Deepfake video of Taylor Swift speaking Mandarin sparks discussion over AI in China
... In a video posted on Chinese social media platform Weibo, the 33-year-old singer can be seen flaunting her Mandarin in what looks to be a talk show.
“Recently, I’ve been to many places, like Italy, France and Japan,” Swift said in Mandarin with a slight American twang. ...
The deepfake videos were generated by an AI tool from Chinese start-up HeyGen, reported news outlet The China Project.
The lip movements were synced to make it look like she was speaking in Mandarin, and it sounded like Swift’s voice, making the videos so realistic that it led Chinese netizens to express their amazement at the technology and leave comments such as: “This is awesome.” ...
Its AI-powered video generator allows users to create text-to-speech videos in over 300 voices in more than 40 languages, using over 100 AI avatars with different ethnicities, ages, poses and clothes. ...
See the full story here: https://www.straitstimes.com/asia/east-asia/deepfake-video-of-taylor-swift-speaking-mandarin-sparks-discussion-over-ai-in-china
Pokémon Go creator John Hanke says it’s not game over for the app
... "I'm most excited about those that don't cut you off from other people," says John.
John says he's taking a particular interest in mixed reality devices, such as Instagram owner Meta's new headsets and Apple's Vision Pro, launching next year.
"Those are the kinds of devices that we're leaning into," he says.
"Because they're devices you can use when you're out for a walk and with your friends, and they don't get in your way".
John predicts that "2024 is going to be a breakthrough year for AR", especially as it converges with artificial intelligence (AI).
He says Niantic is looking to deploy this in Pokémon Go to deepen the interplay between its real and virtual elements. ...
The latest, called Party Play, focuses on something Pokémon Go has always done well - bringing people together.
Trainers, as players call themselves, can team up with three friends and see them on the same screen as they go out completing missions together.
It mirrors the lively community scene that's built up around the game, with events still regularly taking place that bring fans on joint walks around their local areas. ...
See the full story here: https://www.bbc.com/news/newsbeat-67226243
Virtual Influencers: Meet the AI-Generated Figures and Your New Online “Friends”
... Virtual influencers are transforming the way content is created, consumed and marketed online. They represent an electrifying dance between cutting-edge technology and our desire for connection. But, at the same time, they are yet another product being peddled by marketers that want our money. ...
There are three main types of virtual influencers: non-humans, animated humans and life-like CGI humans. Each one provides an innovative way to connect with audiences. ...
While real influencers with millions of followers may demand hundreds of thousands of dollars per post, one 2020 estimatesuggested virtual influencer Lil Miquela charged a more reasonable £6,550 (currently about US$7,965).
Virtual influencers have clear benefits when it comes to online engagement and marketing. They don’t age, they’re free from (real) scandals and they can be programmed to speak any language. ...
One major concern is transparency. Many virtual influencers already present as human-like, and it may become increasingly difficult to distinguish between them and real people. This is particularly problematic in an advertising context. ...
Similarly, TikTok has updated its community guidelines to say: “Synthetic or manipulated media that shows realistic scenes must be clearly disclosed. This can be done through the use of a sticker or caption, such as “synthetic,” ‘fake’, ‘not real’, or ‘altered.’” ...
On one hand, celebrity deepfake porn is on the rise. On the other, celebrities are including “simulation rights” in their contracts so their likeness may be used in the future. Take global football star Lionel Messi, who allowed PepsiCo to use a digital version of him to promote Lay’s potato chips. ...
Sphere: The Future of Immersive Entertainment Wins Prestigious Judges Award for Creativity and Innovation
Sphere, a groundbreaking next-generation entertainment medium in Las Vegas, has been honored with the prestigious Judges Award for Creativity and Innovation at the HPA Awards for 2023. The award recognizes Sphere’s revolutionary technology and visionary design, which have redefined the entertainment experience and storytelling, taking audiences to new heights of immersion and engagement. ...
This year’s esteemed jury, co-chaired by Carolyn Giardina and Joachim Zell, comprised notable members Paul Debevec, Jay Holben, Joanne Kim, and Karen Raz. ...
See the full story here: https://hpaonline.com/sphere-the-future-of-immersive-entertainment-wins-prestigious-judges-award-for-creativity-and-innovation/
What Is Retrieval-Augmented Generation (RAG)?
... The bad news is that the information used to generate the response is limited to the information used to train the AI, often a generalized LLM. The LLM’s data may be weeks, months, or years out of date and in a corporate AI chatbot may not include specific information about the organization’s products or services. ...
That’s where retrieval-augmented generation (RAG) comes in. RAG provides a way to optimize the output of an LLM with targeted information without modifying the underlying model itself; that targeted information can be more up-to-date than the LLM as well as specific to a particular organization and industry. That means the generative AI system can provide more contextually appropriate answers to prompts as well as base those answers on extremely current data.
RAG first came to the attention of generative AI developers after the publication of “Retrieval-Augmented Generation for Knowledge-Intensive NLP Tasks,” a 2020 paper published by Patrick Lewis and a team at Facebook AI Research. The RAG concept has been embraced by many academic and industry researchers, who see it as a way to significantly improve the value of generative AI systems. ...
See the full story here: https://www.oracle.com/artificial-intelligence/generative-ai/retrieval-augmented-generation-rag/
Joe Hunting, Director of ‘We Met in Virtual Reality,’ Launching Virtual Reality Production Company Painted Clouds
Joe Hunting, director of the HBO film “We Met In Virtual Reality,” is starting a production company called Painted Clouds, focusing exclusively on virtual reality filmmaking. The company aims to create original films and series using real time VR, including the popular platform VRChat. ...
Hunting’s “We Met In Virtual Reality,” which premiered at the Sundance Film Festival in 2022, is a documentary showcasing online connection through virtual reality, also filmed exclusively in VR. Taking place amidst the Covid-19 pandemic, the film illustrates the nature of virtual expression and interaction whilst people were physically isolated. Variety‘s Peter Debruge called Hunting’s documentary “a warmhearted, often humorous look at the sociology of such spaces.” ...
See the full story here: https://variety.com/2023/digital/news/joe-hunting-vr-painted-clouds-we-met-in-virtual-reality-1235769359/
From 65 to 103, Older Adults Experience Virtual Reality Differently Depending on Their Age: Evidence From a Large-scale Field Study in Nursing Homes and Assisted Living Facilities
Abstract
There is growing interest in applications of virtual reality (VR) to improve the lives of older adults, but the limited research on older adults and VR largely treats older adults as a monolith, ignoring the substantial differences across 65 to 100+ year olds that may affect their experience of VR. There are also few existing studies examining the experiences and challenges facing those who facilitate VR for older adults (e.g., caregiving staff). We address these limitations through two studies. In Study 1, we explore variation within older adults’ experiences with VR via a field study of VR use among a large (N = 245) and age-diverse (M age = 83.6 years old, SD age = 7.9, range = 65 - 103 years old) sample of nursing home and assisted living facility residents across 10 US states. Age was negatively associated with the extent to which older adults enjoyed VR experiences. However, the negative relationship between age and older adults’ attitudes toward VR was significantly less negative than the relationship between age and their attitudes towards other technologies (cell phones and voice assistants). In Study 2, we surveyed caregiving staff (N = 39) who facilitated the VR experiences for older adult residents and found that the caregiving staff generally enjoyed the activity relative to other activities and felt it to be beneficial to their relationship with residents.
Read the full research paper here: https://papers.ssrn.com/sol3/papers.cfm?abstract_id=4570743
What Did Ancient Egyptian Art Look Like in Its Glory Days? The Louvre and Snapchat’s New A.R. Program Offers a Glimpse
Snap, owner of the video messaging app Snapchat, has partnered with the Louvre on a free program called Egypt Augmented, which will give users an interactive deep dive into the museum’s renowned Egyptian treasures using augmented reality (A.R.).
The experience, designed by Snap’s own A.R. Studio, leverages the technology to bring iconic works like the Chamber of Ancestors and the Dendera Zodiac to life in dynamic ways. Viewers scan the QR codes posted near the chosen artworks, open up Snapchat’s camera, and see these ancient artworks in a new light. The technology will virtually restore faded paint, revealing “shapes, materials, colors, and decorations” of selected works that have faded to bare stone over the years, according to the press release.
See the full story here: https://news.artnet.com/art-world/louvre-museum-snapchat-egyptian-treasures-2382519
LinkedIn Tests Generative AI to Field Cybersecurity Questions From Employees and Suppliers
... Response times to employees’ questions on the chatbot were found to be about five seconds or less, which compares with around 15 minutes when a human helper responded, according to a spokeswoman for the company. Early tests show the chatbots are around 90% accurate. ...
“Security has to be a business differentiator, security has to be part of your strategic plan to grow your business,” he said. Belknap said he is looking into other ways generative AI could help his cybersecurity team defend the company against hackers, such as by detecting malware. ...
When the final version of the chatbot is released, LinkedIn plans to track how many of its interactions with employees require little to no human engagement, according to the spokeswoman for the company. ...
See the full story here: https://www.wsj.com/articles/linkedin-tests-generative-ai-to-field-cybersecurity-questions-from-employees-and-suppliers-d61d35a6
OpenAI Developing ‘Provenance Classifier’ for GenAI Images
OpenAI is developing an AI tool that can identify images created by artificial intelligence — specifically those made in whole or part by its Dall-E 3 image generator. Calling it a “provenance classifier,” company CTO Mira Murati began publicly discussing the detection app last week but said not to expect it in general release anytime soon. This, despite Murati’s claim it is “almost 99 percent reliable.” That is still not good enough for OpenAI, which knows there is much at stake when the public perception of artists’ work can be impacted by a filter applied by AI, which is notoriously capricious. ...
TechCrunch writes that in addition to achieving something closer to 100 percent accuracy across the board, OpenAI is also concerned about “the philosophical question of what, exactly, constitutes an AI-generated image.”
Artwork wholly generated by DALL-E 3 is an obvious example, “but what about an image from DALL-E 3 that’s gone through several rounds of edits, has been combined with other images and then was run through a few post-processing filters?” muses TechCrunch. ...
The usefulness of an OpenAI AI image detector optimized for Dall-E 3 may be limited if it cannot spot pictures generated by competing technologies like Midjourney, Stable Diffusion, and Firefly, Digital Trends points out, adding that “anything that can highlight fake images could have a positive impact.” ...
See the full story here: https://www.etcentric.org/openai-developing-provenance-classifier-for-genai-images/
Pages
- About Philip Lelyveld
- Mark and Addie Lelyveld Biographies
- Presentations and articles
- Trustworthy AI – A Market-Driven approach
- Tufts Alumni Bio