Write Warz is an upcoming party game centered around communal creative writing. Boltz Entertainment describes how it came together.Read More

After a public callout, the developers of Hades took to social media to clarify that they are not intending to recast union voice actors.Read More

Cory Barlog and Neil Druckmann had a remarkable conversation at the Dice Summit about their approaches to creativity.Read More

As the games industry moves on from survive til 25, 2024 games investment and M&A data revealed recovery and improvement, DDM said.Read More

Join our daily and weekly newsletters for the latest updates and exclusive content on industry-leading AI coverage. Learn MoreWhen DeepSeek-R1 first emerged, the prevailing fear that shook the industry was that advanced reasoning could be achieved with less infrastructure.As it turns out, thats not necessarily the case. At least, according to Together AI, the rise of DeepSeek and open-source reasoning has had the exact opposite effect: Instead of reducing the need for infrastructure, it is increasing it.That increased demand has helped fuel the growth of Together AIs platform and business. Today the company announced a $305 million series B round of funding, led by General Catalyst and co-led by Prosperity7. Together AI first emerged in 2023 with an aim to simplify enterprise use of open-source large language models (LLMs). The company expanded in 2024 with the Together enterprise platform, which enables AI deployment in virtual private cloud (VPC) and on-premises environments. In 2025, Together AI is growing its platform once again with reasoning clusters and agentic AI capabilities.The company claims that its AI deployment platform has more than 450,000 registered developers and that the business has grown 6X overall year-over-year. The companys customers include enterprises as well as AI startups such as Krea AI, Captions and Pika Labs.We are now serving models across all modalities: language and reasoning and images and audio and video, Vipul Prakash, CEO of Together AI, told VentureBeat.The huge impact DeepSeek-R1 is having on AI infrastructure demandDeepSeek-R1 was hugely disruptive when it first debuted, for a number of reasons one of which was the implication that a leading edge open-source reasoning model could be built and deployed with less infrastructure than a proprietary model.However, Prakash explained, Together AI has grown its infrastructure in part to help support increased demand of DeepSeek-R1 related workloads.Its a fairly expensive model to run inference on, he said. It has 671 billion parameters and you need to distribute it over multiple servers. And because the quality is higher, theres generally moredemand on the top end, which means you need more capacity.Additionally, he noted that DeepSeek-R1 generally has longer-lived requests that can last two to three minutes. Tremendous user demand for DeepSeek-R1 is further driving the need for more infrastructure.To meet that demand, Together AI has rolled out a service it calls reasoning clusters that provision dedicated capacity, ranging from 128 to 2,000 chips, to run models at the best possible performance.How Together AI is helping organizations use reasoning AIThere are a number of specific areas where Together AI is seeing usage of reasoning models. These include:Coding agents: Reasoning models help break down larger problems into steps.Reducing hallucinations: The reasoning process helps to verify the outputs of models, thus reducing hallucinations, which is important for applications where accuracy is crucial.Improving non-reasoning models: Customers are distilling and improving the quality of non-reasoning models.Enabling self-improvement: The use of reinforcement learning with reasoning models allows models to recursively self-improve without relying on large amounts of human-labeled data.Agentic AI is also driving increased demand for AI infrastructureTogether AI is also seeing increased infrastructure demand as its users embrace agentic AI.Prakash explained that agentic workflows, where a single user request results in thousands of API calls to complete a task, are putting more compute demand on Together AIs infrastructure.To help support agentic AI workloads, Together AI recently has acquired CodeSandbox, whose technology provides lightweight, fast-booting virtual machines (VMs) to execute arbitrary, secure code within the Together AI cloud, where the language models also reside. This allows Together AI to reduce the latency between the agentic code and the models that need to be called, improving the performance of agentic workflows.Nvidia Blackwell is already having an impactAll AI platforms are facing increased demands.Thats one of the reasons why Nvidia keeps rolling out new silicon that provides more performance. Nvidias latest product chip is the Blackwell GPU, which is now being deployed at Together AI.Prakash said Nvidia Blackwell chips cost around 25% more than the previous generation, but provide 2X the performance. The GB 200 platform with Blackwell chips is particularly well-suited for training and inference of mixture of expert (MoE) models, which are trained across multiple InfiniBand-connected servers. He noted that Blackwell chips are also expected to provide a bigger performance boost for inference of larger models, compared to smaller models.The competitive landscape of agentic AIThe market of AI infrastructure platforms is fiercely competitive.Together AI faces competition from both established cloud providers and AI infrastructure startups. All the hyperscalers, including Microsoft, AWS and Google, have AI platforms. There is also an emerging category of AI-focussed players such as Groq and Samba Nova that are all aiming for a slice of the lucrative market.Together AI has a full-stack offering, including GPU infrastructure with software platform layers on top. This allows customers to easily build with open-source models or develop their own models on the Together AI platform. The company also has a focus on research developing optimizations and accelerated runtimes for both inference and training.For instance, we serve the DeepSeek-R1 model at 85 tokens per second and Azure serves it at 7 tokens per second, said Prakash. There is a fairly widening gap in the performance and cost that we can provide to our customers.Daily insights on business use cases with VB DailyIf you want to impress your boss, VB Daily has you covered. We give you the inside scoop on what companies are doing with generative AI, from regulatory shifts to practical deployments, so you can share insights for maximum ROI.Read our Privacy PolicyThanks for subscribing. Check out more VB newsletters here.An error occured.

Join our daily and weekly newsletters for the latest updates and exclusive content on industry-leading AI coverage. Learn MoreVery small language models (SLMs) can outperform leading large language models (LLMs) in reasoning tasks, according to a new study by Shanghai AI Laboratory. The authors show that with the right tools and test-time scaling techniques, an SLM with 1 billion parameters can outperform a 405B LLM on complicated math benchmarks.The ability to deploy SLMs in complex reasoning tasks can be very useful as enterprises are looking for new ways to use these new models in different environments and applications.Test-time scaling explainedTest-time scaling (TTS) is the process of giving LLMs extra compute cylces during inference to improve their performance on various tasks. Leading reasoning models, such as OpenAI o1 and DeepSeek-R1, use internal TTS, which means they are trained to think slowly by generating a long string of chain-of-thought (CoT) tokens.An alternative approach is external TTS, where model performance is enhanced with (as the name implies) outside help. External TTS is suitable for repurposing exiting models for reasoning tasks without further fine-tuning them. An external TTS setup is usually composed of a policy model, which is the main LLM generating the answer, and a process reward model (PRM) that evaluates the policy models answers. These two components are coupled together through a sampling or search method.The easiest setup is best-of-N, where the policy model generates multiple answers and the PRM selects one or more best answers to compose the final response. More advanced external TTS methods use search. In beam search, the model breaks the answer down into multiple steps. For each step, it samples multiple answers and runs them through the PRM. It then chooses one or more suitable candidates and generates the next step of the answer. And, in diverse verifier tree search (DVTS), the model generates several branches of answers to create a more diverse set of candidate responses before synthesizing them into a final answer.Different test-time scaling methods (source: arXiv)What is the right scaling strategy?Choosing the right TTS strategy depends on multiple factors. The study authors carried out a systematic investigation of how different policy models and PRMs affect the efficiency of TTS methods.Their findings show that efficiency is largely dependent on the policy and PRM models. For example, for small policy models, search-based methods outperform best-of-N. However, for large policy models, best-of-N is more effective because the models have better reasoning capabilities and dont need a reward model to verify every step of their reasoning.Their findings also show that the right TTS strategy depends on the difficulty of the problem. For example, for small policy models with fewer than 7B parameters, best-of-N works better for easy problems, while beam search works better for harder problems. For policy models that have between 7B and 32B parameters, diverse tree search performs well for easy and medium problems, and beam search works best for hard problems. But for large policy models (72B parameters and more), best-of-N is the optimal method for all difficulty levels.Why small models can beat large modelsSLMs outperform large models at MATH and AIME-24 (source: arXiv)Based on these findings, developers can create compute-optimal TTS strategies that take into account the policy model, PRM and problem difficulty to make the best use of compute budget to solve reasoning problems.For example, the researchers found that a Llama-3.2-3B model with the compute-optimal TTS strategy outperforms the Llama-3.1-405B on MATH-500 and AIME24, two complicated math benchmarks. This shows that an SLM can outperform a model that is 135X larger when using the compute-optimal TTS strategy.In other experiments, they found that a Qwen2.5 model with 500 million parameters can outperform GPT-4o with the right compute-optimal TTS strategy. Using the same strategy, the 1.5B distilled version of DeepSeek-R1 outperformed o1-preview and o1-mini on MATH-500 and AIME24.When accounting for both training and inference compute budgets, the findings show that with compute-optimal scaling strategies, SLMs can outperform larger models with 100-1000X less FLOPS.The researchers results show that compute-optimal TTS significantly enhances the reasoning capabilities of language models. However, as the policy model grows larger, the improvement of TTS gradually decreases.This suggests that the effectiveness of TTS is directly related to the reasoning ability of the policy model, the researchers write. Specifically, for models with weak reasoning abilities, scaling test-time compute leads to a substantial improvement, whereas for models with strong reasoning abilities, the gain is limited.The study validates that SLMs can perform better than larger models when applying compute-optimal test-time scaling methods. While this study focuses on math benchmarks, the researchers plan to expand their study to other reasoning tasks such as coding and chemistry.Daily insights on business use cases with VB DailyIf you want to impress your boss, VB Daily has you covered. We give you the inside scoop on what companies are doing with generative AI, from regulatory shifts to practical deployments, so you can share insights for maximum ROI.Read our Privacy PolicyThanks for subscribing. Check out more VB newsletters here.An error occured.

StreamElements latest report shows the success of Marvel Rivals, which jumped to the top 10 most-watched games on Twitch.Read More

Carter Lipscomb, a game industry veteran with a big personality, has passed away.Read More

Patching shouldn't be the action item teams get to when other higher-priority tasks are completed. It's core to keeping a business alive.Read More

Why enterprises must embrace an AI-first strategy that unifies endpoint, identity and network security within a zero-trust framework.Read More

NYU Langone has built an LLM research companion and medical advisor, and is pioneering what it calls AI-driven precision medical education."Read More

OpenAI's ChatGPT reaches 400M weekly users and doubles enterprise adoption to 2M+ users, signaling major AI shift as company prepares GPT-5 launch amid growing competition from DeepSeek and xAI.Read More

Cisco AI Defense is purpose-built to address the security paradox AI creates with its exponential growth in enterprisesRead More

Cybersecurity insurance is increasingly critical, particularly as AI transforms (and simplifies) hackers' methodologies.Read More

Vanguard Exiles, the next game by Magic: The Gathering creator Richard Garfield, will launch into early access on Steam on March 11.Read More

The California Institute of the Arts (CalArts) has launched the D.R.E.A.M.S. (Digital Research Entertainment Arts Media Storytelling) Initiative supported by a gift from Tom Dolan and the Dolan Family Foundation. The initiative is intended to prepare students for dynamic careers in the rapidly growing industry of location-based entertainment (LBE)Read More

Voltron Data partners with Accenture to revolutionize enterprise data processing with GPU-powered analytics engine Theseus, promising up to 100x performance gains for AI-driven companies facing massive data challenges.Read More

Nvidia has unveiled a new AI platform for teaching people how to use American Sign Language to help bridge communication gaps. The Signs platform is creating a validated dataset for sign language learners and developers of ASL-based AI applications. It so happens that American Sign Language is the third most prevalent language in the United StatesRead More

Grok-3 still hasn't fully shipped yet. But it will surely set the tone for how other AI labs release future models.Read More

Immutable, with King River Capital and Polygon Labs, described naugural returns on liquid investments of the Inevitable Games Fund.Read More

Heading to SXSW? You wont want to miss the first GamesBeat Engage an invite-only, salon-style gathering in Austin on March 11.Read More

Earlier today, the heads of Microsofts Xbox division revealed Muse, a generative AI model that intends to create both visuals and gameplay for games. The model, which was trained on the largely forgotten Ninja Theory multiplayer game Bleeding Edge, is not a shocking leap for Microsofts Xbox division. The company as a whole, from CEO SRead More

Stytch Connected Apps enables AI agents and third-party apps to authenticate, access data and take action on behalf of users.Read More

The Gayming Awards 2025 show will air on WOW Presents on July 8 as the event enters its fifth year.Read More

Microsoft new Muse AI system learns to generate interactive 3D game worlds by watching gameplay footage, marking a major advance in artificial intelligence's ability to understand spatial environments and human interactions.Read More

Crunchbase pivots to AI-powered predictions engine with 95% accuracy for startup funding rounds, abandoning historical data in favor of predictive intelligence that forecasts acquisitions and company growth.Read More

Alpha3D, an AI-powered platform for 3D asset creation, has acquired MoonlanderAI, an AI-driven game development platform.Read More

Sawmills claims to provide significant data transmission volume and associated cost savings for its enterprise customers.Read More

Chinese game publisher NetEase raised fears with recent moves that it is unloading as much as all of its overseas holdings in U.S. game companies and elsewhere due to rising costs and other reasons.Such a move, which coincidentally is happening at the same time as Chinas retaliation to U.S. tariffs, could send earthquakes through the global gaming industry and hurt U.S. game companies in particular. However, once contacted by GamesBeat about these fears, NetEase denied it is dismantling its far-flung investments and overseas game studios.Based on our reporting, game industry insiders said that NetEase has been directed to divest its overseas holdings. Among NetEases big recent hits is Marvel Rivals, but yesterday members of the Seattle team that developed the game along with a team in China said they had been laid off. This was after the game brought in $136 million in its launch month of December.But sources we interviewed said that the Marvel Rivals team being fired by NetEase is just the start.Those sources said the company is losing its will to make games using overseas staff. Part of that reason is the high cost of U.S. developers in particular. Another reason is that Chinas game developers have also matured enough to make triple-A games, as seen by the success of China-based Game Sciences Black Myth: Wukong, which has sold tens of millions of copies.Its possible that NetEase plans to exit all international investments and holdings, including wholly-owned studios. Some of that will happen through layoffs, and some could happen through sales of studios. But its unlikely that NetEase will turn a profit on those investments if it plans to act quickly. Again, NetEase denied this and weve included their statements below.Based in Hangzhou, NetEase is the second largest game publisher in China, publicly traded on Hong Kong Stock Exchange and NASDAQ. David Kaye, founder of the venture capital firm F4 Fund, said in an interview with GamesBeat that talk of the retreat of Chinas game companies was all over the show at the Dice Summit, the event that drew gamings elite to Las Vegas last week.NetEase has made a strategic investment in Behaviour Interactive.In a post, Kaye said, China is in retreat: geopolitical tensions, some big bets not paying off and the whims of certain CEOs mean that a massive pullback has begun. One MAJOR strategic who has made dozens of investments in the past several years is reportedly pulling the plug and divesting ALL investments outside of China. Some will likely find buyers, others will not be so lucky.Other sources told us the firm in question is NetEase, and the size of its impact is very big. Newzoo ranks the company as the No. 5 game publisher by revenue. Traditionally, 90% of NetEases gaming income came from China and 10% from overseas. Just two years ago, William Ding, the company founder and current chief executive, spoke about plans to have international game revenue reach up to 50%. NetEase executive Simon Zhu became head of investments in 2016 and he increased the investments dramatically.While Dings goal could have effectively doubled the size of the company, those dreams may be on pause.The company has been investing steadily for years, putting money into start-ups, growth stage game companies, as limited partners into venture capital firms, doing acquisitions, and starting internal studios.This activity has buoyed the number of studios and people in the game industry today. NetEase put $100 million into Bungie, which later sold to Sony for $3.6 billion.NetEase backed Devolver Digital when it was started.And it anchored Makers Fund, which had $960 million in assets under management.NetEase bought Quantic Dram and Grasshopper Manufacture outright.These efforts were led by Simon Zhu, based in Seattle.But things have soured internally.Sources say Ding, NetEases CEO, has lost confidence in expensive overseas teams.Add to that the success of Black Myth: Wukong, which illustrates that making triple-A games in China can be less costly than those from Europe, the US, and Japan. In the case of Marvel Rivals, the team in China is expected to take over the live services for the game.Gaming M&A slowed after the pandemic and is slowly coming back, hampered in part because of interest rate policies. Many of the studios will not find buyers, and that will add to the already excruciating number of job losses in the industry. More than 34,000 people were laid off in the last 2.5 years, and more than 900 have been laid off in January. Those who were demoralized by these moves were hopeful that 2025 would bring better news for the game industry. Kaye noted that South Korean firms are filling the void by launching expansion plans and funding new studios. Saudi Arabias Savvy Games Group and its Scopely division are also looking for acquisitions. That could help soften the blow for those looking for jobs or companies looking to be funded or acquired. One source told us that everything that NetEases Zhu touched (all investments and holdings) were going to be unwound. NetEases trailing 12 months of gaming revenue is $11.7 billion. One question is whether the world conditions like the rise of Chinese triple-A game development could impact other firms in China and cause them to pull back as well. In a statement to GamesBeat, NetEase moved allay fears it is abandoning its non-Chinese game developers. The statement said, As far as overseas business efforts are concerned, NetEase has not wavered in its global expansion plans. Our two-pronged approach, proposed in 2022 (combining self-research and investments to explore overseas markets), is still actively progressing and yielding positive results. For titles developed by the self-owned studios, we successfully launched games likeOnce HumanandMarvel Rivalsin 2024. These projects demonstrate NetEases ability, along with our talented development teams, to produce high-quality games loved by players worldwide. For 2025, we have an extensive pipeline of titles in development, featuring a variety of genres, including FragPunk, Ananta and more.However, NetEase confirmed that it has conducted a retreat in some respects. The statement continued, As part of our investment strategy, we started scaling down two of our studios at the end of 2024. This decision was based purely on business evaluations and was not influenced by other factors. And this represents only a small portion of our overseas studio portfolio. Our studios in North America, UK, Spain, and Japan all continue to refine and develop their ongoing game projects. Some additional background information that has not yet been shared to clarify our approachNetEase said it is establishing unified goals across studios. Whether overseas first-party studio or domestic ones, the goals and vision remain consistent. The aim is to develop fun, high-quality games that players will enjoy; and focus on the long-term development and continuous operation of games, providing players with a sustained, quality gaming experience and service.We are very open and aim to leverage our companys strengths and accumulated expertise to support all developers. In order to achieve this vision, we have implemented an evaluation process, which applies equally to all NetEase studios globally both domestic and overseas, NetEase said.Meanwhile, the company said the trade war has not affected its strategy yet. The statement said, Regarding the tariff trade war, yes of course we are mindful of any developments in international trade. However, our decisions have been based solely on business evaluations and not influenced by other factors. We highly respect our gaming community in the US and their love and enthusiasm for our games is valued and appreciated.Will the rest of China follow suit?SAN FRANCISCO, CALIFORNIA NOVEMBER 05: A general view of atmosphere at the League of Legends World Championship Finals on November 5, 2022 in San Francisco, CA. (Photo by Colin Young-Wolff/Riot Games)Tencent, which is the worlds biggest game company has trailing 12 months of game revenue is $26.1 billion, including $7.7 billion international revenue and $18.4 billion of domestic Chinese revenue. Tencents market value is $599.7 billion. Based on what we know, Tencent is not pulling back. But the environment in the market with Donald Trump imposing tariffs on Chinese goods possibly including games and Chinese-made game consoles isnt helping with the overall environment. It also may not help that the Biden administration in its final days classified Tencent as a Chinese military organization. Tencent denied it was a military company.Tencent is highly invested in the United States, with 40% ownership of Epic Games and 100% ownership of Riot Games. Tencent has hundreds of game companies in its portfolio worldwide, including the United States. NetEase is far less dependent on the US companies and user base than Tencent. There have been other Chinese companies with US gamer customers and some that have invested in US-based studios, said Lisa Cosmas Hanson, president of market analyst firm Niko Partners, in a message to GamesBeat.She noted that in 2020 there was a CFIUS inquiry about Tencents ownership of Riot and Epic, during the Biden Administration. There was concern about the privacy or security of U.S. gamer data, but the concerns were put to rest for the most part, Cosmas Hanson said. The investments differ greatly from ByteDances ownership of TikTok and the TikTok US matter that is currently in discussion between US and China. The difference is primarily that Riot and Epic are U.S. companies themselves.NetEase Games acquired Quantic Dream.I do not see how a divestiture, if proposed or enforced, would be related to tariffs directly. I just dont see it that way, Cosmas Hanson said. Regarding tariffs, I have publicly responded to some industry observers who have posted on LinkedIn about their suspicions that the U.S. games industry will suffer because the Trump Administration will set tariffs on products from China, including consoles. My feeling is that even if it were a blanked tariff on all products manufactured in China, there would be a carve out for many items, including consoles. This happened in the past.But she noted that the tariffs, if implemented without a carve out, would be a detrimental impact to the US and Japanese companies (Microsoft, Sony, Nintendo) for tariffs on their consoles sold in the U.S. but manufactured in China, but have no impact on China.So what is the point. These are not Chinese consoles or companies, just manufacturing in China, she said. Tencent recently canceled a Splash Damage game and it also restructured Sumo Digital. But it has expanded as well in the West. For instance, in November 2024, Tencent acquired EasyBrain from Embracer and also acquired the UK studio Space Ape in the same month, said Charles Yu, a partner at game-focused Pillar Legal in Shanghai, in a message to GamesBeat. And Tencent sources said the firm may become more active, so long as it does not attract unwanted attention.Yu said that he doesnt think its accurate that all Chinese game companies will divest their U.S. or overseas game properties. Some of the firms clients are interested in U.S.However, it appears that Chinese game companies are less inclined to set up development teams in the U.S. or other Western countries due to concerns about high costs and management inefficiencies, Yu said. I think it is probably true that NetEase is scaling back its overseas investments. A recent news indicated that NetEase laid off the entire overseas strategy investment team and shut down several studios in 2024. However, NetEase never officially announced this news.By market capitalization, NetEase is worth $68.4 billion. It is among the worlds top ten game companies (alongside Apple, Tencent, Google, Xbox, and Nintendo). NewZoo lists NetEase as No. 5 by annual publisher revenue. NetEase acts as local publisher for Chinese World of Warcraft (recently renewing that deal last year with Microsoft) and Chinese Minecraft. The change in strategy is a big reversal of expansion plans in the past.In 2021, the China Audiovisual and Digital Publishing Association encouraged the local industry to makeoverseas efforts. This was thematically in keeping the Belt and Road Initiative of the Chinese government. The countrys strategy is similar to that of Saudi Arabia, which has authorized $37 billion in investments into gaming through its Public Investment Fund.A public company like indie publisher Devolver Digital might have effects if NetEase has to liquidate its 8% of Devolvers stock onto the market. NetEase also has a stake in privately held developer Behaviour Interactive.Over the years, NetEase invested in companies including Bungie, Nagoshi Studio, Ouka Studios, Satelight, Grasshopper Manufacture, Humanoid Origin, Quantic Dream, Something Wicked Games, Reel Wolves, PlayPulse, Liquid Swords, Skybox Labs, Studio Flare, Anchor Point Studios, Bad Brain Game Studios, PinCool, Youdao, Maestro, Xiaoice, Jobtong, Build A Rocket Boy, MyDearest, Fantastic Pixel Castle, Worlds Untold, Jumpship, Second Dinner, Astrid Entertainment, Kepler Interactive, BulletFarm, Sandsoft and T-Minus Zero Entertainment.It going to be complex for NetEase if it does choose to unwind.As one executive said, There is no buyer for these shares.As for the U.S. games industry, Kaye said theres new dry powder for VCs.While not on the scale of the go go years, LPs are continuing to deploy. A well known Series A-focused gaming fund has just raised over $100 million in their first close, and there are reports of a large new media and entertainment focused fund backed by a major Korean company with a gaming focused partner who will be based in LA. They will focus mainly on Series B and later, filling an underserved spot in the capital stack, Kaye said.In general, I think the entire game industry is still feeling the impact of over-expansion during the Covid, Yu said. Many studios may face challenges or be at risk if their next flagship game underperforms. For NetEase, I anticipate a significant improvement in their financial performance in Q4 2024 and Q1 2025, driven by the successful launch of Marvel Rivals and Yan Yun Shi Liu Sheng in December 2024.Yu added, Chinese game companies remain highly motivated to explore overseas markets, and we may see more overseas headquarters established by these companies for game publishing. However, it is likely that Chinese game companies will slow down their investments in overseas studios. The global success of games developed by Chinese studios, such as Black Myth: Wukong, Marvel Rivals, and Delta Force, has boosted confidence in their domestic capabilities. As a result, I expect Tencent, NetEase, and other companies to allocate more resources to their China-based teams or local studios.Daily insights on business use cases with VB DailyIf you want to impress your boss, VB Daily has you covered. We give you the inside scoop on what companies are doing with generative AI, from regulatory shifts to practical deployments, so you can share insights for maximum ROI.Read our Privacy PolicyThanks for subscribing. Check out more VB newsletters here.An error occured.

The OpenAI rival will focus on strong infrastructure, multimodality and human-AI collaboration (as opposed to purely agentic systems).Read More