We examine the potential of ChatGPT and other large language models in predicting stock market returns using news headlines. We use ChatGPT to assess whether each headline is good, bad, or neutral for firms' stock prices. We document a significantly positive correlation between ChatGPT scores and subsequent daily stock returns. We find that ChatGPT outperforms traditional sentiment analysis methods. More basic models such as GPT-1, GPT-2, and BERT cannot accurately forecast returns, indicating return predictability is an emerging capacity of complex language models. Long-short strategies based on ChatGPT-4 deliver the highest Sharpe ratio. Furthermore, we find predictability in both small and large stocks, suggesting market underreaction to company news. Predictability is stronger among smaller stocks and stocks with bad news, consistent with limits-to-arbitrage also playing an important role. Finally, we propose a new method to evaluate and understand the models' reasoning capabilities. Overall, our results suggest that incorporating advanced language models into the investment decision-making process can yield more accurate predictions and enhance the performance of quantitative trading strategies.

As generative AI settles into the mainstream, growing numbers of courses and certifications are promising entry into the "hot job" of prompt engineering. Having skills in using natural language (such as English) to "prompt" useful content out of AI models such as ChatGPT and Midjourney seems like something many employers would value. But is it as simple as doing a short course and riding the wave to a six-figure salary? A Washington Post article published in February did a lot to seed the notion that prompt engineers are "AI whisperers" who "program in prose". It dropped some big salary numbers and quoted a job ad by Silicon Valley company Anthropic calling for people who have "a creative hacker spirit and love solving puzzles".

In the past year or so, since the public release of OpenAI's ChatGPT, people have been making their peace with the idea that an omnipotent AI might be on the horizon. Sam Altman, the company's CEO, "believes that people need time to reckon with the idea that we may soon share Earth with a powerful new intelligence, before it remakes everything from work to human relationships," my colleague Ross Andersen reported after the two had several conversations. "ChatGPT was a way of serving notice." But OpenAI isn't Altman's only project, and it's not even his only project with ambitions to change the world. He is also a co-founder of a company called Tools for Humanity, which has the lofty goal of protecting people from the economic devastation that may arise from AI taking human jobs. The company's first major project is Worldcoin, which uses an evil-looking metallic orb--called the Orb--to take eyeball scans from people all over the world.

ChatGPT has proven it can help students with their homework, but now it is helping teachers create those very courses, a computer science professor told Fox News. The United Kingdom-based The Guardian newspaper announced that it was blocking ChatGPT owner OpenAI for being able to trawl content on its website. The Guardian announced in a report on its website last week that it is blocking OpenAI from using the paper's online content, citing concerns that its ChatGPT platform is "using unlicensed content to create its AI tools have led to writers bringing lawsuits against the company and creative industries calling for safeguards to protect their intellectual property." The move comes after OpenAI announced last month that it would enable websites to block the company's web crawler from accessing their content, with many online publishers joining The Guardian in choosing to block the crawler, according to the report. Other outlets listed as blocking the crawler, which uses information on websites to help generate AI content, include CNN, Reuters, Washington Post, Bloomberg, New York Times and The Athletic.

Judging whether an integer can be divided by prime numbers such as 2 or 3 may appear trivial to human beings, but can be less straightforward for computers. Here, we tested multiple deep learning architectures and feature engineering approaches on classifying integers based on their residues when divided by small prime numbers. We found that the ability of classification critically depends on the feature space. We also evaluated Automated Machine Learning (AutoML) platforms from Amazon, Google and Microsoft, and found that they failed on this task without appropriately engineered features. Furthermore, we introduced a method that utilizes linear regression on Fourier series basis vectors, and demonstrated its effectiveness. Finally, we evaluated Large Language Models (LLMs) such as GPT-4, GPT-J, LLaMA and Falcon, and demonstrated their failures. In conclusion, feature engineering remains an important task to improve performance and increase interpretability of machine-learning models, even in the era of AutoML and LLMs.

Diffusion-model-based text-guided image generation has recently made astounding progress, producing fascinating results in open-domain image manipulation tasks. Few models, however, currently have complete zero-shot capabilities for both global and local image editing due to the complexity and diversity of image manipulation tasks. In this work, we propose a method with a mixture-of-expert (MOE) controllers to align the text-guided capacity of diffusion models with different kinds of human instructions, enabling our model to handle various open-domain image manipulation tasks with natural language instructions. First, we use large language models (ChatGPT) and conditional image synthesis models (ControlNet) to generate a large number of global image transfer dataset in addition to the instruction-based local image editing dataset. Then, using an MOE technique and task-specific adaptation training on a large-scale dataset, our conditional diffusion model can edit images globally and locally. Extensive experiments demonstrate that our approach performs surprisingly well on various image manipulation tasks when dealing with open-domain images and arbitrary human instructions. Please refer to our project page: [https://oppo-mente-lab.github.io/moe_controller/]

Large language models (LLMs) have emerged as versatile tools in various daily applications. However, they are fraught with issues that undermine their utility and trustworthiness. These include the incorporation of erroneous references (citation), the generation of hallucinated information (correctness), and the inclusion of superfluous or omission of crucial details (fluency). To ameliorate these concerns, this study makes several key contributions. First, we build a dataset to train a critic model capable of evaluating the citation, correctness, and fluency of responses generated by LLMs in QA systems. Second, we propose an automated feedback mechanism that leverages the critic model to offer real-time feedback on heterogeneous aspects of generated text. Third, we introduce a feedback learning loop that uses this critic model to iteratively improve the performance of the LLM responsible for response generation. Experimental results demonstrate the efficacy of our approach, showing substantial improvements in citation and fluency metrics for ChatGPT, including a 4% precision increase in citation and an approximately 8% enhancement in the MAUVE metric for fluency, while maintaining high levels of correctness.

Communication games, which we refer to as incomplete information games that heavily depend on natural language communication, hold significant research value in fields such as economics, social science, and artificial intelligence. In this work, we explore the problem of how to engage large language models (LLMs) in communication games, and in response, propose a tuning-free framework. Our approach keeps LLMs frozen, and relies on the retrieval and reflection on past communications and experiences for improvement. An empirical study on the representative and widely-studied communication game, ``Werewolf'', demonstrates that our framework can effectively play Werewolf game without tuning the parameters of the LLMs. More importantly, strategic behaviors begin to emerge in our experiments, suggesting that it will be a fruitful journey to engage LLMs in communication games and associated domains.

Large language models (LLMs), such as GPT-4, PaLM, and LLaMa, have been shown to achieve remarkable performance across a variety of natural language tasks. Recent advancements in instruction tuning bring LLMs with ability in following user's instructions and producing human-like responses. However, the high costs associated with training and implementing LLMs pose challenges to academic research. Furthermore, the availability of pretrained LLMs and instruction-tune datasets for Vietnamese language is limited. To tackle these concerns, we leverage large-scale instruction-following datasets from open-source projects, namely Alpaca, GPT4All, and Chat-Doctor, which cover general domain and specific medical domain. To the best of our knowledge, these are the first instructional dataset for Vietnamese. Subsequently, we utilize parameter-efficient tuning through Low-Rank Adaptation (LoRA) on two open LLMs: Bloomz (Multilingual) and GPTJ-6B (Vietnamese), resulting four models: Bloomz-Chat, Bloomz-Doctor, GPTJ-Chat, GPTJ-Doctor.Finally, we assess the effectiveness of our methodology on a per-sample basis, taking into consideration the helpfulness, relevance, accuracy, level of detail in their responses. This evaluation process entails the utilization of GPT-4 as an automated scoring mechanism. Despite utilizing a low-cost setup, our method demonstrates about 20-30\% improvement over the original models in our evaluation tasks.

Graph structured data are widely existed and applied in the real-world applications, while it is a challenge to handling these diverse data and learning tasks on graph in an efficient manner. When facing the complicated graph learning tasks, experts have designed diverse Graph Neural Networks (GNNs) in recent years. They have also implemented AutoML in Graph, also known as AutoGraph, to automatically generate data-specific solutions. Despite their success, they encounter limitations in (1) managing diverse learning tasks at various levels, (2) dealing with different procedures in graph learning beyond architecture design, and (3) the huge requirements on the prior knowledge when using AutoGraph. In this paper, we propose to use Large Language Models (LLMs) as autonomous agents to simplify the learning process on diverse real-world graphs. Specifically, in response to a user request which may contain varying data and learning targets at the node, edge, or graph levels, the complex graph learning task is decomposed into three components following the agent planning, namely, detecting the learning intent, configuring solutions based on AutoGraph, and generating a response. The AutoGraph agents manage crucial procedures in automated graph learning, including data-processing, AutoML configuration, searching architectures, and hyper-parameter fine-tuning. With these agents, those components are processed by decomposing and completing step by step, thereby generating a solution for the given data automatically, regardless of the learning task on node or graph. The proposed method is dubbed Auto$^2$Graph, and the comparable performance on different datasets and learning tasks. Its effectiveness is demonstrated by its comparable performance on different datasets and learning tasks, as well as the human-like decisions made by the agents.