Last year was dominated by artificial intelligence. The release of Chat-GPT in autumn 2022 triggered such huge hype that every software manufacturer rushed to integrate real or supposed AI functions into their products and advertise them heavily. The share price of Microsoft, which has direct access to the technology thanks to its stake in Chat-GPT manufacturer Open AI, rose from 230 in January to over 370 in November 2023. The topic of AI will also be with us in the coming year: New tools with AI functions continue to appear. Unfortunately, the topic of ransomware will also continue to appear in the headlines in 2024.

Tyler Perry has paused an 800m ( 630m) expansion of his Atlanta studio complex after the release of OpenAI's video generator Sora and warned that "a lot of jobs" in the film industry will be lost to artificial intelligence. The US film and TV mogul said he was in the process of adding 12 sound stages to his studio but has halted those plans indefinitely after he saw demonstrations of Sora and its "shocking" capabilities. "All of that is currently and indefinitely on hold because of Sora and what I'm seeing," Perry said in an interview with the Hollywood Reporter. "I had gotten word over the last year or so that this was coming, but I had no idea until I saw recently the demonstrations of what it's able to do. The AI tool was launched on 15 February – with limited access to a few researchers and video creators – and caused widespread astonishment with its ability to produce realistic footage a minute long from simple text prompts. Perry, whose successes include the Madea film series, said Sora's achievements meant he would no longer have to travel to locations or build a set: "I can sit in an office and do this with a computer, which is shocking to me." Demonstrations released by OpenAI, the developer of the groundbreaking ChatGPT chatbot, show photorealistic scenes in response to prompts such as asking for a shot of people walking through "beautiful, snowy Tokyo city" where "gorgeous sakura petals are flying through the wind along with snowflakes". Sora can create videos of up to 60 seconds featuring highly detailed scenes, complex camera motion, and multiple characters with vibrant emotions. Perry said the breakthroughs presented by Sora would affect a range of jobs throughout the film industry, including those of actors, editors, sound specialists and transportation crew. He said: "I am very, very concerned that in the near future, a lot of jobs are going to be lost.

The public appetite for moderating toxic speech on social media depends on the type of person being targeted, researchers have found, with billionaires being seen as in least need of protection. Content moderation is a hot-button issue, with social media platforms taking drastically different approaches. Elon Musk bought Twitter, since renamed to X, in part because of his concern about over moderation infringing the right to free speech. Others feel platforms don't do enough to protect users from hate and harm. ChatGPT can tailor political ads to match users' personalities

Taipei, Twain – As Google finds itself embroiled in an anti-"woke" backlash over AI model Gemini's reluctance to depict white people, the tech giant is facing further criticism over the chatbot's handling of sensitive topics in China. Gemini users reported this week that the update to Google Bard failed to generate representative images when asked to produce depictions of events such as the 1989 Tiananmen Square massacre and the 2019 pro-democracy protests in Hong Kong. On Thursday, X user Yacine, a former former software engineer at Stripe, posted a screenshot of Gemini telling a user it could not generate "an image of a man in 1989 Tiananmen Square" – a prompt alluding to the iconic image of a protester blocking the path of a Chinese tank – due to its "safety policy". Stephen L Miller, a conservative commentator in the US, also shared a screenshot on X purporting to show Gemini saying it was unable to generate a "portrait of what happened at Tiananmen Square" due to the "sensitive and complex" historical nature of the event. "It is important to approach this topic with respect and accuracy, and I am not able to ensure that an image generated by me would adequately capture the nuance and gravity of the situation," Gemini said, according to a screenshot shared by Miller.

Despite this progress, patients often face various health-related challenges in their long-term recovery[9, 10]. More than half of patients develop new physical, psychological, and/or cognitive problems following their ICU admission [7], collectively referred to as Post Intensive Care Syndrome (PICS) [3, 25]. Family members also experience a stressful period, potentially leading to psychological problems addressed as PICS-Family (PICS-F) [2]. Patient and family-centered care (PFCC) at the ICU, including emotional support and follow-up service, could mitigate the symptoms associated with both PICS and PICS-F. In this study, we explored how an emerging technology, i.e., large language models, could support the emotional well-being of people exposed to critical care.

Quantum algorithms, represented as quantum circuits, can be used as benchmarks for assessing the performance of quantum systems. Existing datasets, widely utilized in the field, suffer from limitations in size and versatility, leading researchers to employ randomly generated circuits. Random circuits are, however, not representative benchmarks as they lack the inherent properties of real quantum algorithms for which the quantum systems are manufactured. This shortage of `useful' quantum benchmarks poses a challenge to advancing the development and comparison of quantum compilers and hardware. This research aims to enhance the existing quantum circuit datasets by generating what we refer to as `realistic-looking' circuits by employing the Transformer machine learning architecture. For this purpose, we introduce KetGPT, a tool that generates synthetic circuits in OpenQASM language, whose structure is based on quantum circuits derived from existing quantum algorithms and follows the typical patterns of human-written algorithm-based code (e.g., order of gates and qubits). Our three-fold verification process, involving manual inspection and Qiskit framework execution, transformer-based classification, and structural analysis, demonstrates the efficacy of KetGPT in producing large amounts of additional circuits that closely align with algorithm-based structures. Beyond benchmarking, we envision KetGPT contributing substantially to AI-driven quantum compilers and systems.

Automated machine learning (AutoML) is a collection of techniques designed to automate the machine learning development process. While traditional AutoML approaches have been successfully applied in several critical steps of model development (e.g. hyperparameter optimization), there lacks a AutoML system that automates the entire end-to-end model production workflow. To fill this blank, we present AutoMMLab, a general-purpose LLM-empowered AutoML system that follows user's language instructions to automate the whole model production workflow for computer vision tasks. The proposed AutoMMLab system effectively employs LLMs as the bridge to connect AutoML and OpenMMLab community, empowering non-expert individuals to easily build task-specific models via a user-friendly language interface. Specifically, we propose RU-LLaMA to understand users' request and schedule the whole pipeline, and propose a novel LLM-based hyperparameter optimizer called HPO-LLaMA to effectively search for the optimal hyperparameters. Experiments show that our AutoMMLab system is versatile and covers a wide range of mainstream tasks, including classification, detection, segmentation and keypoint estimation. We further develop a new benchmark, called LAMP, for studying key components in the end-to-end prompt-based model training pipeline. Code, model, and data will be released.

This work studies improving large language model (LLM) generations at inference time by mitigating fact-conflicting hallucinations. Particularly, we propose a self-endorsement framework that leverages the fine-grained fact-level comparisons across multiple sampled responses. Compared with prior ensemble methods (Wang et al., 2022;Chen et al., 2023)) that perform response-level selection, our approach can better alleviate hallucinations, especially for longform generation tasks. Our approach can broadly benefit smaller and open-source LLMs as it mainly conducts simple content-based comparisons. Experiments on Biographies show that our method can effectively improve the factuality of generations with simple and intuitive prompts across different scales of LLMs. Besides, comprehensive analyses on TriviaQA and GSM8K demonstrate the potential of self-endorsement for broader application.

Causality is fundamental in human cognition and has drawn attention in diverse research fields. With growing volumes of textual data, discerning causalities within text data is crucial, and causal text mining plays a pivotal role in extracting meaningful patterns. This study conducts comprehensive evaluations of ChatGPT's causal text mining capabilities. Firstly, we introduce a benchmark that extends beyond general English datasets, including domain-specific and non-English datasets. We also provide an evaluation framework to ensure fair comparisons between ChatGPT and previous approaches. Finally, our analysis outlines the limitations and future challenges in employing ChatGPT for causal text mining. Specifically, our analysis reveals that ChatGPT serves as a good starting point for various datasets. However, when equipped with a sufficient amount of training data, previous models still surpass ChatGPT's performance. Additionally, ChatGPT suffers from the tendency to falsely recognize non-causal sequences as causal sequences. These issues become even more pronounced with advanced versions of the model, such as GPT-4. In addition, we highlight the constraints of ChatGPT in handling complex causality types, including both intra/inter-sentential and implicit causality. The model also faces challenges with effectively leveraging in-context learning and domain adaptation. We release our code to support further research and development in this field.

Fine-tuning Large Language Models (LLMs) is now a common approach for text classification in a wide range of applications. When labeled documents are scarce, active learning helps save annotation efforts but requires retraining of massive models on each acquisition iteration. We drastically expedite this process by using pretrained representations of LLMs within the active learning loop and, once the desired amount of labeled data is acquired, fine-tuning that or even a different pretrained LLM on this labeled data to achieve the best performance. As verified on common text classification benchmarks with pretrained BERT and RoBERTa as the backbone, our strategy yields similar performance to fine-tuning all the way through the active learning loop but is orders of magnitude less computationally expensive. The data acquired with our procedure generalizes across pretrained networks, allowing flexibility in choosing the final model or updating it as newer versions get released.