Ubisoft Unveils Ghostwriter: A Tool That Automatically Generates Lines For Game Characters Using AI

By

-

March 24, 2023

In the world of video games, open-world games have become increasingly popular due to their level of immersion. One of the essential elements of an open-world game is the chatter by non-playable characters (NPCs) in the game. These dialogues, known as “bark,” are typically heard in the game’s background, adding a layer of realism to the game’s environment. However, writing barks can be time-consuming and labour-intensive for game developers.

To address this issue, Ubisoft, the video game company responsible for popular open-world game series like Assassin’s Creed and Watch Dogs, has developed a machine-learning tool called Ghostwriter. The tool generates the first draft of barks, reducing the burden on game developers. Ubisoft has emphasized that Ghostwriter is not intended to replace video game writers but rather to assist them in their work by freeing up time devoted to other core plot elements.

According to Ubisoft, crowd chatter and barking are crucial features of player immersion in games, making the game world feel more real and separate from the player’s actions. However, creating barks can be a time-consuming and challenging task for writers.

Ghostwriter allows screenwriters to create a character and the type of dialogue they wish to generate. The tool then generates variations from which the scriptwriter can select and edit. The process uses a pairwise approach to evaluate and improve the generated variations. For each generated variation, the tool offers two options for the screenwriter to compare and choose from. Once the screenwriter chooses one, the tool learns from the chosen alternative. With thousands of choices made by screenwriters, the tool becomes more effective and accurate over time.

Ben Swanson, the research and development director at Ubisoft, emphasizes that technologies like Ghostwriter require screenwriters to learn how to use the tools and incorporate them into the video game production process. Swanson’s team’s goal is to make this AI power available to scenario designers so they can create machine learning models tailored to their design needs. To achieve this goal, Ubisoft has developed a tool called Ernestine that allows anyone to easily create machine-learning models for Ghostwriter.

Google Maps will Introduce Immersive 3D Views of Cities Soon

By

Adwaith

-

March 24, 2023

Google Maps is rolling out one of its most impressive updates in years — a 3D view of some of the most important cities on the planet that almost appears real.

The immersive 3D view was announced at last year’s Google I/O, but the deployment has been very slow, making it one of the slowest deployments of Google Maps. However, a Reddit user has discovered that the feature is now available in cities such as London and Berlin, so it is likely that the feature will soon be available in other parts of the world, including Spain.

This immersive view offers a 3D perspective of the most important areas of cities, allowing users to contemplate the best perspectives of cities they have not yet visited and get a better idea of the urban decoration before actually setting foot in them. The view is available after navigating to a localized area and looking at the photo gallery for that location. The label “Immersive View” appears, and clicking on it enables users to experience the view of 3D graphics that increases the vision almost as if it were real over a point in a city.

One advantage of the new immersive view of Google Maps is that it allows users to enter very popular buildings, such as Madison Square Garden in New York, via an option on the left side of the screen. This feature provides data on weather and traffic, making the experience more realistic and giving Google Maps a new perspective from which to view the most visited sites in said cities.

However, there is a downside to this view in Google Maps — it seems to consume a lot of data. Using this new function of the maps app for 30 minutes consumed 2 GB of data, which can be counterproductive if users are under a limited amount of mobile data.

NVIDIA Announces H100 NVL: A New GPU Variant for Large Language Models

By

Adwaith

-

March 23, 2023

NVIDIA has announced the H100 NVL, a new variant of its Hopper GPU specifically designed for Large Language Models (LLMs) like OpenAI’s GPT-4.

The dual GPU card has two H100 PCIe boards already bridged together and a large memory capacity, making it the most memory per GPU within the H100 family and of any NVIDIA product to date.

The H100 NVL is aimed at a single market of large-scale language model usage on the bandwagon and to further NVIDIA’s AI success. Large language models like the GPT family are, in many ways, constrained in memory capacity. Even the H100 accelerator fills up quickly to hold all parameters (175B for the largest GPT-3 model). Therefore, NVIDIA has developed new H100 SKUs with more memory per GPU than the regular H100 parts that cap at 80GB per GPU.

All GH100 GPUs come with six stacks of HBM memory (HBM2e or HBM3) with 16GB per stack capacity. However, due to yield concerns, NVIDIA only ships regular H100 parts with five of the six HBM stacks enabled, which means that each GPU has a nominal 96GB of VRAM, but the regular SKUs only have 80GB available.

The H100 NVL will be a special SKU with all six stacks enabled, offering 14GB of memory per GH100 GPU, a 17.5% memory increase. The dual GPU/dual card H100 NVL looks like the SXM5 version of the H100 arranged on a PCIe card. Regular H100 PCIe is being held back somewhat by slower HBM2e memory usage, fewer active SM/tensor cores usage, and lower clock speeds, but Tensor Cores NVIDIA is presenting for H100 NVL. The performance numbers are all on par with the H100 SXM5, indicating that this card isn’t further reduced like regular PCIe cards.

The H100 NVL’s big feature is its large memory capacity, as the dual GPU card offers 188GB of HBM3 memory (94GB per card). It also has a total memory bandwidth of 7.8TB/s, and 3.9TB/s on individual boards, making it the most memory per GPU within the H100 family and of any NVIDIA product to date.

The provision of this SKU is probably in anticipation of the explosive hit of ChatGPT and the rapid increase in AI demand due to the subsequent appearance of Bing and Bard. The price will go up accordingly, but given the booming LLM market, there will be demand to pay a high enough premium for a near-perfect GH100 package.

In conclusion, the H100 NVL is a new variant of NVIDIA’s Hopper GPU specifically designed for Large Language Models (LLMs) like OpenAI’s GPT-4. Its large capacity makes it the most memory per GPU within the H100 family and of any NVIDIA product. Its increased memory capacity will help to further NVIDIA’s AI success. The H100 NVL is a significant step forward in the LLM market, and it will be interesting to see how it performs when compared to other GPUs in the future.

Nothing Unveils Ear (2) Wireless Headphones with Adaptive ANC and Hi-Res Audio Streaming Support

By

Sabarinath

-

March 23, 2023

OnePlus founder Carl Pei’s company, Nothing, has unveiled its latest product — the Ear (2) wireless headphones. While the design looks the same as the first-generation Ear (1) earphones, Ear (2) boasts several improvements in terms of components and capabilities.

One significant upgrade is the LHDC 5.0 audio codec support, allowing for Hi-Res audio streaming. The Ear (2) also features a new personalized sound profile that allows you to calibrate the headphones to your specific hearing after taking a quick test in the companion app, Nothing X. Additionally, you can customize the equalizer settings to your preference.

Ear (2) also has adaptive active noise cancellation that filters out unwanted noise up to 40dB based on your ear profile and surroundings. The headphones also have updated 11.6mm drivers with a new diaphragm made from graphene and polyurethane and improved wind protection for clearer voice capture in challenging environments.

Battery life has also been improved, with up to 36 hours with the charging case and 6 hours of playtime on the earbuds alone. The charging case can be charged via USB-C or wirelessly at 2.5W, and the headphones support Google Fast Pair for Android and Microsoft Swift Pair for Windows. The Ear (2) is also rated IP55 dust and splash-resistant, while the earbuds are IP54.

Ear (2) is priced at $149 or €149, the same price as Ear (1) after its price increase. The headphones are available for pre-order starting today, with sales set to begin on March 28. Ear (2) only comes in white, unlike Ear (1), available in black and white.

Despite looking similar to Ear (1), the Ear (2) packs several improvements, making it a worthy upgrade for those seeking high-quality wireless headphones.

Valve Announces Counter-Strike 2 as Free Upgrade: What to Expect from the Source 2 Engine and Other Innovations

By

Vishak

-

March 23, 2023

Valve has announced that Counter-Strike: Global Offensive (CS:GO) will be upgraded to Counter-Strike 2 (CS2) this summer, with the technical switch to the Source 2 engine promised to be the greatest technical advance in the history of Counter-Strike. The game will be available as a free upgrade to existing CS:GO players, with revised maps, new features, and improvements to visual and audio effects.

Valve has not yet revealed the full scope of the innovations in CS2; the upgrades will include audio, maps, and visual effects revisions. In addition, the game will introduce dynamic smoke grenades and gameplay independent of the tick rate. A player’s CS:GO items will be transferred to CS2.

According to Valve, all visual effects have been polished, and new effects such as explosions, fire, smoke, and water have been added thanks to the lighting and particle systems of the Source 2 engine. The user interface has also been completely revised. The sound effects have been retuned to better reflect the physical environment and create a more enjoyable listening experience.

Valve describes the use of the new smoke grenades as a revolutionary change. The smoke can interact with the environment, so bullets and frag grenades can throw up smoke obscuring the player. The smoke also spreads naturally and pulls through broken windows or downstairs.

The tick rate, which refers to the intervals at which the server calculates the game situation, will no longer affect movements, shots, or throws. Regardless of the tick rate, movements and shots will react at the same speed, and grenades will fly and land the same way.

Instead of new maps, CS2 will feature revised versions of existing maps, with changes ranging from new lighting and improved readability to physics-based rendering. “Overpass” will receive a general overhaul with even bigger changes. Comparison images are provided on the Counter-Strike 2 official website.

A limited test phase for CS2 has already started, with selected players able to try out the new game based on criteria such as recent playtime on Valve’s official servers. The testers will be notified via the CS: GO main menu.

The Ethics of Parental Control Apps: Balancing Protection and Trust

By

Jenna Jose

-

March 23, 2023

As our lives become more intertwined with technology, the need for parental control apps has become increasingly apparent. With the rise of social media and the internet, parents are looking for ways to protect their children from the dangers of the online world. However, the use of parental control apps like Bark, Qustodio, Google Family Link, and eyeZy raises ethical questions about trust and privacy.

On the one hand, parental control apps can provide a sense of security and control for parents. These apps allow parents to monitor their children’s online activity, restrict access to certain websites, and even limit screen time. This can be especially important for younger children who may not fully understand the risks associated with the internet.

However, the use of parental control apps can also be seen as an invasion of privacy. Children may feel like they are being watched or controlled, which can damage the trust between parents and children. In addition, parental control apps can be circumvented by tech-savvy children, which can lead to a false sense of security for parents.

One of the main ethical concerns surrounding parental control apps is the balance between protection and trust. While parents may have good intentions in monitoring their children’s online activity, they also need to respect their children’s privacy and autonomy. This means finding a balance between protecting children from the dangers of the online world and allowing them to explore and learn on their own.

Another ethical concern is the potential for abuse of parental control apps. While these apps can be useful for monitoring children’s online activity, they can also be used to control or manipulate them. For example, parents may use these apps to punish or reward their children based on their online behavior, which can lead to a toxic environment of control and fear.

To address these ethical concerns, it is important for parents to have open and honest conversations with their children about their use of parental control apps. Children should be made aware of the reasons behind the use of these apps and should be included in discussions about what is and is not acceptable online behavior. This can help to build trust and respect between parents and children, while still allowing for a level of protection.

Another way to address these ethical concerns is to use parental control apps in a way that promotes responsible and safe online behavior. For example, instead of using these apps to simply restrict access to certain websites, parents can use them to teach children about online safety and how to navigate the online world in a responsible and respectful way.

The use of parental control apps raises important ethical questions about trust and autonomy. While these apps can be useful for protecting children from the dangers of the online world, it is important to find a balance between protection and trust. By having open and honest conversations with children about their use of these apps and promoting responsible online behavior, parents can help to ensure that their children are safe and secure online, while also respecting their independence and privacy.

Meta Explores New Advertising Possibilities on Instagram with Ads in Search Results and Reminder Ads

By

Avinash A

-

March 22, 2023

New Advertising Possibilities on Instagram

Meta is testing new ways to increase advertising revenue on its popular platform. As the global economy slows down, the demand for advertising becomes more apparent. The company has introduced two new tools to Instagram to explore new possibilities.

One of the tactics is the inclusion of advertisements in Instagram search results. When users tap a post from the search results, this advertising will show in a scrollable feed. According to Instagram, the ad reaches users actively looking for a brand, product, or content. Meta intends to publish the advertising globally in the following months after the initial testing phase.

The “reminder ads” tool is the second new tool. This tool enables companies to publicise, remind, and alert users about upcoming events and product releases in which they may be interested. Instagram users may set reminders and get three alerts. This application will assist businesses in creating excitement for upcoming events. The adverts will be available in all advertisers’ feeds.

These new technologies will assist Meta in increasing income when ad revenues are decreasing. According to Meta’s fourth-quarter 2022 financial results, advertising revenue declined to $31.2 billion from $32.6 billion in the same period the previous year.

In addition, the company is introducing a new ad format for Facebook Reels, which will allow marketers to post advertising on the Explore homepage and profile feed.

NVIDIA Releases RTX Path Tracing SDK To Integrate Path Tracing Technology Into Games

By

Adwaith

-

March 22, 2023

NVIDIA has released the RTX Path Tracing SDK, a tool for integrating path-tracing technology into video games. Path tracing is a sophisticated variant of traditional ray tracing, simulating light’s physical behaviour as closely as possible to the real world. It requires more graphics resources than conventional ray tracing technology.

Path tracing technology is commonly used in three-dimensional architectural projects or projects for the development of the design of future products when maximum accuracy and realism of image rendering are required. However, it has hardly been used in games and has only been implemented in Quake RTX and Portal RTX so far.

It should be noted that the RTX Path Tracing SDK release does not offer full integration of the technology into games. NVIDIA sees two options for using the technology in game projects. Firstly, it can be used to create a reference path tracer that ensures the “reality” of lighting in games during their development. Secondly, path-tracing technology can create high-quality photo modes in games.

The Path Tracing SDK uses video cards based on the Ada Lovelace architecture. Cyberpunk 2077 will soon receive path-tracing support, according to NVIDIA. The release of the Path Tracing SDK indicates that the GeForce RTX 40 series will not be retired anytime soon.

While the benefits of path tracing in games remain unclear, NVIDIA believes it can be useful for creating high-quality visuals and improving the overall realism of game lighting. With the RTX Path Tracing SDK, game developers now have the tools to explore the potential of path-tracing technology and its use in future games.