A large portion of society united around the same vision and ideas carries enormous energy. That is precisely what political figures would like to accumulate for their cause. With this goal in mind, they can sometimes resort to distorting or hiding the truth, unintentionally or on purpose, which opens the door for misinformation and disinformation. Tools for automatic detection of check-worthy claims would be of great help to moderators of debates, journalists, and fact-checking organizations. While previous work on detecting check-worthy claims has focused on text, here we explore the utility of the audio signal as an additional information source. We create a new multimodal dataset (text and audio in English) containing 48 hours of speech. Our evaluation results show that the audio modality together with text yields improvements over text alone in the case of multiple speakers. Moreover, an audio-only model could outperform a text-only one for a single speaker.

Sentiment analysis (SA) has been a long-standing research area in natural language processing. It can offer rich insights into human sentiments and opinions and has thus seen considerable interest from both academia and industry. With the advent of large language models (LLMs) such as ChatGPT, there is a great potential for their employment on SA problems. However, the extent to which existing LLMs can be leveraged for different sentiment analysis tasks remains unclear. This paper aims to provide a comprehensive investigation into the capabilities of LLMs in performing various sentiment analysis tasks, from conventional sentiment classification to aspect-based sentiment analysis and multifaceted analysis of subjective texts. We evaluate performance across 13 tasks on 26 datasets and compare the results against small language models (SLMs) trained on domain-specific datasets. Our study reveals that while LLMs demonstrate satisfactory performance in simpler tasks, they lag behind in more complex tasks requiring deeper understanding or structured sentiment information. However, LLMs significantly outperform SLMs in few-shot learning settings, suggesting their potential when annotation resources are limited. We also highlight the limitations of current evaluation practices in assessing LLMs' SA abilities and propose a novel benchmark, \textsc{SentiEval}, for a more comprehensive and realistic evaluation. Data and code during our investigations are available at \url{https://github.com/DAMO-NLP-SG/LLM-Sentiment}.

Obtaining labeled data to train a model for a task of interest is often expensive. Prior work shows training models on multitask data augmented with task descriptions (prompts) effectively transfers knowledge to new tasks. Towards efficiently building task-specific models, we assume access to a small number (32-1000) of unlabeled target-task examples and use those to retrieve the most similar labeled examples from a large pool of multitask data augmented with prompts. Compared to the current practice of finetuning models on uniformly sampled prompted multitask data (e.g.: FLAN, T0), our approach of finetuning on cross-task nearest neighbors is significantly more data-efficient. Using only 2% of the data from the P3 pool without any labeled target-task data, our models outperform strong baselines trained on all available data by 3-30% on 12 out of 14 datasets representing held-out tasks including legal and scientific document QA. Similarly, models trained on cross-task nearest neighbors from SuperNaturalInstructions, representing about 5% of the pool, obtain comparable performance to state-of-the-art models on 12 held-out tasks from that pool. Moreover, the models produced by our approach also provide a better initialization than single multitask finetuned models for few-shot finetuning on target-task data, as shown by a 2-23% relative improvement over few-shot finetuned T0-3B models on 8 datasets.

But China could catch up, according to analysts, as AI solutions take years to be perfected. Chinese internet companies "are arguably more advanced than US internet companies, depending on how you're measuring advancement," Kendra Schaefer, head of tech policy research at Trivium China tells the BBC.

Universal Music Group (UMG) is partnering with Endel, an "AI sound wellness company" specializing in personalized algorithmic soundscapes, the companies announced today. The partnership aims to let UMG artists create machine-learning-generated sounds for activities like sleep, relaxation and focus. The record label "will use Endel's proprietary AI technology to enable UMG artists to create science-backed soundscapes," the companies said. The soundscapes can contain new music and updated versions of back-catalog tracks. The companies emphasize that the project "will always respect creators' rights and put artists at the center of the creative process," adding that musicians and their teams have the final say on the results.

In The Legend of Zelda, Hyrule is a land constantly imperiled by maleficent lords of shadow, cataclysmic volcanic eruptions, and an intangible sense of paranormal gloom that sucks the will to live out of every man, Zora, and Goron. Its nations are stratified across the land, and all of them live under the muzzling bounds of an autocratic royal bloodline. In other words, the people of Zelda need a free press, and in the newest game of the franchise--called Tears of the Kingdom--Hylians have discovered that occasionally, the pen is mightier than the sword. Those who embark on the adventure will discover ancient vistas, glorious ruins, and, most surprisingly, a proud celebration of the power of journalism. At last, Link is asking the tough questions.

Artists concerned about others passing off AI-generated art as their own will now be able to breathe a bit easier: Microsoft has agreed to sign all AI art that its apps generate with a cryptographic watermark indicating it was made with an algorithm. The Coalition for Content Provenance and Authority (C2PA) began work in 2021 to develop an open standard for indicating the origin of digital images, and whether they were authentic or AI-generated. The issue was thrust into the spotlight in March, when AI-generated images of the Pope in a stylish puffy jacket went viral, and AI-art generator Midjourney clamped down to prevent even more. Microsoft, a founding member of the C2PA, will announce at its Microsoft Build developer conference this week that it will cryptographically sign AI-generated images from Bing Image Creator and Microsoft Designer. Images made with Bing Image Creator already include a small "b" for the Bing logo in the bottom right-hand corner.

This is part of the data recipe for Facebook's new large language model, which the company claims is able to detect and rein in harmful content in over 100 languages. Bumble uses similar technology to detect rude and unwanted messages in at least 15 languages. Google uses it for everything from translation to filtering newspaper comment sections. All have comparable recipes and the same dominant ingredient: English-language data. For years, social media companies have focused their automatic content detection and removal efforts more on content in English than the world's 7,000 other languages.

Software giant Adobe has announced it will integrate generative AI into its widely used Photoshop program, while downplaying fears the move will lead to job losses and mass fakes. The brand most associated with image editing will incorporate the generative AI product Adobe Firefly, which launched as a beta six weeks ago, creating a tool the company says will become a "co-pilot" to graphic design rather than a replacement for humans. Using the "generative fill" feature, Photoshop users will be able to add to, expand or remove unwanted items from images using a text prompt similar to those used by Dall-E and Midjourney, such as "long haired dachshund with long flowing rainbow hair". The generative fill feature will be available in the desktop beta from Tuesday, with a wider release set for later in 2023. Adobe has been using AI in its tools for over a decade, such as the background replacement tool in Photoshop.

A fake image appearing to show a large explosion near the Pentagon was shared on social media on Monday prompting a brief dip in the stock market. Within minutes, a wave of social media accounts including some verified accounts shared the fake picture, further amplifying the confusion. Officials later confirmed that no such incident had occurred. Confident that this picture claiming to show an "explosion near the pentagon" is AI generated. Check out the frontage of the building, and the way the fence melds into the crowd barriers.