philip lelyveld The world of entertainment technology

HAI at Davos: Key Insights on AI From the World Economic Forum

...  James Landay: Professor of Computer Science and the Anand Rajaraman and Venky Harinarayan Professor in the School of Engineering; Co-Director, Stanford HAI

...the need for large foundation models that reflect global cultural diversity. Currently, AI models are predominantly built by U.S. companies with an emphasis on English and Western content. That creates cultural biases and even mistranslations. We need culturally and linguistically diverse training data in AI development. ...

DeepSeek was getting in people’s heads during Davos, but really, the company just used several known (and developed some new) clever optimizations for both training time and inference time compute. Now there’s a lot of skepticism over exactly how much money and how many GPUs were required, but it’s clear that they were able to train really great models much more efficiently. And now these are techniques that everyone else can use because it’s open source and in published technical reports. So that, again, leads to way more efficiency. ...

Alex (Sandy) Pentland: Toshiba Professor of Media Arts and Sciences; Professor of Information Technology; Media Lab Entrepreneurship Program Director, MIT Management Sloan School; Center Fellow, Stanford HAI; Faculty Lead of Digital Platforms and Society at the Digital Economy Lab

... People seemed to be girding themselves for challenges — geopolitical tensions, tariff wars, real wars, climate disasters — by a change in focus from idealistic goals and dreams to making social systems work and social contracts sustainable. ...

A major theme I think is emerging is what you might call the third way: not the US-EU, not China, but the way of India, Eastern Africa, the Middle East and Indopacific. These are middle income countries, no longer poor, and with sophisticated technical populations, and they are busy deploying digital technologies everywhere, including all but the most cutting-edge AI. ...

Erik Brynjolfsson: Director, Stanford Digital Economy Lab; Jerry Yang and Akiko Yamazaki Professor and Senior Fellow, Stanford HAI; Ralph Landau Senior Fellow, Stanford Institute for Economic Policy Research 

... I think it's a healthy pivot to start focusing more on identifying the specific tasks where AI can be helpful. Ultimately that will lead not only to more business value, but also to more productivity, better healthcare, a cleaner environment, and a more prosperous society.

See the full post here: https://hai.stanford.edu/news/hai-davos-key-insights-ai-world-economic-forum

The irony — using generative AI in a case about the dangers of generative AI

... The underlying lawsuit challenged the constitutionality of a Minnesota state statute aimed at curbing the use of deepfakes to influence elections by creating criminal penalties for the dissemination of AI-generated content 90 days before an election. ...

The Court excluded the declaration after finding that the expert's unchecked use of generative AI to create the declaration "shatters [the expert's] credibility" and rendered the declaration unreliable. The Court noted that "signing a declaration under penalty of perjury is not a mere formality" and that it could not "accept false statements — innocent or not — in an expert's declaration submitted under penalty of perjury." ...

See the full story here: https://www.reuters.com/legal/legalindustry/irony-using-generative-ai-case-about-dangers-generative-ai-2025-01-30/

What DeepSeek’s breakthrough says (and doesn’t say) about the ‘AI race’ with China

...

Such arguments emphasize the need for the United States to outpace China in scaling up the compute capabilities necessary to develop artificial general intelligence (AGI) at all costs, before China “catches up.” This has led some AI companies to convincingly argue, for example, that the negative externalities of speed-building massive data centers at scale are worth the longer-term benefit of developing AGI. Such an argument has significant business upside for AI companies, as they amass greater numbers of chips to gain a competitive advantage. What the DeepSeek example illustrates is that this overwhelming focus on national security—and on compute—limits the space for a real discussion on the tradeoffs of certain governance strategies and the impacts these have in spaces beyond national security.

To plug this gap, the United States needs a better articulation at the policy level of what good governance looks like. This should include a proactive vision for how AI is designed, funded, and governed at home, alongside more government transparency around the national security risks of adversary access to certain technologies. It also requires the US government to be clear about what capabilities, technologies, and applications related to AI it is specifically aiming to regulate.  ...

See the full story here: https://www.atlanticcouncil.org/blogs/new-atlanticist/what-deepseeks-breakthrough-says-and-doesnt-say-about-the-ai-race-with-china/

New Vatican document examines potential and risks of AI

...

On privacy and control, the Note points out that some types of data can go so far as to touch “upon the individual’s interiority, perhaps even their conscience” [90], with the danger of everything becoming “a kind of spectacle to be examined and inspected” [92]. Digital surveillance “can also be misused to exert control over the lives of believers and how they express their faith” [90]. ...

See the full PR here: https://www.vaticannews.va/en/vatican-city/news/2025-01/new-vatican-document-examines-potential-and-risks-of-ai.html

Augmented Reality and Projection Mapping Technologies

One significant area of research is the development of dynamic projection mapping techniques that can adapt to changing scenes and surfaces. For instance, a study introduced a method for dynamic multi-projection mapping that utilizes real-time pixel-parallel calculations to control projector intensity with low latency, achieving impressive performance at 360 frames per second (fps) with minimal delay[1]. This advancement allows for seamless integration of projections onto moving or changing surfaces, enhancing the overall user experience. ...

See the full story here: https://www.nature.com/research-intelligence/augmented-reality-and-projection-mapping-technologies

Vuzix introduces new AR platforms and ultra-thin waveguide technology

... Another innovation is an ultra-thin full-color waveguide with a thickness of just 1.0 millimeter. This uses the company's proprietary Incognito technology to optimize AR display performance. Vuzix also presented new display technologies. These included MicroLEDs and ultra-compact full-color LCoS projectors designed for use in next-generation AI-powered devices.

About a year ago, the company also launched the developer edition of its Z100 smart glasses. Weighing just 36 grams, they are among the lightest smart glasses with a display. Designed for industrial environments, the headset has a monocular display with a 30-degree field of view and a battery life of up to two days. ...

See the full story here: https://mixed-news.com/en/vuzix-introduces-new-ar-platforms-and-ultra-thin-waveguide-technology/

Year of the Full-Length AI Movie

...

The Heist follows a thief on the run driving a green muscle car through an urban hellscape trying to evade the cops, and it stoked a lot of passionate conversation from the millions who saw it. Unlike the movies whose run times are pushing three hours, though, The Heistclocks in at 1:56 — as in one minute, 56 seconds.

The short is shockingly good. It’s dynamic, fast-paced — and yes, made with AI. Three things make The Heist worthy of deeper consideration: It was made with Google’s Veo 2 generative video model, an unheralded entrant into the AI video wars that debuted on Dec. 16. It was created by Jason Zada, a veteran Hollywood filmmaker who directed the 2016 horror film The Forest, which grossed almost $41 million globally. Zada is now the founder of an AI film studio called Secret Level. And the film has a level of polish that should make Hollywood take notice. ...

See the full story here; https://theankler.com/p/year-of-the-full-length-ai-movie-2025-heist-jason-zada?utm_source=substack&utm_medium=email

Watch The Heist here: https://www.instagram.com/jasonzada/reel/DD5hBbZS6B8/?hl=en

Davos 2025 – in the heated debate around AGI, will no-one think about where Godzilla got his fish?

... In the 1940s and 50s, there was a lot of concern about nuclear weapons, and then, of course, we got Godzilla. There are some real concerns about nuclear weapons. They are dangerous. It's also led to some of the greatest peace in mankind ever, because of people's fear of using them. But I'll leave you with this with respect to Godzilla - where did Godzilla get all the fish? That is an organism that is massive. There's no way that that thing could exist. You have to look at the physics. AI requires energy, and each token that comes out of one of these systems is one to three joules. So if intelligence started going out of control, you would notice it would use a lot of energy.

While everyone digested that thesis, Choi made an attempt at a more grounded pitch for a middle ground:

We need more investment on scientific understanding about gen AI. The fact that we know how to create that doesn't mean that we actually understand it. The fact that you can give life to humans doesn't mean that we actually understand how the the body actually works, or mental health works. We still have a lot to find out. Similarly, the fact that we know how to create gen AI doesn't mean that we know how to control it, or we know how things work. We do not know the limits of it. ...

See the full story here: https://diginomica.com/davos-2025-heated-debate-around-agi-will-no-one-think-about-where-godzilla-got-his-fish

Robotic hand helps pianists overcome “ceiling effect”

...

“I’m a pianist, but I [injured] my hand because of overpracticing,” coauthor Shinichi Furuya of Kabushiki Keisha Sony Computer Science Kenkyujo told New Scientist. “I was suffering from this dilemma, between overpracticing and the prevention of the injury, so then I thought, I have to think about some way to improve my skills without practicing.” Recalling that his former teachers used to place their hands over his to show him how to play more advanced pieces, he wondered if he could achieve the same effect with a robotic hand.

So Furuya et al. used a custom-made exoskeleton robot hand capable of moving individual fingers on the right hand independently, flexing and extending the joints as needed.  ...

A total of 118 pianists participated in three different experiments. In the first, 30 pianists performed a designated "chord trill" motor task with the piano at home every day for two weeks: first simultaneously striking D and F keys with the right index and ring fingers, then striking the E and G keys with the right middle and little fingers. "We used this task because it has been widely recognized as technically challenging to play quickly and accurately," the authors explained. ...

After two weeks, they were assigned to one of two laboratory groups, in which the robot exoskeleton hand passively moved those four fingers for 30 minutes to perform the chord trill—faster than the players could do so themselves. ...

The results: "Even when the skill plateaued after weeks of piano practice, passive training of the fast and complex motor skill with the robot further facilitated the maximum rate of repetitive piano keystrokes involving fast and complex multifinger movements," the authors wrote, and the training effect also showed up in the untrained hand, so there was an "inter-manual transfer effect." ...

See the full story here: https://arstechnica.com/science/2025/01/robotic-hand-helps-pianists-overcome-ceiling-effect/

What is ‘authentic’ music in the age of AI?

....

The Recording Industry Association of America and other industry groups have advocated for stronger regulations to prevent the unauthorized use of artists’ voices and likenesses in AI-generated content. This push is not just about protecting individual artists but also preserving the unique qualities that make each performer special and feel inclined to create their own art. ...

In addition to the legal challenges, there is also a philosophical debate about what constitutes ‘authentic’ music. As AI continues to improve, it becomes harder to distinguish between human-created and AI-generated songs. While some view this as an exciting frontier for creativity, others worry that it may devalue human artistry, turning unique voices into easily replicable templates. This anxiety is particularly pronounced in genres like hip-hop and blues, where AI-generated content has sometimes been accused of cultural appropriation or misrepresentation. ...

See the full story here: https://www.thetriangle.org/entertainment/ai-in-music-industry/