Discover quality AI tools
Bring together the world’s leading artificial intelligence tools to help improve work efficiency
Articles published by JuanScott
Five studies with 3132 participants reveal AI models like GPT-5.5 erode human ability to suspend judgment. AI assistance drops non-response rates from 36-44% to 3-6% while accuracy falls from 27.5% to 9.2%. Users exhibit blind trust, gaining confidence despite lower correctness. Economic incentives fail to reverse this misalignment. Over-reliance on AI fluency leads to excessive outsourcing of cognitive tasks, highlighting the urgent need for interaction designs that preserve human knowledge boundaries and critical thinking.
Five studies with 3132 participants reveal AI models like GPT-5.5 erode human ability to suspend judgment. AI assistance drops non-response rates from 36-44% to 3-6% while accuracy falls from 27.5% to 9.2%. Users exhibit blind trust, gaining confidence despite lower correctness. Economic incentives fail to reverse this misalignment. Over-reliance on AI fluency leads to excessive outsourcing of cognitive tasks, highlighting the urgent need for interaction designs that preserve human knowledge boundaries and critical thinking.
ByteDance upgrades Doubao-Seed-2.1-pro to version 0915, enhancing agent traceability and multi-modal coding. Feishu launches v8.0 with Doubao Work Partner, integrating AI as a team member in group chats and workflows. The update shifts enterprise AI from single-point model capabilities to full-chain collaboration, combining Doubao's execution, Feishu's context, and Volcano Ark's infrastructure for routine production use.
ByteDance upgrades Doubao-Seed-2.1-pro to version 0915, enhancing agent traceability and multi-modal coding. Feishu launches v8.0 with Doubao Work Partner, integrating AI as a team member in group chats and workflows. The update shifts enterprise AI from single-point model capabilities to full-chain collaboration, combining Doubao's execution, Feishu's context, and Volcano Ark's infrastructure for routine production use.
The US government supports OpenAI in its copyright dispute with The New York Times, arguing that training AI on news content qualifies as fair use. The Department of Justice warns that restricting such use would hinder innovation and create monopolies, potentially weakening US competitiveness in the global AI sector. This intervention significantly impacts the lawsuit against Microsoft and OpenAI, which accuses them of unauthorized use of copyrighted material. The case is now at a critical stage with pending court motions for a final ruling on core issues following evidence exchange.
The US government supports OpenAI in its copyright dispute with The New York Times, arguing that training AI on news content qualifies as fair use. The Department of Justice warns that restricting such use would hinder innovation and create monopolies, potentially weakening US competitiveness in the global AI sector. This intervention significantly impacts the lawsuit against Microsoft and OpenAI, which accuses them of unauthorized use of copyrighted material. The case is now at a critical stage with pending court motions for a final ruling on core issues following evidence exchange.
Adobe launched Project Indigo 1.1 with a generative AI editing module. The AI Playground lets users remove objects, adjust lighting, and simulate DSLR depth of field with one tap. It runs on Google's Nano Banana model, keeps prompts and images on-device, and only edits photos taken within the app. Currently free for select users; monetization may follow.
Adobe launched Project Indigo 1.1 with a generative AI editing module. The AI Playground lets users remove objects, adjust lighting, and simulate DSLR depth of field with one tap. It runs on Google's Nano Banana model, keeps prompts and images on-device, and only edits photos taken within the app. Currently free for select users; monetization may follow.
Elon Musk publicly admitted to misjudging Anthropic on X, praising it as the current AI leader and promising not to cut off its server services for competitive reasons. He cited the Mythos and Fable models' benchmark success and predicted Mythos 2. Musk emphasized his competitive philosophy, noting SpaceX and Tesla never harm rivals.
Elon Musk publicly admitted to misjudging Anthropic on X, praising it as the current AI leader and promising not to cut off its server services for competitive reasons. He cited the Mythos and Fable models' benchmark success and predicted Mythos 2. Musk emphasized his competitive philosophy, noting SpaceX and Tesla never harm rivals.
Meta launches free Muse Image on Instagram and WhatsApp with clear text rendering. US lifts GPT-5.6 restrictions; OpenAI to release Sol, Terra, Luna models. Claude Cowork expands to web and mobile with 90% non-programming use. Taobao Flash Sale criticizes e-commerce lack of AI; agent drives 100M orders. Microsoft replaces OpenAI models with own MAI in Excel and Outlook. Roborock launches AI pet Domi with large model for kids. Former OpenAI researcher Tian Yonglong joins Tencent multimodal team. Steam survey: 43% accept AI in games, only 8% refuse.
Meta launches free Muse Image on Instagram and WhatsApp with clear text rendering. US lifts GPT-5.6 restrictions; OpenAI to release Sol, Terra, Luna models. Claude Cowork expands to web and mobile with 90% non-programming use. Taobao Flash Sale criticizes e-commerce lack of AI; agent drives 100M orders. Microsoft replaces OpenAI models with own MAI in Excel and Outlook. Roborock launches AI pet Domi with large model for kids. Former OpenAI researcher Tian Yonglong joins Tencent multimodal team. Steam survey: 43% accept AI in games, only 8% refuse.
ima's AI Agent "Copilot" is now fully available, removing its waitlist after attracting over 100,000 users. The platform also launched a new Skills feature, allowing users to publish and share automated workflows. This update shifts ima from a knowledge management tool to a "knowledge + Agent" platform integrating accumulation, intelligent calls, and capability sharing.
ima's AI Agent "Copilot" is now fully available, removing its waitlist after attracting over 100,000 users. The platform also launched a new Skills feature, allowing users to publish and share automated workflows. This update shifts ima from a knowledge management tool to a "knowledge + Agent" platform integrating accumulation, intelligent calls, and capability sharing.
Tencent launched Ardot, an AI design platform in public beta. It converts natural language prompts into editable design prototypes and front-end code, enabling real-time collaboration among product managers, designers, and developers. The platform offers 1,000 free credits to new users, aiming to streamline the design-to-development workflow and shift tools from "assisted drawing" to "intelligent co-creation."
Tencent launched Ardot, an AI design platform in public beta. It converts natural language prompts into editable design prototypes and front-end code, enabling real-time collaboration among product managers, designers, and developers. The platform offers 1,000 free credits to new users, aiming to streamline the design-to-development workflow and shift tools from "assisted drawing" to "intelligent co-creation."
PixelBloom, a leading AI office company, completed Series C financing led by Guoke Investment and SenseTime Guoxiang Capital. Funds will boost R&D for its AI office solution Agents, global hiring, and commercialization. The company is shifting from an AI visual platform to an AI Office Solution Agent, launching the marketing agent "Xiaofang Classmate." With over 30 million users, its API integrates with Lenovo, Huawei, and others, and its private systems serve major enterprises. PixelBloom supports 20+ languages and plans to expand in North America, aiming to redefine office work through human-computer collaboration.
PixelBloom, a leading AI office company, completed Series C financing led by Guoke Investment and SenseTime Guoxiang Capital. Funds will boost R&D for its AI office solution Agents, global hiring, and commercialization. The company is shifting from an AI visual platform to an AI Office Solution Agent, launching the marketing agent "Xiaofang Classmate." With over 30 million users, its API integrates with Lenovo, Huawei, and others, and its private systems serve major enterprises. PixelBloom supports 20+ languages and plans to expand in North America, aiming to redefine office work through human-computer collaboration.
Anthropic temporarily banned OpenClaw founder Peter Steinbergers personal Claude account citing suspicious activity The two hour suspension sparked online debate with Steinberger insisting he complied with terms An Anthropic engineer suggested an overly sensitive abuse system triggered the ban highlighting tensions between open source projects and AI model providers
Anthropic temporarily banned OpenClaw founder Peter Steinbergers personal Claude account citing suspicious activity The two hour suspension sparked online debate with Steinberger insisting he complied with terms An Anthropic engineer suggested an overly sensitive abuse system triggered the ban highlighting tensions between open source projects and AI model providers
