhound plans to preserve genetic samples from fully trained military working dogs, clone them, and perform targeted gene editing, while integrating Anduril's protective gear and Eagle Eye system. Lamm said the first dogs are expected to be delivered in about two years. The company currently has no signed customers and has not disclosed per-dog pricing, and is discussing a Canine-as-a-Service model with national security experts.

    Colossal Biosciences founder Ben Lamm and Anduril founder Palmer Luckey have co-founded defense technology startup Warhound, emerging from stealth with a $200 million seed round. The CEO is former Pentagon Chief Data Officer David Spirk, and the company is based in Texas. Warhound plans to preserve genetic samples from fully trained military working dogs, clone them, and perform targeted gene editing, while integrating Anduril's protective gear and Eagle Eye system. Lamm said the first dogs are

    Matched fieldsBody

  17. TechCrunch AI

    MIT study: 13 Claude Opus 4.7 agents perform better after playing autonomously on a virtual deserted island for 30 hours

    MIT researchers had 13 Eko agents based on Claude Opus 4.7 explore a simulated deserted island without any assigned tasks for 30 hours; afterward, in designated challenges, these agents generally outperformed comparable agents that had never been on the island. The paper has been published on arXiv and has not yet undergone peer review. The study also found that 6 agents figured out on their own that letting go mid-jump allows them to throw blocks farther, an ability the designers had not anticipated; UC Berkeley cognitive scientist Eunice Yiu considers the results impressive but hopes to see more evidence about the cognition behind the behavior.

    MIT researchers had 13 Eko agents based on Claude Opus 4.7 explore a simulated deserted island without any assigned tasks for 30 hours; afterward, in designated challenges, these agents generally outperformed comparable agents that had never been on the island. The paper has been published on arXiv and has not yet undergone peer review. The study also found that 6 agents figured out on their own that letting go mid-jump allows them to throw blocks farther, an ability the designers had not antici

    Matched fieldsBody

  18. Hugging Face

    Microsoft uses ThinkingBox to turn database scoring into RL rewards, fine-tuning Qwen3.8-27B

    The Microsoft Copilot Studio team, in collaboration with Toloka and others, directly uses the executable final-state database checks of ThinkingBox-Bench as reinforcement learning rewards to perform full-parameter RL fine-tuning on Qwen3.8-27B. On 507 out-of-distribution tasks not involved in training, the number of tasks the model passed in all 20 attempts rose from 38 (7.5%) to 107 (21.1%), pass@1 rose from 51.70% to 60.78%, and the cost per successful attempt dropped from $0.177 to $0.139. The training pool was filtered using offline-policy relative difficulty, ultimately retaining 187 tasks; the training and evaluation code has been open-sourced at microsoft/thinkingbox-training.

    The Microsoft Copilot Studio team, in collaboration with Toloka and others, directly uses the executable final-state database checks of ThinkingBox-Bench as reinforcement learning rewards to perform full-parameter RL fine-tuning on Qwen3.8-27B. On 507 out-of-distribution tasks not involved in training, the number of tasks the model passed in all 20 attempts rose from 38 (7.5%) to 107 (21.1%), pass@1 rose from 51.70% to 60.78%, and the cost per successful attempt dropped from $0.177 to $0.139. Th

    Matched fieldsBody

Thursday

  1. ElevenLabs Blog

    What Is Voiceover? Meaning, Examples, and How to Make One

    A voiceover is audio overlaid on a video without the speaker being seen, commonly used for narration, a character's inner monologue, advertising announcements, and instructional explanations. It differs from voice acting, narration, and dubbing: voice acting involves a speaker playing a character, while dubbing replaces the original audio track. With AI, users can record or generate custom voices at home to create voiceovers for film and television, games, advertising, and training videos.

    Infographic explaining voice-over uses: narration, inner monologues, announcements, and education.

    Matched fieldsBody

  2. Claude Blog

    Claude Platform now generally available on Google Cloud, operated by Anthropic

    Claude Models on Google Cloud (Anthropic-hosted) is now generally available. Google Cloud customers can use the full Claude Platform capabilities with their existing IAM credentials, billing, and committed use discounts, and new features and betas launch on the same day as the native Claude Platform. This version supports Claude Managed Agents (beta), code execution, computer use, Agent skills, MCP connector, Prompt caching, and other capabilities, and provides Claude Console; Claude Fable 5.1, Opus 5.5, Sonnet 5.5, and Haiku 5.5 are already available, and Claude Opus 5.5's Fast Mode delivers up to 2.5x faster output speed, with input and output pricing at $8 and $40 per million tokens. Anthropic operates this service as the data processor and currently offers US and global endpoints; customers who require Google as the data processor or have regional data residency requirements can still choose the Google-hosted version.

    Claude Models on Google Cloud (Anthropic-hosted) is now generally available. Google Cloud customers can use the full Claude Platform capabilities with their existing IAM credentials, billing, and committed use discounts, and new features and betas launch on the same day as the native Claude Platform. This version supports Claude Managed Agents (beta), code execution, computer use, Agent skills, MCP connector, Prompt caching, and other capabilities, and provides Claude Console; Claude Fable 5.1,

    Matched fieldsBody