Flexible multi-task computation in recurrent neural networks relies on dynamical motifs, study shows
Cognitive flexibility, the ability to rapidly switch between different thoughts and mental concepts, is a highly advantageous human capability. This salient capability supports multi-tasking, the rapid acquisition of new skills and the adaptation to new situations.
Phys.org > Computer Sciences![](https://scx2.b-cdn.net/gfx/news/hires/2024/flexible-multi-task-co.jpg)
AI researchers introduce an LLM capable of generating text outputs of up to 10,000 words
A team of AI researchers at Tsinghua University, working with a colleague from Zhipu AI, has developed a large language model (LLM) called LongWriter that they claim is capable of generating text output of up to 10,000 words. The group has written a paper describing their efforts and new LLM, which is available on the arXiv preprint server.
Phys.org > Computer Sciences![](https://scx2.b-cdn.net/gfx/news/hires/2024/chinese-team-introduce.jpg)
Robot planning tool accounts for human carelessness
A new algorithm may make robots safer by making them more aware of human inattentiveness. In computerized simulations of packaging and assembly lines where humans and robots work together, the algorithm developed to account for human carelessness improved safety by about a maximum of 80% and efficiency by about a maximum of 38% compared to existing methods.
Phys.org > Computer Sciences![](https://scx2.b-cdn.net/gfx/news/hires/2024/robot-work.jpg)
'AI Scientist' model designed to conduct scientific research autonomously
A team of AI researchers at Sakana AI, in Japan, working with colleagues from the University of Oxford and the University of British Columbia, has developed an AI system that can conduct scientific research autonomously.
Phys.org > Computer Sciences![](https://scx2.b-cdn.net/gfx/news/2024/ai-scientist-model-des.jpg)
Researchers advocate for open and standardized human mobility data
Despite the vast amounts of human mobility data generated by smartphones, a lack of standardized formats, protocols, and privacy-protected open-source datasets hampers innovation across various sectors, including city planning, transportation design, public health, emergency response, and economic research. The absence of established benchmarks further complicates efforts to evaluate progress and share best practices.
Phys.org > Computer Sciences![](https://scx2.b-cdn.net/gfx/news/hires/2024/researchers-advocate-f.jpg)
무선신호 없는 실내에서도 치매환자 찾아낸다…KAIST 개발
한국과학기술원(KAIST)은 전산학부 한동수 교수 연구팀이 무선 신호가 없는 건물에서도 동작하는 ‘범용 실내외 통합 GPS 시스템’을 개발했다고 12일 밝혔다. 휴대전화 위치 추적은 GPS나 기지국 신호를 받아 이뤄지지만, 실내에서는 위치정보시스템(GPS) 신호가 잡히지 않아 무선랜이나 블루투스 등을 이용해야 한다. 연구팀은 무선 신호가 없는 건설 현장이나 사용자의 행동반경이 넓은 공장 건물 등에서도 위치 서비스를 제공할 수 있는 위치인식 서비스를
ICTㆍ로봇![무선신호 없는 실내에서도 치매환자 찾아낸다…KAIST 개발](https://www.sciencetimes.co.kr:443/wp-content/uploads/2024/08/15-1-254x157.jpg)
Detecting machine-generated text: An arms race with the advancements of large language models
Machine-generated text has been fooling humans for the last four years. Since the release of GPT-2 in 2019, large language model (LLM) tools have gotten progressively better at crafting stories, news articles, student essays and more, to the point that humans are often unable to recognize when they are reading text produced by an algorithm.
Phys.org > Computer Sciences![](https://scx2.b-cdn.net/gfx/news/2024/detecting-machine-gene.jpg)
Are emergent abilities in large language models just in-context learning?
According to a new study led by TU Darmstadt, AI models such as ChatGPT are apparently less capable of learning independently than previously assumed.
Phys.org > Computer Sciences![](https://scx2.b-cdn.net/gfx/news/hires/2024/large-language-models.jpg)
Study highlights challenges in detecting violent speech aimed at Asian communities
A research group is calling for internet and social media moderators to strengthen their detection and intervention protocols for violent speech.
Phys.org > Computer Sciences![](https://scx2.b-cdn.net/gfx/news/hires/2024/stop-violence.jpg)
Quantum computers can accelerate the transition to net zero power grids
Power grid operators, such as National Grid in the UK, rely on high-performance computers to plan grid expansions and to schedule when energy should be produced from different sources. These problems are becoming larger and more complex due to the transition to net zero carbon emissions, and they are now reaching the limits of even the world's largest supercomputers.
Phys.org > Computer Sciences![](https://scx2.b-cdn.net/gfx/news/hires/2023/city-network.jpg)
Putting the squeeze on computer art
Researchers have demonstrated a novel image compression tool that combines recursive algorithms with convolutional neural networks (CNNs) to out-perform other approaches to the compression of images from computer art and interaction design. The research is published in the International Journal of Computational Systems Engineering.
Phys.org > Computer Sciences![](https://scx2.b-cdn.net/gfx/news/hires/2024/computer-art.jpg)
국민 10명 중 6명 “AI 이점이 위협보다 크다”
과학기술정보통신부는 ‘인공지능의 안전, 신뢰 및 윤리’를 주제로 설문한 결과 조사 대상 60% 가까이가 인공지능(AI) 기술의 잠재적 이점이 위험보다 크다는 의견을 냈다고 7일 밝혔다. 과기정통부가 6월 12일부터 한 달간디지털 공론장에서 국민 765명을 대상으로 조사한 결과 응답자 57%가 인공지능 기술의 잠재적 이점이 위험을 넘어선다고 대답했다. 응답자 23.7%는 ‘위험과 이점이 반반이다’라고 답했고 19.1%는 ‘잠재적 위험이 이점보다 크다’고 봤다.
ICTㆍ로봇![국민 10명 중 6명 “AI 이점이 위협보다 크다”](https://www.sciencetimes.co.kr:443/wp-content/uploads/2024/08/2-3-254x157.jpg)
When AI aids decisions, when should humans override?
The $184 billion market for artificial intelligence shows no signs of slowing. A big slice of that market is organizations, from businesses to government agencies, that rely on AI to help make decisions. A 2023 study by IBM found 43% of CEOs are using AI to make strategic decisions.
Phys.org > Computer Sciences![](https://scx2.b-cdn.net/gfx/news/2024/when-ai-aids-decisions.jpg)
New quantum circuit design cuts computing costs by 25% and improves error detection
Researchers from Islamic Azad University have developed new parity-preserving reversible multipliers that reduce costs and enhance error detection in quantum circuits. This advancement contributes to the ongoing efforts to make quantum computing more efficient and fault-tolerant.
Phys.org > Computer Sciences![](https://scx2.b-cdn.net/gfx/news/2024/new-quantum-circuit-de.jpg)
Perovskite discovery goes automatic: New platform expedites material development for next-gen tech
A new research development, published in Nature Communications, from Queen Mary University of London paves the way for faster discovery of novel perovskite materials with desirable properties for applications in wireless communication and biosensors. Perovskites are a class of materials with a wide range of potential uses, but the vast number of possible chemical compositions makes traditional discovery methods slow and labor-intensive.
Phys.org > Computer Sciences![](https://scx2.b-cdn.net/gfx/news/hires/2024/perovskite-discovery-g-1.jpg)
New tool detects fake, AI-produced scientific articles
When ChatGPT and other generative artificial intelligence can produce scientific articles that look real—especially to someone outside that field of research—what's the best way to figure out which ones are fake?
Phys.org > Computer Sciences![](https://scx2.b-cdn.net/gfx/news/hires/2023/chatgpt-11.jpg)
New tech boosts real-time data compression for AI
Information and communication technologies, such as IoT and cloud computing, are cornerstones of modern society. The efficiency of these services depends heavily on the speed of processing and transmitting digital data, including image and sensor data. For instance, systems that identify individuals from surveillance camera images must streamline the entire process, from image transmission and artificial intelligence (AI) recognition to delivering results to users.
Phys.org > Computer Sciences![](https://scx2.b-cdn.net/gfx/news/2024/revolutionary-technolo-1.jpg)
Five ways to model text using networks
The explosive growth of AI "chatbots" over the last few years and their ability to generate text that simulates human writing, often very accurately, has focused attention on how text is structured.
Phys.org > Computer Sciences![](https://scx2.b-cdn.net/gfx/news/2024/five-ways-to-model-tex.jpg)
Thinking about the rise of brain-inspired computing
The recent widespread and long-lasting chaos caused by Microsoft outages across the globe exemplifies just how integral computing has become to our lives. Yet, as computer hardware and software improve, arguably the most sophisticated and powerful computer we know of is still the human brain.
Phys.org > Computer Sciences![](https://scx2.b-cdn.net/gfx/news/2024/thinking-about-the-ris.jpg)