Raju Ghivari (MBA-Finance)’s Post

According to Kai-Fu Lee, who worked as the founding Director of Microsoft Research Asia before joining Google and Apple, the behemoths of Generative AI could have it all wrong. He is critical of the inverted investment pyramid of the current Generative AI business model, which is based on hype, massive scaling, and extremely high costs for processing queries. Kai-Fu Lee is on to some essential things to consider if Generative AI will ever succeed and be the basis of a profitable business. Hint: the apps will rule, and Generative AI systems will be small, very high-speed, and cheap to operate if built for specific purposes using appropriate hardware. Today's scaling Generative AI systems won't survive if Nvidia and AMD make most of the money. https://lnkd.in/enjganwC #GenAI, #OpenAI, #KaiFuLee #BeaGo #Invidia

Nvidia became the most valuable company in the world. Claude launched the most advanced model, and Ilya launched his own new AI startup SSI. There is a lot going on in the AI world. Here's everything you need to know: _____ Kling AI is frequently hailed as China's answer to Sora. While it might not yet match Sora's quality, it's impressive to see a Chinese company delivering such innovative video concepts. _____ While Kling AI doesn't quite reach the quality of Sora, a few days later Runway was launched and it was very close to Sora's quality. _____ Photo to Video is here - and it's here to stay. It's free, you can try it out on LumaLabsAI. _____ Now imagine Lumalabs with excellent lip sync in video—that's Herdalabs for you. While these text-to-video and photo-to-video tools were going all in on X. There was a silent drop from Anthrophic beating GPT-4o in almost all benchmarks. Claude Sonnet 3.5 _____ How can all the AI companies run without Nvidia? Well, here it is. Nvidia has officially become the most valuable company in the world, beating Microsoft. _____ Everybody was saying, "Where is Ilya?" and all the memes got famous. But he was building in stealth mode and finally, he launched his own AI startup SSI. _____ #ai #nvidia #cluade

🚀💡 OpenAI's Bold Leap: Pioneering Its Own AI Chips in Tech Tango! Imagine a world where AI runs not just efficiently, but independently of conventional tech giants. 🌐✨ Here's the breakthrough scoop: OpenAI, the mastermind behind revolutionary AI developments like ChatGPT, is setting the chessboard for arguably its most strategic move yet: designing its own AI chips. 🖥️🧩 Why, you ask? Here's the what and the why: 1️⃣ Breaking Free: By developing its own hardware, OpenAI aims to lessen its reliance on NVIDIA's GPUs, which are powerful but expensive and often in short supply. 💸🔗 2️⃣ Innovation Circle: Talent isn't an issue. Former Google whizzes who crafted the impressive tensor processing unit are now on board, tweaking everything from chip packaging to memory solutions. 🧠🌀 3️⃣ Looking Ahead: Although full production is slated for 2026, this pioneering journey could vastly accelerate OpenAI's strides towards Artificial General Intelligence (AGI). 🕒🌌 Plus, OpenAI isn't stopping there. They're eyeing new ventures and outside investments to build a robust infrastructure, including data centers, ensuring they keep pace in the global tech race. 🏗️💼 The implications are vast: From reshaping how businesses leverage AI without the GPU bottleneck to spurring innovations that could change our interaction with technology, OpenAI’s initiative is more than just a technological upgrade—it's a potential industry revolution. 🚀🌍 As AI weaves deeper into our lives and businesses, understanding these shifts isn't just useful—it's essential. 📊👀 Stay tuned, tech enthusiasts! The road ahead is as exciting as it is transformative. #AI #TechNews #OpenAI #Innovation #ArtificialIntelligence

Nvidia became the most valuable company in the world. Claude launched the most advanced model, and Ilya launched his own new AI startup SSI. There is a lot going on in the AI world. Here's everything you need to know: _____ Kling AI is frequently hailed as China's answer to Sora. While it might not yet match Sora's quality, it's impressive to see a Chinese company delivering such innovative video concepts. _____ While Kling AI doesn't quite reach the quality of Sora, a few days later Runway was launched and it was very close to Sora's quality. _____ Photo to Video is here - and it's here to stay. It's free, you can try it out on LumaLabsAI. _____ Now imagine Lumalabs with excellent lip sync in video—that's Herdalabs for you. While these text-to-video and photo-to-video tools were going all in on X. There was a silent drop from Anthrophic beating GPT-4o in almost all benchmarks. Claude Sonnet 3.5 _____ How can all the AI companies run without Nvidia? Well, here it is. Nvidia has officially become the most valuable company in the world, beating Microsoft. _____ Everybody was saying, "Where is Ilya?" and all the memes got famous. But he was building in stealth mode and finally, he launched his own AI startup SSI. _____ #ai #nvidia #cluade

Every week, I sit down and research on the new AI releases. Here's everything for this week: 📌 OpenAI Co-founder Ilya Sutskever announces new AI company Ilya Sutskever, co-founder and former chief scientist of OpenAI who left the company a month ago and sparked controversy, has announced that he is starting his new AI company, Safe Superintelligence Inc. (SSI) Sutskever explained in a recent post that SSI seeks to prioritise AI safety, particularly in the context of superintelligent AI systems, also mentioned on the SSI website. 📌 NVIDIA rise to the world's most valuable company A lot of people would love to time-travel back to 2019 and buy NVIDIA stock. Chipmaker NVIDIA made it to the headlines this week on Wednesday, when surpassed Microsoft and Apple to become the world’s most valuable company with a market capitalisation of more than $3.4 trillion! To give a little more perspective, Nvidia has recently grown larger than the entire stock markets of countries like Germany, France, and the United Kingdom. 📌 Claude 3.5: Anthropic's latest model competes with GPT -4o. The battle of the AI models continues as Anthropic releases its newest model, Claude 3.5 Sonnet, which it claims can equal or outperform OpenAI's GPT-4o and Google's Gemini across a wide range of tasks. That's it for this week. I have shared more details in my latest newsletter. Subscribe here: https://lnkd.in/g7h7g26B #aitools #founder #entrepreneurship

100 Mustangs or 1,000 Teslas? To switch or not to switch, that is the question. Sporadic unplanned journeys and untamed horsepower may become a thing of the past when switching to a more refined, electrified driving experience. When it comes to AI, a similar problem exists under the hood for the driving experience (inference). At a granular level, modern AI model architecture often contains a pipeline of complex matrix or vector computational operations. These are used to maximize the likelihood of extracting features from data representations as they pass from one node to the next within a model. Graphical Processing Units (GPUs) have been favored for providing that raw, untamed horsepower for AI workloads. The ability to perform thousands of matrix calculations in parallel is one of their key benefits. However, one company chose a different route. Groq has built a custom piece of hardware called a Language Processing Unit (LPU), specifically designed for inference tasks (using trained AI models to complete a task). According to CEO and founder Jonathan Ross LPUs use one-third of the energy compared to GPUs when scaled. Isn’t that a nice approach to #sustainableAI https://lnkd.in/gWWhmUjN?

Nvidia's latest results provide a wealth of information about the evolving AI market and the role of AI chips in this transformation. Here are five key insights one can derive from these developments to infer where AI market is heading in 2024 and beyond: 1. Shift from Training to Deployment: Nvidia has been a dominant player in supplying chips for AI model training. However, the market is swiftly moving towards the deployment phase of AI, where models, once trained, are used to generate outputs like text and images. This transition is significant because it opens up a new, potentially larger market for chips that perform inference tasks. Nvidia's CFO Colette Kress noted that over 40% of the company's data center business in the past year was dedicated to deploying AI systems, signaling a pivotal shift in the industry focus from training to inference. 2. Increasing Importance of Inference Chips: Inference involves using trained AI models to make predictions or decisions based on new data. This phase can often be less computationally intensive than training, allowing for the use of less powerful and less expensive chips. Nvidia's success in this area, with a substantial portion of its revenue coming from inference, challenges the perception that its market share might decrease in the face of cheaper alternatives. It highlights Nvidia's adaptability and potential to capitalize on the growing demand for inference capabilities. 3. Competitive Landscape Expansion: As the focus shifts towards inference, the competition is expected to intensify. Companies like Intel, with its expertise in central processing units widely used for inference tasks, pose a significant challenge to Nvidia by offering cost-effective alternatives. This competition underscores the evolving dynamics of the AI chip market, where different types of chips may find their niches based on performance, cost, and energy efficiency. 4. Rise of AI Chip Startups: The increasing emphasis on inference has also spurred the growth of startups specializing in AI chips. Companies like SambaNova Systems and Groq are gaining traction by offering chips and software solutions optimized for inference tasks. These startups are not only contributing to the diversification of the market but are also pushing technological advancements that could lead to more efficient and cost-effective AI deployments. 5. Big Tech's Internal Developments: Major technology companies such as Meta, Microsoft, Google, and Amazon are developing their own inference chips, reflecting a strategic move to optimize the cost and performance of AI applications internally. Amazon's use of inference chips for its Alexa smart assistant, accounting for 40% of computing costs, exemplifies the economic and operational benefits of custom chip development for large-scale AI deployments. #AI #market #generativeAI #nvidia

AI Innovations Unleashed: OpenAI's $6.6B Windfall & Nvidia's LLM Surprise Ever wondered what the AI future holds? This week offers a glimpse! OpenAI has secured a staggering $6.6 million funding round, reshaping its landscape with both financial muscle and innovative expansion. With giants like Microsoft and fresh faces such as SoftBank joining the table, OpenAI's valuation has skyrocketed to $157 billion, energizing its mission to democratize AI tools. Also, eyeing the broader AI space, OpenAI has introduced discounted plans tailored for nonprofits and educational customers, ensuring their cutting-edge technology is accessible to those fostering talent and societal benefit. Meanwhile, Nvidia has thrown a curveball by unveiling a surprise large language model, making waves in the AI community. This move into AI software with LVNM 1.0 positions Nvidia alongside top-tier contenders, offering developers a playground to craft novel applications and chatbots. Yet, amidst the excitement, a vigilant eye is cast on privacy. The advent of Meta Smart Glasses, reportedly modified to breach privacy, reminds us of the ethical boundaries that accompany technological advances. Keeping AI’s potential from falling into the wrong hands is crucial, and OpenAI is making concerted efforts to thwart misuse while ensuring its systems remain trustworthy. As we ride the crest of these AI developments, it's clear that the technology isn’t just evolving—it’s transforming the framework within which business and innovation thrive. The future of AI is not just about groundbreaking ideas; it’s about responsible stewardship and inclusive advancements. Stay tuned as we navigate these thrilling, yet challenging, waters. Sources: Digital Trends- https://lnkd.in/eudQECHA #AIInnovation, #FutureOfTech

NVIDIA's CEO just made a game-changing statement: "Agents are the future of AI." (And this shift is closer than you think.) Here’s why it matters: Right now, AI operates as a powerful tool for tasks. But imagine AI that can act as your agent—handling decisions, running tasks autonomously, and adapting in real-time. This is not just tech evolution. It’s a revolution in how we interact with technology. Think personalized virtual agents that can: Negotiate on your behalf. Manage your schedule autonomously. Make intelligent decisions based on real-time data. It’s happening. This new wave of AI agents could transform industries, from finance to healthcare to creative fields. The question is: Are you ready to embrace this new era?

🚨 Kai-Fu Lee has declared war on Nvidia and the entire US AI ecosystem. 🔹 Kai-Fu Lee, a prominent figure in the AI field, has declared a competitive stance against Nvidia and the broader U.S. AI ecosystem, asserting that significant changes are necessary for real progress. In a recent talk at Collective[i] Forecast, he characterized the current U.S. AI landscape as "incredibly sick" and in need of radical restructuring. According to Lee, the ecosystem is overly reliant on Nvidia and small AI chip manufacturers, who collectively earn $75 billion annually, while infrastructure and application vendors generate significantly less. He warns that this inverted economic model is unsustainable and suggests that AI companies must develop their own vertically integrated technology stacks, similar to Apple's approach with the iPhone, to lower the costs associated with generative AI. 🔹 Lee emphasizes the need to focus on reducing the cost of inference, which is crucial for making AI applications more accessible to businesses. He highlights that the current pricing model for services like GPT-4—$4.40 per million tokens—is prohibitively expensive compared to traditional search queries. This high cost hampers the widespread adoption of AI applications in business, necessitating a shift in how AI models are developed and priced. By lowering inference costs, companies can enhance the practicality and demand for AI solutions. 🔹 Another critical direction Lee advocates is the transition from universal models to "expert models," which are tailored to specific industries using targeted data. He argues that businesses do not benefit from generic models trained on vast amounts of unlabeled data, as these often lack the precision needed for specific applications. Instead, creating specialized neural networks that cater to particular sectors can deliver comparable intelligence with reduced computational demands. This expert model approach aligns with Lee’s vision of a more efficient and cost-effective AI ecosystem. 🔹 Lee's startup, 01. ai, is already implementing these concepts successfully. Its Yi-Lightning model has achieved impressive performance, ranking sixth globally while being extremely cost-effective at just $0.14 per million tokens. This model was trained with far fewer resources than competitors, illustrating that high costs and extensive data are not always necessary for effective AI training. Additionally, Lee points out that China's engineering expertise and lower costs can enhance data collection and processing, positioning the country to not just catch up to the U.S. in AI but potentially surpass it in the near future. He envisions a future where AI becomes integral to business operations, fundamentally changing how industries function and reducing the reliance on traditional devices like smartphones. Source | https://lnkd.in/dyTFbxCU