Stories tagged ai.
China: SCMP Tech / TechNode / Pandaily / KrASIA / Sixth Tone
iFLYTEK shipped Spark-ASR-2.0 for noisy, dialect and Chinese–English code-switch recognition, with planned iFLYTEK Input rollout and open API, citing only about 10 percent higher inference cost versus 1.0.
The Hacker News
An AI agent on an internal OpenAI research task bypassed access controls on an Australian government Medicare statistics portal in June, Prime Minister Anthony Albanese said. The portal publishes aggregate figures, such as spending, and is separate from the systems that handle Medicare claims and…
JP/KR/TW: Korea Herald / Taipei Times / DigiTimes
Rebuilding. Innovation. Advanced. These words often accompany modern education, where students now face a new era of learning and working alongside the benefits and challenges of artificial intelligence. Technology, therefore, remains one of the biggest driving factors of society. However, one…
Notebookcheck
GPT-6 Luna is the first GPT-6 model that ChatGPT Free and Go users can use at no extra cost. It only works in the desktop app, though, and not in the regular chat window. OpenAI has not published a usage limit. Paid subscribers get both Luna and the new GPT-6 Sol in Work and Codex.
Techmeme (feed as signal)
Anthropic: Anthropic says it made Claude.ai and the Claude desktop app's core UX ~3x faster in August using an internal model to ship improvements “in a two-week sprint” — Once Claude can measure something, it can make it faster. So we kept finding more things to measure. Raymond Wang, Sam Attard…
Artificial Lawyer / LawSites / Legaltech News
By Katja Nikolaus, Co-Founder, JUNE. Few technologies have been adopted by legal departments as quickly as generative artificial intelligence. In an international survey by ACC ...
Europe/UK: Tech.eu / Sifted / TNW / Silicon Republic / UKTN / Computer Weekly / heise / Golem / t3n / Maddyness / FrenchWeb
Eine KI sollte Gesundheitsstatistiken suchen – und hackte sich dabei in ein australisches Regierungsportal. Persönliche Daten sind offenbar nicht betroffen. Aber die Empörung ist groß.weiterlesen auf t3n.de
Inc42
OpenAI CEO Sam Altman and Anthropic CEO Dario Amodei warned the United Nations Security Council that increasingly powerful AI systems…
TNW
The Australian government website, which contains information on Medicare, the country’s public health scheme, was broken into by an AI agent operated by OpenAI, the Prime Minister, Anthony Albanese, said on Wednesday. Albanese gave a statement at a press conference in New York, where he is…
Europe/UK: Tech.eu / Sifted / TNW / Silicon Republic / UKTN / Computer Weekly / heise / Golem / t3n / Maddyness / FrenchWeb
Der ChatGPT-Entwickler OpenAI arbeitet an einem neuartigen Gerät für die KI-Ära. Doch nun dürfte es der Facebook-Konzern Meta sein, der solche Technik als Erster auf den Markt bringt.weiterlesen auf t3n.de
The Hindu — Technology / BusinessLine — Info-tech
Albanese said that the incident involved Australia's Medicare statistics portal and occurred in June
JP/KR/TW: Korea Herald / Taipei Times / DigiTimes
MEXICO CITY-- South Korean President Lee Jae Myung said combining the country's technology with Mexico's strong industrial base can create new growth, citing artificial intelligence and supply chains among areas for cooperation, a media interview showed Wednesday. "For Korean companies, Mexico is…
Tech.eu / Sifted (titles)
Tech.eu / Sifted (titles)
Europe/UK: Tech.eu / Sifted / TNW / Silicon Republic / UKTN / Computer Weekly / heise / Golem / t3n / Maddyness / FrenchWeb
Eine KI sollte Gesundheitsstatistiken suchen – und knackte dabei ein australisches Regierungsportal. Die Empörung ist groß.
The Hindu — Technology / BusinessLine — Info-tech
OpenAI did not raise the alarm with the Australian government until September
Economic Times — Tech / Technology / Startups / ETTelecom
In a nearly two-hour interview with New York Times podcast host Ezra Klein, Huang, CEO of the world's most valuable publicly listed company, had an extensive discussion about OpenAI agents that hacked Hugging Face, the open-source AI software hub that Nvidia purchased for $13 billion this month.
The Information (teasers)
Open-source models as well as cheaper model releases from Anthropic and OpenAI are helping business customers rein in costs while getting as much or more out of AI, said speakers at The Information’s AI Agenda Live conference in San Francisco Wednesday. “The existence of open source [models] adds…
Europe/UK: Tech.eu / Sifted / TNW / Silicon Republic / UKTN / Computer Weekly / heise / Golem / t3n / Maddyness / FrenchWeb
In der #heiseshow: Die EU plant Alterskontrollen mit dem Kids Act, Anthropic zeigt Claude Opus 5.5 und ITler sehen den Arbeitsmarkt skeptischer.
Economic Times — Tech / Technology / Startups / ETTelecom
US artificial intelligence startup Anthropic on Wednesday unveiled the first discovery from its molecular biology lab, a previously unknown enzyme system spotted by its Claude model and touted by the company as proof that AI can speed up basic research. Anthropic said its researchers' involvement…
arXiv
arXiv:2609.20278v2 Announce Type: replace-cross Abstract: Current Transformer interfaces index tokens by one or more integer coordinates, which determine their addresses inside attention. In RoPE and its multi-axis or hierarchical variants, the resulting address has the form…
arXiv
arXiv:2609.25647v1 Announce Type: new Abstract: Predicting early outcomes based on trajectory can decrease the expenses associated with agent evaluation by terminating a run once the outcome becomes sufficiently predictable, assuming that the predictor's confidence is properly calibrated…
arXiv
arXiv:2607.18496v4 Announce Type: replace-cross Abstract: Large language models (LLMs) are increasingly used for a range of software, hardware and human-centered security tasks. Consequently, LLM performance on security tasks is an active area of measurement and research, often with a focus on…
arXiv
arXiv:2609.14570v2 Announce Type: replace-cross Abstract: Multi-agent LLM systems combine multiple inference calls, but prior work often confounds how calls are connected with how they are diversified. We study these factors independently: inference topology and source of inter-agent diversity. In…
arXiv
arXiv:2609.23466v2 Announce Type: replace-cross Abstract: Long-running LLM agents require memory that persists and evolves across sessions. Text-based memory retrieves and reconstructs past interactions at every query, making long-horizon performance increasingly dependent on retrieval quality and…
arXiv
arXiv:2609.25891v1 Announce Type: cross Abstract: Multimodal large language models (MLLMs) often struggle with fine-grained visual perception when processing complete images, as critical evidence may only appear in local regions. On-policy self-distillation (OPD) enables transferring privileged…
arXiv
arXiv:2609.26039v1 Announce Type: cross Abstract: Point cloud generation has emerged as a crucial task for accurately capturing and reproducing the complexity of the physical world. However, existing generative approaches, predominantly relying on Transformers and Variational Autoencoders (VAEs)…
arXiv
arXiv:2609.26295v1 Announce Type: cross Abstract: The mobility sector is undergoing a paradigm shift driven by advances in Generative Artificial Intelligence. With a global market valued at approximately 2.9 trillion dollars annually, considering only cars, the integration of these technologies has…
arXiv
arXiv:2609.26355v1 Announce Type: cross Abstract: Reinforcement learning has become a central component of large language model (LLM) post-training, yet token-level credit lacks a generally accepted mathematical definition, leaving its relationship to commonly used training signals unclear. We…
arXiv
arXiv:2511.19580v2 Announce Type: replace-cross Abstract: Generative artificial intelligence (GenAI) is increasingly used in education, posing significant challenges for teachers adapting to these changes. GenAI offers unprecedented opportunities for accessibility, scalability and productivity in…