Half-Life: Alyx update adds unassuming but impressive liquid effect to bottles

http://feedproxy.google.com/~r/venturebeat/SZYF/~3/HZm2qYgbvRA/

An amazing new detail arrives in the latest Half-Life: Alyx update: Bottles now have liquid in them.

No, seriously, it’s pretty impressive.

Update 1.4 adds incredibly convincing liquid effects to the many, many bottles strewn along the streets of City 17. Some Valve developers have been showing the effect in action, with liquid that realistically sloshes around as you twist and turn bottles in your hand. Plus, when a significant portion of your game is set in a vodka distillery, it really helps add to the immersion.

It’s just a small detail but it adds a lot. We can’t help but wonder if Valve is planning other similar additions to the game in the months ahead. What would you want to see?

Elsewhere this update adds new features to the recently launched Workshop feature for modding. Plus there are some fixes for a few cases of crashing. Fans around the globe will also be happy to see official subtitles for Brazilian, Czech, Hungarian, Italian, Polish, Portuguese, Ukrainian and Vietnamese.

Alyx remains one of our favorite VR games of the year and a towering achievement in triple-A VR design.

This story originally appeared on Uploadvr.com. Copyright 2020

New AI technique speeds up language models on edge devices

http://feedproxy.google.com/~r/venturebeat/SZYF/~3/qKicMC1LRoM/

Researchers at the MIT Computer Science and Artificial Intelligence Laboratory (CSAIL) and MIT-IBM Watson AI Lab recently proposed Hardware-Aware Transformers (HAT), an AI model training technique that incorporates Google’s Transformer architecture. They claim that HAT can achieve a 3 times inferencing speedup on devices like the Raspberry Pi 4 while reducing model size by 3.7 times compared with a baseline.

Google’s Transformer is widely used in natural language processing (and even some computer vision) tasks because of its cutting-edge performance. Nevertheless, Transformers remain challenging to deploy on edge devices because of their computation cost; on a Raspberry Pi, translating a sentence with only 30 words requires 13 gigaflops (one billion floating-point operations per second) and takes 20 seconds. This obviously limits the architecture’s usefulness for developers and companies integrating language AI with mobile apps and services.

The researchers’ solution employs neural architecture search (NAS), a method for automating AI model design. HAT performs a search for edge device-optimized Transformers by first training a Transformer “supernet” — SuperTransformer — containing many sub-Transformers. These sub-Transformers are then trained simultaneously, such that the performance of one provides a relative performance approximation for different architectures trained from scratch. In the last step, HAT conducts an evolutionary search to find the best sub-Transformer, given a hardware latency constraint.

HAT AI Trasnformers

HAT AI Trasnformers

VB Transform 2020 Online – July 15-17. Join leading AI executives:
Register for the free livestream.

To test HAT’s efficiency, the coauthors conducted experiments on four machine translation tasks consisting of between 160,000 and 43 million pairs of training sentences. For each model, they measured the latency 300 times and removed the fastest and slowest 10% before taking the average of the remaining 80%, which they ran on a Raspberry Pi 4, an Intel Xeon E2-2640, and an Nvidia Titan XP graphics card.

According to the team, the models identified through HAT not only achieved lower latency across all hardware than a conventionally trained Transformer, but scored higher on the popular BLEU language benchmark after 184 to 200 hours of training on a single Nvidia V100 graphics card. Compared to Google’s recently proposed Evolved Transformer, one model was 3.6 times smaller with a whopping 12,041 times lower computation cost and no performance loss.

“To enable low-latency inference on resource-constrained hardware platforms, we propose to design [HAT] with neural architecture search,” the coauthors wrote, noting that HAT is available in open source on GitHub. “We hope HAT can open up an avenue towards efficient Transformer deployments for real-world applications.”

Amazon discontinues the Echo Look and migrates AI style recommendations to other apps and devices

http://feedproxy.google.com/~r/venturebeat/SZYF/~3/OKpDSgBOJp0/

Amazon this week discontinued the Echo Look after pulling the Alexa-powered camera from store shelves in December 2019, a spokesperson confirmed to VentureBeat via email. As spotted by Voicebot.ai, Echo Look customers began receiving an email yesterday stating that their devices and the companion app will “no longer function” beginning July 2020, as the Echo Look’s capabilities migrate to other apps and services.

Style by Alexa, which suggests, compares, and rates apparel using a combination of AI and human curation, is available in the Amazon Shopping app after launching exclusively on the Echo Look. (It’s under the Programs and Features section in the left slide-out menu.) And now, asking any Alexa-enabled device “Alexa, what should I wear?” will prompt style recommendations based on the weather, shopping preferences, and other factors.

As Amazon explains, the Echo Look was always intended to help train the AI algorithms behind features like Style by Alexa, whose expanded launches come as the apparel market suffers pandemic-related declines. With style recommendations and programs like Prime Wardrobe, which lets users try on clothes and send back what they don’t want to buy, Amazon is vying for a larger slice of sales while algorithmically surfacing products customers might not normally choose. It’s a win for businesses on its face — excepting cases where the recommended accessories are Amazon’s own, of course.

Amazon Alexa Style recommendationsAmazon Alexa Style recommendations

Above: Alexa style recommendations from the Alexa mobile app.

Image Credit: Amazon

“When we introduced Echo Look three years ago, our goal was to train Alexa to become a style assistant as a novel way to apply AI and machine learning to fashion. With the help of our customers we evolved the service, enabling Alexa to give outfit advice and offer style recommendations,” a spokesperson told VentureBeat. “We look forward to continuing to support our customers and their style needs with Alexa.”

VB Transform 2020 Online – July 15-17. Join leading AI executives:
Register for the free livestream.

Amazon says Echo Look customers have been offered steep discounts on purchases of an Echo Show or Echo Show 5. The company is also offering free Amazon Photos accounts to store users’ Echo Look photos and videos through July 24, 2021, but those who don’t create an account by the deadline won’t be able to access their content.

Amazon Style by Alexa

Amazon Style by Alexa

Interestingly, the demise of the Echo Look — which launched in a closed, invite-only beta three years ago and was made available for general purchase in April 2017 — follows Facebook’s push toward an AI- and machine learning-driven ecommerce assistant. Last week, the social network detailed the algorithms behind its shopping experiences, including GrokNet, which can detect exact, similar (via related attributes), and co-occurring products across billions of photos; the company says GrokNet performs searches and filtering on Facebook Marketplace at least twice as accurately as the algorithm it replaced.

“We envision a future in which [a] system could … incorporate your friends’ recommendations on museums, restaurants, or the best ceramics class in the city — enabling you to more easily shop for those types of experiences,” Facebook wrote in a blog post. “Our long-term vision is to build an all-in-one AI lifestyle assistant that can accurately search and rank billions of products, while personalizing to individual tastes. That same system would make online shopping just as social as shopping with friends in real life. Going one step further, it would advance visual search to make your real-world environment shoppable. If you see something you like (clothing, furniture, electronics, etc.), you could snap a photo of it and the system would find that exact item, as well as several similar ones to purchase right then and there.”

The DeanBeat: Gamers count down the days to The Last of Us Part II

http://feedproxy.google.com/~r/venturebeat/SZYF/~3/nO2zWLVknfU/

The countdown to the launch of Sony’s The Last of Us Part II has gamers spellbound. It surged as a topic of conversation this week as Sony and Naughty Dog revealed more of the exclusive game coming to the PlayStation 4 on June 19.

The signs of fan anticipation are there. More than a million people watched the latest gameplay reveal, even though the full game leaked and anybody who really wants to know the full story can find it out there somewhere on the internet. I have a copy of the game for review purposes, but I am bound not to talk about that yet. But I have found a way to speak about how I’m feeling now.

I did not want to look at the leaked material, because I wanted to make my way to the ending on my own. It’s been an emotional journey, and I want to get there on my own terms. Playing this game in a pandemic, I cannot tell you what that’s been like, and I am eager to have conversations about it later.

If you’re a fan, you already get the emotional moment that is approaching as we soldier on in our fragile state, sheltering in place without many of our friends or family or coworkers around. Games are a salve, and I can think of nothing better for us than to lose ourselves in the narrative of the sequel.

VB Transform 2020 Online – July 15-17. Join leading AI executives:
Register for the free livestream.

If you’re not a fan, or you’re not a gamer, I have undertaken this story to explain why somebody you know is about to get obsessed with this game and talk about it incessantly. You can choose to ignore them or pretend you understand, but for the sake of your own safety, please do not tell them that their obsession is unjustified. That’s because the story in games like The Last of Us is as good as any you’ll come across. Forgive your friends if they drop out of Call of Duty: Warzone or Animal Crossing: New Horizons for a while.

The story so far

The Last of UsThe Last of Us

Above: The Last of Us featured a deep, evolving relationship between Ellie and Joel.

Image Credit: Naughty Dog/Sony

The Last of Us came out in 2013 on the PlayStation 3. It was a smash hit for developer Naughty Dog. I played the game and was deeply touched by the story of the teenage girl Ellie and gruff smuggler Joel — two survivors of the zombie apocalypse who spend their days just trying to survive. The graphic violence of the original was horrific, but more often than not it was perpetrated by Joel in the name of protecting Ellie, and later on it was Ellie protecting Joel. Fighting the zombies took skill and stealth, and too often the human enemies were worse.

I played the first game across 22 or so hours and declared it to be my favorite game of all time. It had what I call a bookend story, where the beginning of the game resonated at the ending and you had to think about the beginning of your journey just as it came full circle to the close. It was haunting, like experiencing a ghost story where the ghost returns to you in the end.

I spoke with the game directors, Neil Druckmann and Bruce Straley, about the experience after I finished it, and it was clear that it was one of Sony’s greatest hits of all time. While we had zombie games galore, The Last of Us was an intimate game when it came to characters, story, and combat. Each duel with a zombie or a human enemy was like a life-or-death struggle. You barely had enough bullets to get through a section of the game. You learned to feel like what it would be to exist as a survivor, Druckmann said.

Some 20 years after losing everything that was important to him, Joel was a shell of a human with very little humanity left in him, Druckmann told me. He might as well be a zombie. In his time with Ellie, she pulls that remaining humanity out. The 14-year-old girl becomes the hero, Druckmann said, by bringing the older man back to life. And then, at the end of the first game, Joel faced the moral dilemma of choosing whether to sacrifice someone he came to love as a surrogate daughter in the name of stopping the plague. It became a question of how far a father is willing to go to save his kid.

“At first, he’s willing to put his life on the line. That’s almost the easiest thing for him, where he’s at,” Druckmann said. “But then he’s willing to put his friends on the line. Finally, it comes to putting his soul on the line, when he’s willing to damn the rest of humanity. When he has that final lie with Ellie, he’s willing to put his relationship with Ellie on the line in order to save her.”

Joel made his choice, and the repercussions of that choice will play out in Part II. For the choice he made, many people might consider Joel to be a monster. But as a parent, I can understand the dilemma he faced.

Hitting home

Ellie and a fellow survivor in The Last of Us Part II.Ellie and a fellow survivor in The Last of Us Part II.

Above: The Last of Us Part II has finely detailed graphics.

Image Credit: Naughty Dog/Sony

I played this first game with my eldest daughter, and it was good to show her the character of Ellie at a time — even though it wasn’t all that long ago — when we didn’t see so many female protagonists who were normal people (not superheroes) and could handle their own. Back in 2013, and certainly before that, it could be a commercial risk to launch a game with a female protagonist. That is no longer true.

Ellie was a role model for my kids, and my second daughter avidly played through all of The Last of Us as well, as I watched and advised. I think a father and a daughter playing this game together could understand the bond between Ellie and Joel, even if the characters didn’t see it themselves.

They were surprised at the game’s popularity and even taken aback at the criticism that the game had strong female characters. The enemies of diversity in games would later develop a lexicon for this, saying the developers were “woke.”

Druckmann said the intention was always there to have a story with dual protagonists, each with strong arcs that affected the other. It was not just a teenage girl being escorted on a mission by a middle-aged man. As my daughters played, I felt it was good for them to have the agency of being Ellie, as she fought back during her crucible moments.

But this would not be such a strange thing today. We have a lot of strong female characters and improving diversity in game characters. But back in 2013, this was something original to have strong characters who were men, women, black, white, gay, and straight. They were fleshed out, but Naughty Dog took heat for being on a soapbox. Part II adds to the diversity with multiple Asian characters, lesbian relationships, and people with very different points of view than Ellie or Joel, as we have seen in the trailers.

Part II

Above: Naughty Dog shared this scene in its latest trailer for The Last of Us Part II.

Image Credit: Naughty Dog/Sony

As I saw the extended trailer for Part II this week, I felt a new appreciation for the difference between the old game and the new game in terms of technological improvements. The story spans seasons and climates, from the snows of Jackson, Wyoming, to the lush landscapes of Seattle. The cities are so authentic, as you would expect them to look after a pandemic.

As the trailer showed, Naughty Dog wants you to know that Ellie can do more. She can swing on ropes, traverse vertical structures to avoid trouble, navigate boats, ride horses, break glass, and crawl through grass. She faces enemies such as dogs who can trace her footsteps, stealth warriors who can attack her with arrows, and large numbers of zombies. Ellie can sprint, dodge attacks, and time her counterattacks. She can use enemies as a shield, and she can get help from her friends. This makes combat far more diverse than in The Last of Us.

And as she could do in the first game, Ellie can pit enemies against each other, making zombies attack human enemies. Druckmann said that this is Naughty Dog’s largest, most ambitious game.

As you delve into the development story of this game, you’ll come across allegations of crunch, or heavy overtime, at the Naughty Dog staff. Depending on how you feel about that, you may come to resent the company or the development leaders. You are entitled to feel what you want about that. But I do think you should play it, and not dismiss the creative work because of how it was created.

I do not mean to justify crunch at all. But I think that I can understand why perfectionists would work so hard on such a game. So much passion goes into it, they want to keep working and do. Others wonder if they have to do the same, and then a lot of people start doing the same whether they want to or not. It’s something for managers to contemplate. I hope the industry solves this challenge.

And I know I’m lucky to be playing this first. But it’s a bummer not being able to talk about it yet. I trust Naughty Dog more than any other studio when it comes to telling a story. As I finish the game, I’ll have plenty of time to soak it in and figure out what I think about it. For now, we have plenty of anxiety to go around as we all wait to see what happens in The Last of Us Part II.