philip lelyveld The world of entertainment technology

Everything Google announced at I/O today

Android 12, 3D video calls, Wear OS updates, dev tools and more

...

AI conversations get more conversational with LaMDA

It’s a little too easy to stump AIs if you go off script, asking something in a way that to you seems normal but to the language model is totally incomprehensible. Google’s LaMDA is a new natural language processing technique that makes conversations with AI models more resilient to unusual or unexpected queries, making it more like a real person and less like a voice interface for a search function. They demonstrated it by showing conversations with anthropomorphized versions of Pluto and a paper airplane. And yes, it was exactly as weird as it sounds.

Google built a futuristic 3D video calling booth

...

See the full story here: https://techcrunch.com/2021/05/18/everything-google-announced-at-i-o-today/?guccounter=1&guce_referrer=aHR0cHM6Ly93d3cuZ29vZ2xlLmNvbS8&guce_referrer_sig=AQAAAInE1tf01sSKFJiYuJlReP-1LQ0_aEfedKbChXeaIw4AKiHgtiJ2yOqn8WdhYc7TQe6E6p14ou7DpU2gZOTI-8dLL5ZmBZofIDxnO3F27mR8mkmc3rF_uczvLlOpGQxMe02ymdF1JSBEdmmr9d3EOfYE6GE6HuNFJWQMzYKpirHr&fbclid=IwAR2FkBjBYpFO-0wQjI2-Ms1UMUQDQQujfb7JV0tmViYYNCFXHTiKPj3V3RI

(Indian actress deepfakes herself into a role) Make way, Rakhi Sawant is now eyeing Hollywood!

Actress-dancer Rakhi Sawant never fails to keep us hooked to her antics, be it on a reality show or in real life.In a quirky video posted on social media, she gave a glimpse into what would be her Hollywood look.Deepfaking her own face in a Hollywood movie, the 42-year-old presented an eerie gift to her fans on Saturday.Let's check it out!

See the full story here: https://www.newsbytesapp.com/news/entertainment/rakhi-sawant-deepfakes-her-face-in-hollywood-movies/story

ETC Executive Coffee: Verizon on the Future of Entertainment

To kick off ETC@USC’s five-part Spring 2021 Executive Coffee with… series, Serhad Doken, executive, innovation and product realization at Verizon, led a Zoom discussion on March 15. “Expectations for the Future of Entertainment and UI/UX design” was the framing question for a wide-ranging discussion. Six students from USC’s School of Cinematic Arts, Viterbi School of Engineering, and the Iovine and Young Academy participated. Topics included user experience expectations, VR and 5G, “gaming” the algorithm, recommendations versus serendipity, artists earning a living, and what experiences have yet to be adopted.

See the full story and 9 minute highlight reel here: https://www.etcentric.org/etc-executive-coffee-verizon-on-the-future-of-entertainment/

Facebook aims to help AI models forget with Expire-Span

By forgetting, AI will be able to retain more information at scale because it will be able to memorize the most critical items.

The company introduced a tool called Expire-Span, which enables AI models, to forget. Today, AI models including neural networks memorize information and data without distinction. By forgetting, AI will be able to retain more information at scale because it will be able to memorize the most critical items.

Expire-Span evaluates each piece of information in a model and predicts what will be more important to return. The training data is based on context as the model learns. 

In a blog post, Facebook explained that processing information indiscriminately can work on a small scale. But once models take on more data the computational costs surge. 

Aside from the computation and storage savings, Expire-Scan processes can enable models to optimize over time. 

Facebook said Expire-Scan was inspired by the way humans retain memories.

See the full story here: https://www.zdnet.com/article/facebook-aims-to-help-ai-models-forget-with-expire-span/

The unlikely success of DuckDuckGo suggests a way to escape surveillance capitalism.

Last year, the company’s traffic more than doubled. It has done this with no creepy surveillance: All it does is use whatever keywords you type in the search bar—“best inkjet printer,” “Boston hotels”—to customize an ad for that search. This is known as “contextual” targeting, distinct from the secret-police “behavioral” tracking that fuels advertising on many tech platforms and creates a mammoth dossier of your online activity. DuckDuckGo doesn’t even retain your search info. Every time you load the search engine, you’re a stranger.

Indeed, DuckDuckGo’s success suggests, more trenchantly, that a lot of Silicon Valley’s business argument about data harvesting is flat-out wrong. They say they need to do it to produce compelling products: Personalizing their wares helps keep us maximally “engaged,” and thus rakes in advertising money. Yet here’s a tech firm that avoided practicing surveillance capitalism; it just practices regular capitalism.

Even some ad buyers are questioning whether endless tracking works; a survey by Digiday found that 45 percent of ad execs saw “no significant benefit” from behavioral tracking, and 23 percent found it made revenues decline.

See the full story here: https://www.wired.com/story/tech-companies-dont-need-to-be-creepy-to-make-money/

Patriot One Technologies Announces Accuracy Enhancements to Video Recognition Software for Weapons and Threat Detection

Patriot One Technologies Inc. (TSX: PAT) (OTCQX: PTOTF) (FRANKFURT: 0PL) (“Patriot One” or the “Company”) today announced enhancements to its Video Recognition Software (VRS) for identifying weapons and other threats in crowds. Patriot One’s VRS is an advanced artificial intelligence (AI)-based video analytics solution that provides unprecedented accuracy in detecting threats and enabling businesses, schools and other organizations to protect employees, students, customers and assets. Businesses can also use its capabilities to gain valuable intelligence on customer traffic flow and behavior to optimize operations.

With the ability to accurately detect weapons and disturbances of up to 150 feet away, Patriot One’s VRS dramatically improves security and efficiency for organizations that host crowds or the general public, including retail stores, schools, casinos, entertainment venues and more.

“False positives are a huge problem with other AI-based video analysis systems, which is why their alerts need to be reviewed and validated by humans before any action can be taken. This squanders critical minutes in the incident detection and response cycle,” said Peter Evans, CEO of Patriot One. “VRS automatically validates alerts through multiple camera sources and a combination of AI analytic engines, so security staff can be correctly notified and briefed in real-time before a weapons incident or other type of disturbance arises.”

...

Key enhancements to VRS include:

High Fidelity Object Recognition: The system accurately detects guns while automatically distinguishing them from umbrellas, phones or other innocuous items. Its accuracy enables the system to notify security personnel at the earliest possible point in the incident cycle, so they can de-escalate potentially violent situations quickly.

People Counting: VRS Counting provides valuable intelligence both for crowd management and business operations, ...

See the full story here: https://www.streetinsider.com/Globe+Newswire/Patriot+One+Technologies+Announces+Accuracy+Enhancements+to+Video+Recognition+Software+for+Weapons+and+Threat+Detection/18414013.html

U.S. Army Unveils Enhanced Night Vision Tech With Higher Resolution, Augmented Reality

This new technology combines a night vision binocular (which has more features than a goggle) with a clip-on thermal imager, as well as augmented reality technology. That means they can improve sight in the day, detect targets in very low light, bad weather and even through smoke.

See the full story here: https://www.newsweek.com/us-army-unveils-enhanced-night-vision-tech-higher-resolution-augmented-reality-1591395

Oscars, Golden Globes drive streaming activity amid controversy

Reelgood found that Nomadland, having won the most Oscars this year (and all the PR that comes with that achievement) saw 4.1x more streams the day after director Chloé Zhao and actress/producer Frances McDormand’s historic Oscar wins.

Mank, meanwhile, enjoyed a much stronger boost in viewing after news of its 10 Oscar nominations broke compared to when it actually won two of those noms.

For TV, the team looked at the lift for shows that were not as popular as the biggest nominated series such as Schitt’s Creek, The Queen’s Gambit, and The Crown.

See the full story here: https://blog.reelgood.com/oscars-golden-globes-streaming-controversy?mkt_tok=ODUwLVRBQS01MTEAAAF9Cc1JWbhNcnkutGXXU0v8liHeG4l3BHatz11COhiGHF-vP1vC_Dt6uEwt_Rmlucwd6ng61wy0eutkzf-dmWyHDqPMKfKVa728YypT0pippM7c

Deloitte Digital and Snap Inc. Report Reveals the Rich, Untapped Future of Augmented Reality for Customer Experience

With 74% of global consumers expecting augmented reality (AR) to be even more important in their lives in the next five years, AR has the power to redefine a brand's digital customer experience and mobile engagement

Deloitte Digital and Snap Inc. have conducted one of the largest and most comprehensive global "Consumer Augmented Reality" reports to further understand the potential of this technology, the findings of which affirm AR's role as the next frontier of mobile and digital engagement and identify how AR can help brands elevate their customer experience. 

The report, which evaluated responses from 15,000 U.S. domestic and international consumers, about half of which use Snapchat, shows that AR has the power to impact a business' bottom line, not only driving customer connections but significantly driving conversions. AR has made the move to utility: no longer seen as just a toy (used to try out hairstyles or send playful videos between friends), but now as a meaningful engagement tool for brands, where consumers can, for example, virtually try on beauty products or envision a piece of furniture in their home. This growth in AR usage represents how virtual commerce is becoming more accessible, ultimately changing the shopping experience for consumers. 

"As AR evolves, it will revolutionize our lives and become as significant of a technology shift as the web or mobile was to society, changing how we view and interact with the world around us," said Allan Cook, managing director, Deloitte Consulting LLP, and Digital Reality Business leader at Deloitte Digital. "The report shows that AR is really only in the beginning stages of being integrated in our everyday activities and is poised to make the journey from our cameras today to smart glasses tomorrow."

See the Snap report here:https://forbusiness.snapchat.com/blog/the-next-inflection-point-more-than-100-million-consumers-are-shopping-with-ar

See the full story here: https://www.prnewswire.com/news-releases/deloitte-digital-and-snap-inc-report-reveals-the-rich-untapped-future-of-augmented-reality-for-customer-experience-301290445.html

MIT artificial intelligence tech can generate 3D holograms in real-time

Now, MIT researchers have developed a new way to produce holograms almost instantly — and the deep learning-based method is so efficient that it can run on a laptop in the blink of an eye, the researchers say.

Shi believes the new approach, which the team calls "tensor holography," will finally bring that elusive 10-year goal within reach. 

...a hologram encodes both the brightness and phase of each light wave. That combination delivers a truer depiction of a scene's parallax and depth.

...

Shi's team took a different approach: letting the computer teach physics to itself.

They used deep learning to accelerate computer-generated holography, allowing for real-time hologram generation. The team designed a convolutional neural network — a processing technique that uses a chain of trainable tensors to roughly mimic how humans process visual information.

In mere milliseconds, tensor holography can craft holograms from images with depth information — which is provided by typical computer-generated images and can be calculated from a multicamera setup or LiDAR sensor (both are standard on some new smartphones). 

This advance paves the way for real-time 3D holography. What's more, the compact tensor network requires less than 1 MB of memory. "It's negligible, considering the tens and hundreds of gigabytes available on the latest cell phone," he says.

The research "shows that true 3D holographic displays are practical with only moderate computational requirements," says Joel Kollin, a principal optical architect at Microsoft who was not involved with the research.

 Currently, most affordable consumer-grade displays modulate only brightness, though the cost of phase-modulating displays would fall if widely adopted.

The work was supported, in part, by Sony.

See the full story here: https://bigthink.com/technology-innovation/3d-holograms