Artificial intelligence (AI) has already enabled the creation of the'last Beatles song', Now and Then, which raced to the top of the charts this week. Filmmaker Peter Jackson used an AI tool called'machine audio learning' (MAL) to isolate John Lennon's voice from an old 1970s home demo. The vocal performance – rendered'crystal clear' by the AI – was then complemented by new instrumentation from Paul McCartney and Ringo Starr, along with guitar recorded by George Harrison for the song in 1995. But MAL's work may not be finished, as it could be used to salvage the last filmed interview of John Lennon, recorded less than two months before his tragic murder. Long to the frustration of fans, much of Lennon's answers to questions in the historically priceless clip are drowned out by – somewhat bizarrely – the sound of the first Star Wars movie.

From computer-generated "extras" to AI "zombies," new restrictions against the use of artificial intelligence in Hollywood were set out by the actors' union Friday. The Screen Actors Guild (SAG-AFTRA) reached a deal with studios like Disney and Netflix this week to end its nearly four-month strike. Its board members on Friday voted 86% in favor of ratifying the agreement. Besides a 7% minimum pay increase, and a new $40 million-per-year fund to transfer a portion of revenues for hit shows from studios to actors, AI guardrails were a key part of talks.

Netflix's new prestige sci-fi show is delayed until March 22, 2024. The new trailer gives us our first look at the series' key "video game," Three-Body, which involves a nebulous and extremely shiny VR headset. According to John Bradley's character Jack Rooney, the headset has "no screen... no headphone jack... not even a charging port." Donning the headset transports Rooney to a hyper-realistic world, before he's swiftly ejected and the trailer ends. The show's source material is The Three-Body Problem, the first novel in Liu Cixin's Remembrance of Earth's Past series.

In this study, we investigate in-context learning (ICL) in document-level event argument extraction (EAE). The paper identifies key challenges in this problem, including example selection, context length limitation, abundance of event types, and the limitation of Chain-of-Thought (CoT) prompting in non-reasoning tasks. To address these challenges, we introduce the Heuristic-Driven Link-of-Analogy (HD-LoA) prompting method. Specifically, we hypothesize and validate that LLMs learn task-specific heuristics from demonstrations via ICL. Building upon this hypothesis, we introduce an explicit heuristic-driven demonstration construction approach, which transforms the haphazard example selection process into a methodical method that emphasizes task heuristics. Additionally, inspired by the analogical reasoning of human, we propose the link-of-analogy prompting, which enables LLMs to process new situations by drawing analogies to known situations, enhancing their adaptability. Extensive experiments show that our method outperforms the existing prompting methods and few-shot supervised learning methods, exhibiting F1 score improvements of 4.53% and 9.38% on the document-level EAE dataset. Furthermore, when applied to sentiment analysis and natural language inference tasks, the HD-LoA prompting achieves accuracy gains of 2.87% and 2.63%, indicating its effectiveness across different tasks.

Users interact with text, image, code, or other editors on a daily basis. However, machine learning models are rarely trained in the settings that reflect the interactivity between users and their editor. This is understandable as training AI models with real users is not only slow and costly, but what these models learn may be specific to user interface design choices. Unfortunately, this means most of the research on text, code, and image generation has focused on non-interactive settings, whereby the model is expected to get everything right without accounting for any input from a user who may be willing to help. We introduce a new Interactive Text Generation task that allows training generation models interactively without the costs of involving real users, by using user simulators that provide edits that guide the model towards a given target text. We train our interactive models using Imitation Learning, and our experiments against competitive non-interactive generation models show that models trained interactively are superior to their non-interactive counterparts, even when all models are given the same budget of user inputs or edits.

On Wednesday night, the actor Jeremy Allen White, of "The Bear," was working his way down a red carpet in Dallas. It was the première of "The White Claw," an A24 movie about the Von Erich clan of professional wrestlers. On the carpet, an "Entertainment Tonight" reporter informed White, "We just heard moments ago--the strike is over!" and stuck the mike in his face. "That's amazing," White said, seeming taken aback. Asked how he felt, he added, "I don't know the details of the deal, but I'm sure that SAG got what we wanted."

A restaurant in a rural Oregon city couldn't find enough servers to stay fully staffed. So the owner hired a robot named Plato. She had no idea how much pushback she'd get from the community. A San Francisco smoothie startup promised to create customers "one-of-a-kind" recipes, but two months after opening, the store was shuttered. The shop, BetterBlends, used artificial intelligence to generate custom smoothies based on customer preferences.

Do you ever find yourself captivated by a song but struggling to put into words what makes it so special? Have you ever wanted to identify the instrument or genre of a piece of music but found yourself at a loss? Perhaps you've tried to search for a particular song through text, only to hit a dead end in your quest. In the world of music information retrieval, the tasks of transcribing music scores and retrieving music based on its characteristics are critical areas of research and advanced techniques may help you sometimes. However, for everyday music enthusiasts without formal training, achieving these goals in pre-defined scientific terms can often feel elusive.

The thread was 10 tweets long--verbose by X standards--and 219 words, but there was just one word that stuck out. The message, posted on the @SAG-AFTRA account, summed up everything the actors union had fought to get in the tentative agreement with Hollywood studios. In the context of the rapid rise of generative AI, it's worth reading in full: "We have achieved a deal of extraordinary scope that includes'above-pattern' minimum compensation increases, unprecedented provisions for consent and compensation that will protect members from the threat of AI, and for the first time establishes a streaming participation bonus." "Threat," of course, is how many people have come to view artificial intelligence. US president Joe Biden's recent executive order on the technology was seen as, in part, a way to address the risks the technology presents to national security.

Early in the pandemic, an agent--literary, not software--suggested Fei-Fei Li write a book. She has made an indelible mark on the field of artificial intelligence by heading a project started in 2006 called ImageNet. It classified millions of digital images to form what became a seminal training ground for the AI systems that rock our world today. Li is currently the founding codirector of Stanford's Institute of Human-Centered AI (HAI), whose very name is a plea for cooperation, if not coevolution, between people and intelligent machines. Accepting the agent's challenge, Li spent the lockdown year churning out a draft.