Latest news with #multimodal
Fast Company
a day ago
- Business
- Fast Company
The future of AI: Is your business ready to lead responsibly?
The future of AI holds much promise, but it also comes with dangers. As with any invention, the power of whether it becomes beneficial or a detriment depends mostly on who controls it. At Rank Secure, we have delved into AI wholeheartedly because our clients depend on us to remain on the cutting edge of marketing. Last year was trial and error until search engines started setting more rules and guidelines. THE GOOD We are just starting to explore the capability of AI, even though it feels like it has pushed society a lightyear into the future. Its use and capabilities remain at the baseline, with scientists and developers continuing to push the limits. AI is currently being used by most people to do routine tasks. Mundane tasks require objective, analytical thought. The next step is to allow AI to do tasks that require conscious thought, known as artificial general intelligence (AGI). AGI will have the ability to grasp data and information from a variety of sources, rather than just print. This multimodal version understands information derived from text, pictures, audio, video, graphics, and other sources. Multimodal versions will be able to understand someone's tone and facial expressions. AGI will be able to continue learning by teaching itself and learning from its experiences. AGI has much potential. The first phase is to be able to analyze data gathered from all sources. Eventually, we will need to include multimodal into websites. I am not ready to go there yet at my company because I feel more research and guidelines for proper use are needed. AI'S FUTURE USE The focus on AI or AGI usability centers around common-good systems like medicine and education. My company supports these sectors, as well as industries like e-commerce and service businesses, giving us a first look at trends, emerging tech, and new ways to use it. The key element transcending both common-good industries and consumer industries is personalization. Data is highly personalized, and allowing AI and AGI will pinpoint needs even further than they do today. AI can be used to tailor a student's learning to suit their unique needs, and it already plays a supportive role in diagnosis and treatment planning in medicine. It will likely play a larger role over the next few decades in predicting outcomes. So what does that have to do with my company? As a support partner for dental practices and educational providers, we need to implement personalization into their websites and create a user experience that feels like a real human interaction. THE BAD AI advancement comes with ethical and moral questions, as well as potential logistical and security issues. While AI and AGI may become more human-like, they'll always lack what makes us truly human—a soul. AI doesn't have an internal moral compass. And since it's created by human programmers, it carries human biases. These biases show up in what AI leaves out, and they can affect the decisions it makes. AI could even prompt decisions that prioritize societal outcomes over individual needs—for example, denying treatment to a senior citizen with cancer. So the question becomes: Can humans override AI when its decisions are bad or immoral? Right now, we don't have clear standards in place to guide ethical AI use. That's why business leaders should set boundaries for the ethical use of AI and other new tech. My own principles come down to three questions: Does it offer truth? Does it protect privacy? And does it ensure no one is exploited? A few guidelines we follow include: Be clear about any sign-ups to use specialized tech offered on the website. If it's free, it's free—no hidden subscriptions or surprise credit card authorizations in the fine print. Keep user data private. We don't sell the data we collect. Keep all data secure. Clearly distinguish AI-generated content from real people or sources. This is important when it comes to marketing. Never use AI-generated images of brands or celebrities without permission. AI DOUBLES THE REWARD—AND THE RISK AI can have a positive contribution to society, but risk still exists. We already have some experience with constant technology—for example, cameras on public street corners in many cities that aid police—yet, surveillance also invades privacy. Self-driving vehicles can cut down on DUIs, but they can also be hacked. AI can also expand access to mental health support, but it can't offer real companionship like humans can. We may become more efficient, but we risk losing treasures like community, family, and human connection. In the end, intent determines whether new tech like AI and AGI is beneficial or exploitative. It's a morally neutral tool—how we use it is what matters. At my company, our intent is to help clients make good contributions to their communities through their products and services. Others may use the same tools in more exploitative ways just to make a sale. LOOKING AHEAD Gen Z and Gen Alpha are best positioned to fully realize the potential of AI and AGI. They're growing up with it, and they won't carry the same fears or hesitations as older generations. These generations will ultimately shape how far AI goes and whether it's used for good or for bad. Their moral compass will define its limits. It's time for business leaders to consider paths forward using AI and AGI. Consider including a separate mission statement for using new tech in your company's founding principles and outlining policies that support that mission. Doing so will help guide you in the future through ever-growing changes and present a path for younger generations to follow.
Fast Company
2 days ago
- Business
- Fast Company
OpenAI unveils GPT-5 model, featuring improved coding and problem-solving chops
TECH The powerful new multi-modal model boosts reasoning, code generation, and real-time inference, and will be available to both free and paid users. [Images: Nathan Howard/Bloomberg via Getty Images; svetolk/Adobe Stock] BY OpenAI on Thursday unveiled its highly anticipated GPT-5, a powerful multi-modal AI model featuring major advancements in problem-solving and coding. The new flagship model was announced during a Thursday morning livestream. Unlike previous releases that were limited to paid subscribers, GPT-5 will be available to free-tier ChatGPT users as well, OpenAI said. OpenAI CEO Sam Altman described interacting with the new model as akin to conversing with a PhD-level expert, noting that while GPT-3 was comparable to a high school student and GPT-4 to a college student, 'With GPT-5 you get an entire team of PhD experts in your pocket, ready to help you,' Altman said. During the announcement, OpenAI researchers emphasized that GPT-5 was designed to be more reliable and accurate, with fewer hallucinations. GPT-5 offers improvements in reasoning, higher-quality code generation, greater autonomy with reduced need for user input, and seamless integration with platforms like ChatGPT and Google's Gmail and Calendar apps. Earlier language models relied only on pre-training to generate responses. GPT-5, like recent inference-based models, can also incorporate new data from user prompts in real time (a method known as 'test-time computing'). This release also unifies OpenAI's model naming, replacing names like o1 and o4-mini with the GPT-5 family, signaling a shift to models that combine both pre-training and inference. For many, the key question is whether the leap from GPT‑4 to GPT-5 will prove as dramatic as the jump from GPT‑3 to GPT‑4. Independent testing will shed light on that. OpenAI also recently introduced two 'open-weight' models —gpt-oss-120b and gpt-oss-20b—which are freely available and modifiable by developers. This move marked a rare shift from its typically closed model strategy. However, with GPT-5, OpenAI returns to its more traditional 'closed' approach. In addition, the company announced a partnership with the U.S. federal government to provide executive branch agencies access to its enterprise-grade chatbot. Through a landmark agreement with the General Services Administration (GSA), ChatGPT Enterprise will be made available to agencies for just $1 per agency for one year. OpenAI has assured that it will not use government data to train its AI models. The early-rate deadline for Fast Company's Most Innovative Companies Awards is Friday, September 5, at 11:59 p.m. PT. Apply today. Sign up for our weekly tech digest. SIGN UP This site is protected by reCAPTCHA and the Google Privacy Policy and Terms of Service apply. Privacy Policy ABOUT THE AUTHOR Mark Sullivan is a San Francisco-based senior writer at Fast Company who focuses on chronicling the advance of artificial intelligence and its effects on business and culture. He's interviewed luminaries from the emerging space including former Google CEO Eric Schmidt, Microsoft's Mustafa Suleyman, and OpenAI's Brad Lightcap More
National Post
5 days ago
- Business
- National Post
Oxmiq Labs Inc.™: Re-Architecting the GPU Stack: From Atoms to Agents™
Article content Sorry, your browser doesn't support embedded videos. Article content CAMPBELL, Calif. — Oxmiq Labs Inc., the all-new GPU software and IP startup founded by one of the world's top GPU architects and visionaries, Raja Koduri, emerges from stealth after two years of intensive IP development. Raja has assembled a world-class team of GPU and AI architects with over 500 years of combined experience, hundreds of patents, and a collective track record of generating more than $100B in revenue at prior companies. Article content Article content Modern computing has fundamentally shifted toward multimodal experiences where text, audio, video, images, and 3D environments seamlessly interact, establishing GPU architecture as the cornerstone of this transformation. Unlike fixed-function AI accelerators that handle specific tasks, GPUs provide the general-purpose computational flexibility required for these diverse modalities while maintaining deep integration with mainstream operating systems through standardized APIs and unified memory models. This architectural advantage positions GPUs as the essential compute engine for both current applications and the emerging landscape of multimodal AI, where heterogeneous workloads must be processed in harmony. Article content OXMIQ™'s licensable GPU IP rearchitects the GPU from first principles incorporating breakthrough technologies including nano agents in silicon leveraging RISC-V cores, near-memory and in-memory computing, and light transport. OXMIQ delivers solutions that balance multimodal computing flexibility with the radical performance improvements required for next-generation graphics and AI workloads for its customers. Article content Software First: Article content Learning from decades of industry evolution, OXMIQ embraces a Software First strategy that prioritizes developer experience through a comprehensive software stack compatible with both OXMIQ IP-based silicon and third-party GPU and AI accelerator platforms. Article content OXCapsule™, Article content OXMIQ's unified software ecosystem, abstracts away hardware complexity to provide frictionless deployment across diverse computing platforms, eliminating the configuration challenges that traditionally plague heterogeneous environments. Article content A flagship component, OXPython™, enables Python-based NVIDIA® CUDA™ AI applications to execute seamlessly on non-NVIDIA hardware without code modification or recompilation. Launching initially on Tenstorrent's AI platform later this year with multiple vendor integrations in progress, OXPython demonstrates OXMIQ's commitment to breaking down hardware silos and accelerating the democratization of high-performance computing across the industry. Article content 'We're excited to partner with OXMIQ on their OXPython software stack,' said Jim Keller, CEO of Tenstorrent™. 'OXPython's ability to bring Python® workloads for CUDA to AI platforms like Wormhole™ and Blackhole™ is great for developer portability and ecosystem expansion. It aligns with our goal of letting developers open and own their entire AI stack.' Article content The OXMIQ Breakthrough: Article content Beyond software, OXMIQ delivers a complete GPU hardware IP stack that powers silicon solutions scaling from Physical AI in edge devices and autonomous robots to enterprise edge infrastructure and zettascale data centers. Their scalable GPU core, OXCORE™, integrates scalar, vector, and tensor compute engines in a modular architecture customizable for specific workloads, enabling nano-agents, native Python acceleration, and compatibility with SIMD/CUDA paradigms. Article content OXQUILT™ Article content chip-let architecture. Through Article content OXQUILT Article content , customers can configure optimal ratios of compute, memory and interconnect for their needs and achieve significant reduction in time to market, R&D and production costs v/s current industry standard methodologies. Article content Capital Efficiency & Market Traction: Article content OXMIQ has raised $20 million in seed funding from prominent technology investors, including strategic players in mobile and AI silicon such as MediaTek®, and has recorded its first round of software revenue. With its licensing-first model, OXMIQ avoids the heavy capital requirements of chip startups that depend on expensive EDA tools and physical tape-outs — delivering outstanding capital efficiency. Article content 'OXMIQ has an impressive bold vision and world-class team,' said Lawrence Loh, SVP of MediaTek 'The company's GPU IP and software innovations will drive a new era of compute flexibility across devices – from mobile to automotive to AI on the edge.' Article content Mihira™: Article content After co-founding Mihira two years ago with Shobu Yarlagadda and SS Rajamouli, Raja Koduri has transitioned out of day-to-day operations at Mihira to focus full-time on running OXMIQ. Raja now acts as a strategic advisor to Mihira Visual Labs™, which is led by CEO Shobu Yarlagadda. OXMIQ holds a minority stake in Mihira and will continue to support its growth through foundational agentic and GPU IP technologies that power Mihira's cinematic AI platform. Article content 'Raja's early contributions to Mihira helped shape our foundational vision of cinematic AI. Now, with OXMIQ, he and his team are building the deep-tech infrastructure that powers our next chapter,' said Shobu Yarlagadda, Co-Founder and CEO of Mihira Visual Labs™, and acclaimed producer of the Baahubali film series. 'As we scale Mihira into a global creative platform, we're thrilled to continue our close collaboration with OXMIQ and integrate their agentic GPU innovations into our storytelling stack.' Article content Article content Article content Article content
CBS News
24-07-2025
- Business
- CBS News
Emeryville considering creating bus-only, bike-only lanes on 40th Street
A proposal for bus-only and bike-only lanes in part of Emeryville's busiest corridor has some businesses in the area concerned. "It's a third-generation business, and I'm generation two," said Deborah Cohen. She and her family have owned Rug Depot Outlet in Emeryville for the past 55 years. The store has more than 10,000 rugs in the showroom and warehouse. There are massive semi-trucks coming and going almost every day, not just for her business but others nearby like Pottery and Beyond, Granite Expo and more. "A lot of our neighbors are light industrial or in the catering business, and so this street takes delivery on very heavy product," said Cohen. That is why she and about a dozen neighboring businesses are so concerned about the 40th Street Multimodal project that's moving forward in the city council. The plan is to make 40th Street more friendly to pedestrians, cyclists, and buses by closing side streets to vehicles, and eliminating one traffic lane in each direction, and converting it into dedicated bus and bike lanes. The total cost of the project is just over $30 million. "This project is in a big way looking to accommodate all of the users who want to come into the city," said Emeryville's Mayor David Mourra. He said they've been studying the corridor since 2018 and have done traffic studies looking at what makes the most sense. He said this plan helps make the 40th Street area safer for not just pedestrians and cyclists, but people in cars as well. "We are really trying to improve things for everyone, and sometimes there are tradeoffs, and we understand that and we understand that sometimes change can be difficult," said Mourra. For Deborah and other business owners, they feel like the change is only a positive for a small few at the expense of legacy businesses that built Emeryville into what it is today. "40th to me is a throughway that was made for this exact thing, which is deliveries of trucks coming off of freeways, and to make 40th street into one lane car traffic doesn't feel safe or relevant," she said. The city council still has to vote on the plan and that vote is expected sometime toward the end of the year, and if it is approved, the project wouldn't be completed until 2028.
The National
10-07-2025
- Business
- The National
Samsung focuses on voice commands for XR headset
Samsung Electronics' coming extended reality headset will focus on a multimodal experience centred on voice as it prepares to enter its newest hardware category later this year, a senior executive has said. The company is in no rush to release the device dubbed Project Moohan in a category that has yielded mixed results for even the biggest technology companies, said Hyesoon Jeong, executive vice president, and head of framework research and development at Samsung Electronics. In XR, multimodal interaction includes voice commands, hand gestures, eye tracking, haptics and controller inputs. In more advanced cases, brain computer interfaces can be used. Project Moohan, which is being developed in partnership with Google, was teased at January's Unpacked in California. Ms Jeong had previously told The National that it might not be released this year, but Google, at its I/O conference in May, announced that it will be available by the end of 2025. 'For XR, we are already developing the UI [user interface] and it will be applied and moving forward' while the hardware is being built, she said at a round-table on the sidelines of Galaxy Unpacked in Brooklyn on Wednesday. 'The interface we will develop [will be] more say voice centric and multimodal; that's the direction that we will go,' she added, without specifying what interactions will be included in Project Moohan. Samsung's jump into the XR realm will pit it against some of the tech industry's top names in an increasingly crowded sector. It is already well positioned to do so being the world's largest mobile phone manufacturer and a top chipmaker. Meta Platforms, the parent company of Facebook, already has its Meta Quest line-up of AR headsets and teamed up with Ray-Ban to launch smart glasses, while Microsoft already has its HoloLens. Apple, Samsung's biggest rival in smartphones, has the high-end Vision Pro, which has since been discontinued. Another advantage Samsung enjoys is its extensive partnership with Google, allowing it to tap into the Alphabet-owned company's major generative artificial intelligence platform, Gemini – and AI itself is a critical factor in the development of advanced XR systems. For Google, it could be an opportunity to redeem itself from the failure of its Project Iris and, most notably, Google Glass, its first major attempt to bring the device into the mainstream that struggled with a number of problems and a hefty $1,500 price tag. Google's Gemini was chosen by Samsung as its main generative AI partner over other platforms such as Microsoft's Copilot because of its 'openness', Ms Jeong said. California-based Google is the developer of the Android operating system used in Samsung smartphones and is the owner of some of the most used apps globally, including Gmail and YouTube, in which Gemini is integrated. 'We thought that [Gemini] was optimised for the mobile environment. That's why we went with Gemini,' Ms Jeong said. 'Galaxy AI pursues openness, therefore, based on that philosophy, we continuously look at user needs and are always open to adopt to what users call for.' 'Respect' for user app preferences Meanwhile, Samsung said it will maintain the presence of its native apps on its devices despite the dominance of Google's Gemini-backed services out of 'respect' for user preferences, Ms Jeong said. A number of apps that are pre-installed on Samsung phones perform the same tasks such as Google's; for example, Samsung Internet, Samsung Music and Samsung Gallery are equivalent to Chrome, YouTube Music and Google Photos. Samsung apps cannot, however, be considered as bloatware, which are apps that are unnecessary or hinder performance, as they are part of the Galaxy Ecosystem that connects the company's devices, similar to the interconnectivity of Apple's hardware and software. Samsung also has its own Galaxy Store where users can download apps specifically programmed for the company's devices; Ms Jeong said Samsung has a substantial number of users tapping into their native apps, hence the reason they will stick to this strategy. 'There are other instances though where we have to respect the choice of consumers, as well and their feedback,' she said. 'There are still many people who use Samsung [apps] … we try to respect and give options when needed.'
