LLMs

Twitter

Facebook

Copy Link

Key Facts

Year2018

Key PlayersOpenAI, Google, Anthropic

Top News

Dartmouth Study Reveals AI in Healthcare Can Increase Physician Workload With Flawed Responses

A large-scale Dartmouth study analyzing 146,000 patient-doctor conversations found that AI in healthcare often creates more work than it saves. AI-generated responses frequently contain errors and irrelevant details, forcing physicians to spend more time editing AI mistakes than writing messages themselves. However, researchers developed a technique that improves accuracy by 33% and reduces editing by 26%.

Health2 sources

Dartmouth Study Reveals AI in Healthcare Can Increase Physician Workload With Flawed Responses

Health2 sources

Companies slash AI costs 65% by making Claude and Codex talk like cavemen

Developers at OpenAI, Nvidia, and GitHub are using a 'caveman' plugin that strips verbose AI responses down to bare essentials. The tool, created by Julius Brussee, reduces token usage by 65% as companies scramble to curb rising AI costs. A senior OpenAI employee has even contributed code to the project, highlighting the industry's urgent need for cost-saving measures.

Technology2 sources

Companies slash AI costs 65% by making Claude and Codex talk like cavemen

Technology2 sources

AI writing tools inject political bias into posts, Oxford study reveals hidden opinion influence

A groundbreaking study from the Oxford Internet Institute and Hasso Plattner Institute found that AI writing tools consistently shift the meaning of social media posts, even when instructed to preserve original intent. The research demonstrates how AI-mediated communication can subtly steer opinion across millions of interactions, creating measurable shifts in community perspectives on divisive topics from abortion to climate change.

Science3 sources

AI writing tools inject political bias into posts, Oxford study reveals hidden opinion influence

Science3 sources

Amazon Mechanical Turk stops accepting new customers as AI overtakes crowdsourcing platform

Amazon announced that its Mechanical Turk crowdsourcing service will close to new customers on July 30, 2026, though existing users can continue. Launched in 2005, the platform once employed human workers for micro-tasks resistant to full automation. But a 2023 analysis found that 33% to 46% of workers were using large language models to complete tasks, highlighting how artificial intelligence has now overtaken the service that once powered it.

Technology3 sources

Amazon Mechanical Turk stops accepting new customers as AI overtakes crowdsourcing platform

Technology3 sources

US-China AI tensions reach new heights as both nations move to restrict each other's models

The geopolitical standoff over AI intensified this week as Beijing held talks about curbing overseas access to Chinese AI models while Washington probed American companies adopting cost-effective AI alternatives from China. Both nations now treat advanced AI as a critical national asset requiring stringent controls, with officials citing national security concerns and ideological influence of AI systems.

Policy25 sources

US-China AI tensions reach new heights as both nations move to restrict each other's models

Policy25 sources

JPMorgan AI Chief Teresa Heitsenrether to retire after shaping bank's AI strategy for 40 years

Teresa Heitsenrether, JPMorgan Chase's Chief Data and Analytics Officer, is retiring at year-end after nearly four decades with the bank. She led the bank's AI strategy for three years, overseeing more than 600 AI use cases and driving adoption of large language models across the organization. Chief Technology Officer Scot Baldry will assume her responsibilities as the bank continues its AI transformation.

Business2 sources

JPMorgan AI Chief Teresa Heitsenrether to retire after shaping bank's AI strategy for 40 years

Business2 sources

First study on AI ghosts reveals people prefer talking to the dead as reincarnations, not narrators

Researchers at the University of Colorado Boulder conducted the first user experience study on AI ghosts, finding that all 16 participants preferred AI simulations that spoke as their deceased loved ones rather than about them. While everyone found the experience meaningful enough to repeat, nearly all expressed concern about potential emotional dependency on grief technology.

Science3 sources

First study on AI ghosts reveals people prefer talking to the dead as reincarnations, not narrators

Science3 sources

Portugal launches Amalia, its first open-source AI model, in Europe's sovereignty push

Portugal unveiled Amalia, its first national AI model built for European Portuguese, joining Europe's push for AI sovereignty. Funded with €5.5 million in EU recovery funds, the open-source large language model enables public institutions, companies, and researchers to build tailored AI applications while reducing dependence on U.S. providers like OpenAI and Google.

Technology2 sources

Portugal launches Amalia, its first open-source AI model, in Europe's sovereignty push

Technology2 sources

AI Model Claude Helps Physicists Crack 10-Year-Old Jamming Conjecture in Complex Systems

Nobel Prize winner Giorgio Parisi and Francesco Zamponi from La Sapienza University of Rome enlisted AI model Claude to solve a mathematical problem that had stumped researchers for over a decade. The collaboration between AI and human researchers led to a surprisingly simple proof of a jamming conjecture, showing how AI is transforming scientific research while highlighting the continued need for human verification.

Science2 sources

AI Model Claude Helps Physicists Crack 10-Year-Old Jamming Conjecture in Complex Systems

Science2 sources

UBTech unveils companion humanoid robot that recognizes emotions with 90% accuracy

Chinese robotics firm UBTech launched the UWORLD U1 Series, claiming it's the world's first full-size ultra-bionic humanoid robot designed for mass production and AI companionship. The robot can identify over 20 emotional states with 90% accuracy and features biomimetic skin, long-term memory, and customizable appearances. Priced from $17,650, it targets China's 90 million adults living alone and 118 million empty-nest seniors.

Technology3 sources

UBTech unveils companion humanoid robot that recognizes emotions with 90% accuracy

Technology3 sources

San Diego Charter School Spends $500,000 on ChatGPT-Powered Humanoid Robots as Teaching Partners

Altus Schools in San Diego has invested $500,000 in two Ameca humanoid robots to serve as AI-powered teaching partners in classrooms. The ChatGPT-enabled robots can function as tutors, wellness coaches, career guides, and translators. While school officials call it groundbreaking research into physical AI in education, critics question whether the expensive experiment delivers enough value to justify the cost.

Technology3 sources

San Diego Charter School Spends $500,000 on ChatGPT-Powered Humanoid Robots as Teaching Partners

Technology3 sources

Meituan unveils LongCat-2.0 AI model trained entirely on domestic Chinese chips

Chinese tech giant Meituan released LongCat-2.0, a 1.6-trillion-parameter open-source AI model trained on 50,000 domestic chips. The model spent two months anonymously dominating OpenRouter as "Owl Alpha" before its official reveal. This marks a significant shift in China's AI industry toward self-sufficiency amid ongoing U.S. export restrictions on advanced semiconductors.

Technology6 sources

Meituan unveils LongCat-2.0 AI model trained entirely on domestic Chinese chips

Technology6 sources

Meta Brain2Qwerty v2 decodes typed sentences from brain scans at 61% accuracy without implants

Meta unveiled Brain2Qwerty v2, an AI system that converts brain activity into text without surgery, achieving 61% word accuracy using magnetoencephalography scanners. The non-invasive brain-computer interface trained on 22,000 sentences from nine volunteers marks a leap from single-digit accuracy in previous methods, though it requires room-sized equipment and can't operate in real time.

Technology7 sources

Meta Brain2Qwerty v2 decodes typed sentences from brain scans at 61% accuracy without implants

Technology7 sources

DeepSeek open-sources DSpark framework to speed up LLM inference by up to 85%

DeepSeek released DSpark, an open-source framework using speculative decoding to accelerate AI responses by 60-85%. Released under MIT license with DeepSpec training code, the system tackles costly AI deployment challenges. The framework works across multiple open-weight models including Qwen and Gemma, potentially reshaping how developers deploy large language models.

Technology2 sources

DeepSeek open-sources DSpark framework to speed up LLM inference by up to 85%

Technology2 sources

AI finds thousands of hidden vulnerabilities, overwhelming security teams with unprecedented discoveries

Advanced AI models are uncovering thousands of previously hidden bugs in open-source code faster than security teams can patch them. The Linux Foundation launched Akrites alongside tech giants to coordinate fixes, while the Athena coalition has already processed over 20,000 findings across 500 projects. With fewer than 5% of AI-discovered vulnerabilities patched and attackers moving quickly, the industry faces a critical race to secure software infrastructure.

Technology2 sources

AI finds thousands of hidden vulnerabilities, overwhelming security teams with unprecedented discoveries

Technology2 sources

Prev 1…567…29 Next