Apple is in negotiations with an increasingly demanding Indonesia to get its iPhone 16 ban lifted, while Huawei plans to shortly launch its smartphone in the country.Tim Cook (right) with photographer Sofyan Pratama in Indonesia in April 2024 image credit: AppleThe ban on sales of the iPhone 16 in Indonesia originally followed Apple's failure to meet its commitments to the country. Firms have to either source an unfeasibly large 40% of components from the country, or invest in development within the region, which is what Apple had committed to do.Based on previous reports, it appears that Apple's shortfall was around $10 million, and the company did then offer Indonesia that amount to basically make good. It was a token and the Indonesian government rejected it, making Apple up its offer to $100 million worth of investment. Continue Reading on AppleInsider | Discuss on our Forums

Six selected pavilions have been chosen for construction along the Nestaweya River Trail in Winnipeg as part of the latest Warming Huts: An Art + Architecture Competition on Ice. Since its launch in 2009, the open competition has attracted entries from across the globe, with winning schemes constructed towards the end of each January.The 2025 selection of huts seeks to represent fun and warmth, while also speaking to issues such as climate change and homelessness, according to the organizers. The winning schemes range from a colorful grain elevator to a balled-up beaver to a half-sunken car, and from a single-person home to a gift wrapped with a bow.Details on the six huts are expanded upon below, while our ongoing coverage of the initiative, including previous winners, can be explored here.ROSEMARY Skool by invited artists Jaimie Isaac and Suzanne Morrissette of ROSEMARY Gallery

During CES 2025, AMD has announced an update to its line-up of APUs, which meld high-power CPUs with strong graphics processing. Dubbed the Ryzen AI Max series of APUs, at the highest end, these new chips feature up to 16 CPU and 40 GPU cores.At the highest end, AMD has the Ryzen AI Max+ 395, which clock in at a maximum boost frequency of 5.1 GHz, has 16 CPU cores, 40 GPU cores, and a neural processing unit offering up to 50 trillion operations per second (TOPS). These APUs will be available starting in Q1 2025.On the lower end, the Ryzen AI Max PRO 380 will feature 6 CPU cores with 12 threads and 16 GPU cores, with a max boost frequency of 4.9 GHz. Filling out the line-up are the Ryzen AI Max and Max PRO 390 with 12 CPU cores and 32 GPU cores at a max boost of 5 GHz, the and the Ryzen AI Max and Max PRO 385, which features 8 CPU cores and 32 GPU cores at a max boost of 5 GHz. The only difference between the 390 and 385 is the amount of cache memory, with the 385 coming in at 40 MB while the 390 features 76 MB.These new APUs, developed under the codename Strix Halo, run on AMDs Zen 5 architecture for the CPUs, and RDNA 3.5 architecture for the GPUs. The APUs are available in the AI Max form, as well as the AI Max Pro form, the latter of which are designed specifically for business-class workstations.

During CES 2025, Asus has unveiled its new external GPU module that connects to a PC by using Thunderbolt 5. Dubbed the Asus ROG XG Mobile, the latest eGPU is set to offer power equivalent to a GeForce RTX 5090 mobile chip, and rather than using a proprietary cable like Asus previous eGPUs, will instead use Thunderbolt 5.The newest iteration of the Asus ROG XG Mobile also features a host of I/O options, including USB ports, an SD card reader, USB-C ports, and 5 Gbps ethernet connectivity. The XG Mobile will also offer up support for multiple monitors through DisplayPort 2.1 and HDMI 2.1 ports.While Asus has revealed features of the new XG Mobile, it hasnt yet announced raw performance numbers for the system. The system is priced at $2,199.99. Asus has also announced a similar system based on the GeForce RTX 5070 Ti, priced at $1,199.99.For context, these external GPUs are often targeted at users that would like to have as much power as possible in their PC, but cant justify a full tower desktop PC that would take up space and wouldnt be as portable when compared to a laptop. These eGPUs tend to be a great deal for users that would like to bolster their laptops hardware capabilities with more powerful graphics processing, and still manage to stay portable.The new Asus eGPUs make use of mobile versions of desktop GPUs recently unveiled by Nvidia under its new GeForce RTX 50-series lineup, which includes four new graphics cards that are slated to hit the market over the next couple of months.

AMD has unveiled its next generation of graphics cards during CES 2025. Running on the AMD RDNA 4 architecture, the new line-up of Radeon RX GPUs features a new naming convention, and will compete against Nvidias 40-series mid-range graphics cards. Interestingly, the company pointed out during its presentation that its new naming convention for the GPUs is meant to allow users to directly compare them with Nvidias offerings.Dubbed the Radeon RX 9070, RX 9070 XT, RX 9060 and RX 9060 XT, the graphics cards feature upgrades to their compute units over the previous RX 7000-series graphics cards. The chips in the cards are based on the 4nm process node, and will compete directly with the GeForce RTX 4070, RTX 4070 Ti, RTX 4060, and RTX 4060 Ti respectively.While AMD hasnt yet announced a release date for the new GPUs, the company and its partners will start releasing the graphics cards during Q1 2025. Unfortunately, AMD hasnt revealed too many details about the clock speeds and number of compute units in the new graphics cards. We do, however, know that the RX 9070 and its XT counterpart will feature 16 GB of GDDR6 VRAM, while the RX 9060 and 9060 XT will feature 8 GB of GDDR6 VRAM.Along with upgrades to hardware, the new GPUs will also make use of AMD FSR 4, which will allow high quality 4K upscaling, as well as other features like FSR Frame Generation, and AMDs Anti-Lag 2 technology. While games that support FSR 3.1 will also automatically support FSR 4, users will need a GPU running on RDNA 4 to make use of the FSR 4 features.For other graphics cards news, check out details of Nvidias newly-announced GeForce RTX 50-series GPUs.

Horizon Zero Dawn and Helldivers 2 movies announcedHorizon film to be produced in partnership with PlayStation Productions and Columbia Pictures News by Sophie McEvoy Staff Writer Published on Jan. 7, 2025 Sony has announced film adaptations of Horizon Zero Dawn and Helldivers 2.Revealed at the Consumer Electronics Show in Las Vegas yesterday, the Horizon movie will be produced in partnership with PlayStation Productions and Columbia Pictures, as reported by The Verge.The two studios previously collaborated with 2022's Uncharted film starring Tom Holland and Mark Wahlberg.The Helldivers 2 adaptation will be produced by Sony Pictures. No further details about the project have been announced.Sony previously announced a series adaptation of the Horizon franchise produced by Netflix in 2022.The series was reportedly cancelled last July, according to a report by Rolling Stone, which accused its showrunner Steve Blackman of "toxic, bullying, manipulative, and retaliatory behaviour."During the CES presentation, Sony also announced an anime adaptation of Ghost of Tsushima based on its multiplayer mode Legends.

Chris Kerr, News EditorJanuary 7, 20252 Min ReadImage via SonySony is developing a new anime series based on Ghost of Tsushima in partnership with Aniplex and Crunchyroll, both of which are owned by the PlayStation maker.The series will be called Ghost of Tsushima: Legends. As the name suggests, it will be based on the game's 'Legends' cooperative mode and incorporate elements from Japanese folktales and mythology.The series will be produced by PlayStation Pictures and Sony Music in collaboration with Aniplex, the studio behind Demon Slayer: Kimetsu no Yaiba, Solo Leveling, and Sword Art Online.It will premiere on anime streaming platform Crunchyroll in 2027.Sony acquired Crunchyroll in 2021 and merged it with Japanese anime distributor Funimation in a bid to create the "ultimate destination for anime fans." Aniplex was founded in 1995 by Sony Music Entertainment Japan.The project will become the first anime adaptation of a PlayStation property, but Sony is no stranger to the world of transmedia. Massive PlayStation franchises such as The Last of Us and Uncharted have already been turned into an award-winning HBO series and Tom-Holland-led-action-romp, respectively.Others such as Gran Turismo and Twisted Metal have received similar treatment. More are coming, too, with Sony today confirming that movies based on Horizon Zero Dawn and Helldivers 2 are currently in the works (thanks Eurogamer).Sony feels its Ghost of Tsushima anime is a testament to the "creative synergy" on display across its subsidiaries and the "versatility" of its video game properties."Having already proven the immense quality and versatility of our gaming properties across multiple successful film and television projects, we couldn't be more excited to announce our first ever anime adaptation," said Asad Qizilbash, head of PlayStation Productions, in a press release."Ghost of Tsushima's rich, immersive world and its fantastical Legends mode based on Japanese mythology provide the perfect canvas for this project, and Aniplex is the perfect partner to translate Sucker Punch Productions' hit video game into a stunning new anime series."Read more about:Transmedia[Company] PlayStationAbout the AuthorChris KerrNews Editor, GameDeveloper.comGame Developer news editor Chris Kerr is an award-winning journalist and reporter with over a decade of experience in the game industry. His byline has appeared in notable print and digital publications including Edge, Stuff, Wireframe, International Business Times, andPocketGamer.biz. Throughout his career, Chris has covered major industry events including GDC, PAX Australia, Gamescom, Paris Games Week, and Develop Brighton. He has featured on the judging panel at The Develop Star Awards on multiple occasions and appeared on BBC Radio 5 Live to discuss breaking news.See more from Chris KerrDaily news, dev blogs, and stories from Game Developer straight to your inboxStay UpdatedYou May Also Like

Google DeepMind is assembling a new team of artificial intelligence researchers to develop world models that can simulate physical environments. The initiative will be led by Tim Brooks, a former co-lead for OpenAIs Sora project who joined DeepMind in October to work on Googles video generation and world simulators.World models are a relatively new development within AI that could serve a variety of purposes, such as creating real-time interactive media environments for video games and movies, and realistic training scenarios for robots and other AI systems. Its also part of Googles push to achieve an artificial general intelligence system, or AGI, before its competitors.DeepMind has ambitious plans to make massive generative models that simulate the world, Brooks announced in an X post on Monday. Brooks included two open job listings for research engineers and scientists who will help to advance AI world models capable of simulating real-world scenarios by solving problems around training at massive scale, curating training data, and studying how they can be integrated with multimodal language models.We believe scaling pretraining on video and multimodal data is on the critical path to artificial general intelligence, DeepMind said in the job descriptions. World models will power numerous domains, such as visual reasoning and simulation, planning for embodied agents, and real-time interactive entertainment.The race to be the first to declare AGI is heating up, so Googles focus here isnt surprising. OpenAI CEO Sam Altman recently said that the company has cracked how to achieve the tech industrys long-sought benchmark, and that autonomous AI agents may start to meaningfully join workforces this year.There are several competitors who already have a head start over DeepMinds world modeling project, including Nvidias Cosmos platform for advancing physical AI, autonomous vehicle, and robot development; and the World Labs startup created by the godmother of AI, Fei-Fei Li.The new DeepMind team will work alongside existing Google AI projects including its flagship Gemini AI models, Veo video generator, and Genie Googles prior world model for simulating playable 3D environments in real-time.

Philips Hue users will soon be able to create their own AI-powered lighting scenes, the company announced on Tuesday.Philips Hues new generative AI lighting assistant lets users create personalized lighting scenes based on occasion, mood, or style. They can do so by typing commands like Give me a scene for a garden party into the Hue app and also by using their voice. The assistant will then either recommend scenes from the existing Philips Hue gallery or create a new one entirely. In addition to creating scenes, the AI assistant can also be used to adjust the settings of existing lights, including brightness levels and color. The assistant will recommend scenes based on prompts. Image: Philips HueThe news comes after Hueblog reported that Philips Hue was working on a generative AI feature in October. According to Hueblog, the AI Playground lets users create lighting scenes based on typed text like Create a scene for a childrens birthday party, but there was no mention of voice commands.Philips Hue isnt the first to offer an AI-powered lighting assistant. Govees AI Lighting Bot generates similar dynamic lighting effects and also responds to voice prompts. Meanwhile, Nanoleafs Magic Scenes feature creates similar scenes, too, but only responds to typed short phrases like tropical paradise.Philips Hue says the AI assistant will be compatible with all of its lights. The company hasnt divulged details about its exact availability, but stated itll start rolling out in the first quarter of 2025.RelatedThe AI-powered lighting assistant isnt the only piece of news Philips Hue announced on Tuesday. The Philips Hue Sync TV app now supports LG TVs, while thecolor-changing Philips HueDatura ceiling light is now available in the US starting at $299.99. The frameless ceiling panel offers a pair of individually controllable lights and customizable color effects.Philips Hue also announced new smart home security features, including smoke alarm sound detection and the ability to stream live camera feeds to Amazon Alexa and the Google Nest Hub. Users can also now check out live camera views or schedule system settings straight from their devices.

Dense embedding-based text retrieval has become the cornerstone for ranking text passages in response to queries. The systems use deep learning models for embedding text into vector spaces that enable semantic similarity measurements. This method has been adopted widely in applications such as search engines and retrieval-augmented generation (RAG), where retrieving accurate and contextually relevant information is critical. These systems efficiently match queries with relevant content by building on learned representations, driving huge advancements in knowledge-intensive domains.However, the main challenge for embedding-based retrieval systems is their susceptibility to manipulation by adversaries. The reason is that these systems often build on public corpora, which are not immune to adversarial content. Malicious actors can inject crafted passages into the corpus in a way that affects the retrieval systems ranking to prioritize the adversarial entries over the queries containing them. This can threaten the integrity of search results with the spread of misinformation or the introduction of biased content, endangering the reliability of knowledge systems.Previous approaches to counter adversarial attacks have used simple poisoning techniques, such as stuffing targeted queries with repetitive text or embedding misleading information. Although these methods can break single-query systems, they are often ineffective against more complex models that handle diverse query distributions. Existing defenses also do not address the core vulnerabilities in embedding-based retrieval systems, leaving the systems open to more advanced and subtle attacks.Researchers at Tel Aviv University introduced a mathematically grounded gradient-based optimization method called GASLITE for crafting adversarial passages. GASLITE performs better than previous techniques because it focuses precisely on the retrieval models embedding space rather than modifying content in the text. It aligns itself with certain query distributions, which results in adversarial passages achieving high visibility within retrieval results. Thus, this makes it a potent tool for evaluating vulnerabilities in dense embedding-based systems.The GASLITE methodology is grounded in rigorous mathematical principles and innovative optimization techniques. It constructs adversarial passages from attacker-chosen prefixes combined with optimized triggers designed to maximize similarity to targeted query distributions. Optimization takes the form of gradient calculations in the embedding space to find optimal token substitutions. Unlike previous approaches, GASLITE does not edit the corpus or model but instead focuses on generating text that the retrieval systems ranking algorithm can manipulate. This design makes it stealthy and effective; adversarial passages can blend directly into the corpus without being detectable by standard defenses.The authors test GASLITE with nine state-of-the-art retrieval models under various threat scenarios. The method consistently outperformed baseline approaches, achieving a remarkable 61-100% success rate in ranking adversarial passages within the top 10 results for concept-specific queries. These results were achieved with minimal poisoning of the corpus, with adversarial passages comprising just 0.0001% of the dataset. For example, GASLITE demonstrated top-10 visibility across most retrieval models when targeting concept-specific queries, showcasing its precision and efficiency. In single-query attacks, the method consistently ranked adversarial content as the top result, which is effective even under the most stringent conditions.Further analysis of the factors that contributed to the success of GASLITE showed that embedding-space geometry and similarity metrics significantly determined model susceptibility. Models using dot-product similarity measures were particularly vulnerable because the GASLITE method exploited these characteristics to achieve optimal alignment with targeted query distributions. The researchers further emphasized that models with anisotropic embedding spaces, where random text pairs produced high similarities, were more susceptible to attacks. This again points towards the importance of understanding embedding-space properties while designing retrieval systems.It underscores the need for strong defenses against adversarial manipulations in embedding-based retrieval systems. The authors thus recommend utilizing hybrid retrieval approaches like dense and sparse retrieval techniques that can minimize the risks provided by such methods as GASLITE. It serves, on its own, to expose the vulnerability in current retrieval systems to risks and pave the way for more secure and resilient technologies.The researchers urgently call to focus on the risks presented by such adversarial attacks to dense embedding-based systems. The minimal effort that GASLITE could use to manipulate search results shows the potential severity of such attacks. However, by characterizing critical vulnerabilities and developing actionable defenses, this work provides valuable insights into improving this robustness and reliability in retrieval models.Check out the Paper and GitHub Page. All credit for this research goes to the researchers of this project. Also,dont forget to follow us onTwitter and join ourTelegram Channel andLinkedIn Group. Dont Forget to join our60k+ ML SubReddit. Nikhil+ postsNikhil is an intern consultant at Marktechpost. He is pursuing an integrated dual degree in Materials at the Indian Institute of Technology, Kharagpur. Nikhil is an AI/ML enthusiast who is always researching applications in fields like biomaterials and biomedical science. With a strong background in Material Science, he is exploring new advancements and creating opportunities to contribute. [Recommended Read] Nebius AI Studio expands with vision models, new language models, embeddings and LoRA (Promoted)