Companies are also spending time and money improving their models by testing them with real people. A technique called reinforcement learning with human feedback, where human testers manually improve a bot's answers and then feed them back into the system to improve it, is widely credited with making ChatGPT so much better than chatbots that came before it. A popular approach is to connect chatbots up to databases of factual or more trustworthy information, such as Wikipedia, Google search or bespoke collections of academic articles or business documents.

Becky Jefferies made it to the front of the aisle amid her marriage ceremony in Greece when she realized her wedding dress was incomplete. Artificial intelligence is the hot new topic of conversation as AI tools are increasingly threatening to replace the "human" component in a variety of everyday tasks -- or are already doing so. For some, artificial intelligence apps like ChatGPT are being used to write or draft academic papers, answer medical questions and receive advice on a wide array of topics. Some engaged couples are now using the platforms for help when writing their wedding vows. Fox News Digital tried using ChatGPT to write wedding vows -- typing "write my wedding vows" into the platform.

Sam Altman, the CEO of artificial intelligence lab OpenAI, told a Senate panel he welcomes federal regulation on the technology'to mitigate' its risks. A stranger in a coffee shop can watch you and learn virtually everything about you, where you've been and even predict your movements "with greater ease and precision than ever before," experts say. All the user would need is a photo and advanced artificial intelligence technology that already exists, said Kevin Baragona, a founder of DeepAI.org. "There are services online that can use a photo of you, and I can find everything. Every instance of your face on the internet, every place you've been and use that for stalker-type purposes," Baragona told Fox News Digital.

Taiwan's Asustek Computer plans to introduce one of the first services that lets companies tap into the potential of generative artificial intelligence while keeping control over their data. The novelty of the Taipei-based firm offering, called AFS Appliance, is that all of the hardware will be installed at the client's own facilities -- to maintain security and control. The AI computational platform, built on Nvidia Corp.'s chip technology, will be operated and updated with new data by Asustek, also known as Asus. A major concern around services like OpenAI is that they're operated through online data centers that can expose sensitive information. Samsung Electronics Co. banned employees from using OpenAI's ChatGPT after it found workers had uploaded sensitive code to the platform.

Steve Wozniak is no fan of Elon Musk. In February, the Apple co-founder described the Tesla, SpaceX and Twitter owner as a "cult leader" and called him dishonest. Yet, in late March, the tech titans came together, joining dozens of high-profile academics, researchers and entrepreneurs in calling for a six-month pause in training artificial intelligence systems more powerful than GPT-4, the latest version of Chat GPT, the chatbot that has taken the world by storm. Their letter, penned by the United States-based Future of Life Institute, said the current rate of AI progress was becoming a "dangerous race to ever-larger unpredictable black-box models". The "emergent capabilities" of these models, the letter said, should be "refocused on making today's powerful, state-of-the-art systems more accurate, safe, interpretable, transparent, robust, aligned, trustworthy and loyal".

Despite a growing interest in diffusion-based language models, existing work has not shown that these models can attain nontrivial likelihoods on standard language modeling benchmarks. In this work, we take the first steps towards closing the likelihood gap between autoregressive and diffusion-based language models, with the goal of building and releasing a diffusion model which outperforms a small but widely-known autoregressive model. We pursue this goal through algorithmic improvements, scaling laws, and increased compute. On the algorithmic front, we introduce several methodological improvements for the maximum-likelihood training of diffusion language models. We then study scaling laws for our diffusion models and find compute-optimal training regimes which differ substantially from autoregressive models. Using our methods and scaling analysis, we train and release Plaid 1B, a large diffusion language model which outperforms GPT-2 124M in likelihood on benchmark datasets and generates fluent samples in unconditional and zero-shot control settings.

Few-shot fine-tuning and in-context learning are two alternative strategies for task adaptation of pre-trained language models. Recently, in-context learning has gained popularity over fine-tuning due to its simplicity and improved out-of-domain generalization, and because extensive evidence shows that fine-tuned models pick up on spurious correlations. Unfortunately, previous comparisons of the two approaches were done using models of different sizes. This raises the question of whether the observed weaker out-of-domain generalization of fine-tuned models is an inherent property of fine-tuning or a limitation of the experimental setup. In this paper, we compare the generalization of few-shot fine-tuning and in-context learning to challenge datasets, while controlling for the models used, the number of examples, and the number of parameters, ranging from 125M to 30B. Our results show that fine-tuned language models can in fact generalize well out-of-domain. We find that both approaches generalize similarly; they exhibit large variation and depend on properties such as model size and the number of examples, highlighting that robust task adaptation remains a challenge.

We propose a novel task, G4C, to study teacher-student natural language interactions in a goal-driven and grounded environment. Dungeons and Dragons (D&D), a role-playing game, provides an ideal setting to investigate such interactions. Here, the Dungeon Master (DM), i.e., the teacher, guides the actions of several players -- students, each with their own personas and abilities -- to achieve shared goals grounded in a fantasy world. Our approach is to decompose and model these interactions into (1) the DM's intent to guide players toward a given goal; (2) the DM's guidance utterance to the players expressing this intent; and (3) a theory-of-mind (ToM) model that anticipates the players' reaction to the guidance one turn into the future. We develop a novel reinforcement learning (RL) method for training a DM that generates guidance for players by rewarding utterances where the intent matches the ToM-anticipated player actions. Human and automated evaluations show that a DM trained to explicitly model intents and incorporate ToM of the players using RL generates better-quality guidance that is 3x more likely to fulfill the DM's intent than a vanilla natural language generation (NLG) approach.

Strategic reasoning enables agents to cooperate, communicate, and compete with other agents in diverse situations. Existing approaches to solving strategic games rely on extensive training, yielding strategies that do not generalize to new scenarios or games without retraining. Large Language Models (LLMs), with their ability to comprehend and generate complex, context-rich language, could prove powerful as tools for strategic gameplay. This paper introduces an approach that uses pretrained LLMs with few-shot chain-of-thought examples to enable strategic reasoning for AI agents. Our approach uses systematically generated demonstrations of reasoning about states, values, and beliefs to prompt the model. Using extensive variations of simple matrix games, we show that strategies that are derived based on systematically generated prompts generalize almost perfectly to new game structures, alternate objectives, and hidden information. Additionally, we demonstrate our approach can lead to human-like negotiation strategies in realistic scenarios without any extra training or fine-tuning. Our results highlight the ability of LLMs, guided by systematic reasoning demonstrations, to adapt and excel in diverse strategic scenarios.

The open-ended Visual Question Answering (VQA) task requires AI models to jointly reason over visual and natural language inputs using world knowledge. Recently, pre-trained Language Models (PLM) such as GPT-3 have been applied to the task and shown to be powerful world knowledge sources. However, these methods suffer from low knowledge coverage caused by PLM bias -- the tendency to generate certain tokens over other tokens regardless of prompt changes, and high dependency on the PLM quality -- only models using GPT-3 can achieve the best result. To address the aforementioned challenges, we propose RASO: a new VQA pipeline that deploys a generate-then-select strategy guided by world knowledge for the first time. Rather than following the de facto standard to train a multi-modal model that directly generates the VQA answer, RASO first adopts PLM to generate all the possible answers, and then trains a lightweight answer selection model for the correct answer. As proved in our analysis, RASO expands the knowledge coverage from in-domain training data by a large margin. We provide extensive experimentation and show the effectiveness of our pipeline by advancing the state-of-the-art by 4.1% on OK-VQA, without additional computation cost. Code and models are released at http://cogcomp.org/page/publication_view/1010