Why I am not surprised “Further down the value chain, away from the glow of Nvidia, lurk signs of discontent. Businesses have cut back on whizzy new AI tools out of concern for hallucinations, cost and data security. “ “First, tech companies need to identify where their hype machine has gone wrong. They didn’t set expectations for AI’s capabilities too high; they framed its use as being too general purpose.” “Framing AI as a general-purpose Swiss Army knife for productivity inevitably leads to paralysis for its end users: Where do you even start with a technology that can do everything?” “AI isn’t yet a jack-of-all-trades but a master of a few. The sooner business leaders realize they can apply it to an array of niches and not for everything, everywhere, all at once, the sooner they can make the technology useful for them.” *** What prompts this analysis by Bloomberg is the meteoric rise of the stock price of NVIDIA. It is producing great concern because it resembles the internet bubble we saw 20 year ago. There is no basis for the 3 trillion USD value of the GPU company. How do we start correcting course: 1. Stop calling data science and machine learning AI. They are not interchangeable terms 2. Treat AI as a science not a mercantilistic goal 3. Don’t buy “AI” solutions for the sake of being trendy and smart. You will get the opposite effect 4. Start slow. Build your foundations with statistics, data science and data quality processes. Examine the need for machine learning algorithms 5. Inform yourself about what AI is. Put your hands on an old good friend: books. 📚 Be extremely selective with your internet choices of reading, mindful of the internet press makes money by your clicks and your data 6. Set realistic goals and establish a real, concrete result if you get hooked with “AI”. If a vendor is too insistent, try a demo or a pilot of the “AI solution” for six months, at no cost 7. Ask the vendor for real world results. Data is not perfect. You may want to try the “AI” product on your data first 8. Move your focus to the data. See if your project or challenge qualifies for machine learning. If your data is imperfect, solve the issue with the measurement processes upstream. No algorithm or “AI” does miracles. It’s all about the data 9. Be specially cautious if the vendor claims to replace your “old” and “slow” physics based models, or sound engineering workflows, with data-driven only, machine learning, or solve-it-all super fast “AI” 10. Digital Transformation is not about switching old toys for new shiny toys; it is about improving efficiencies, solving business problems, make the intricate reproducible. There is no point spending $100 million to obtain a benefit of $1 million 11. Use common sense. Ask your engineers and end users on the “AI” product. Involve them, use the bottom-up approach, and not the power of your title #AI #artificialIntelligence #spe #petroleumEngineering #machineLearning #DigitalTransformation
Alfonso R. Reyes’ Post
More Relevant Posts
-
#HypeOfAI #WomenInTech&Investing Is GenAI going to deliver on it's promise? And till then ...... Who is making money from GenAI?.... It's an incestuous relationship amongst all .... so all benefit ...from picks & shovels to the companies actually digging & owning the gold.....and you as an investor can be part of the gold rush .... being cautious that the real proof of AI will come after the POC stage..... Listed below is the “AI value chain” which investors can explore: - Layer I - AI infrastructure: This stage focuses on semiconductor companies that provide essential components for computational power and memory, including processors, GPUs, and advanced chips that enable AI algorithms to function. Think Nvidia, TSMC, ASML - Layer II - Software Services: The next layer involves the development of AI algorithms and platforms. This includes machine learning frameworks as well as other software tools such as Large Language Models (LLMs). These software solutions provide the underlying tools and frameworks for AI development. Meta, Google, Open AI - Layer III - Applications: This layer explores the diverse AI applications across various industries and seeks out companies with innovative business models capable of monetising AI through product offerings, thereby driving real-world transformation. - Layer IV - the consultancy companies like Accenture and others and the Cloud Companies holding all these layers... While investors have primarily focused on semiconductor companies, the AI value chain extends far beyond this initial layer. As AI evolves, there could be opportunities in the software services and applications layers.
Who’s making money from GenAI? Big Tech, consultants or data centres?
livemint.com
To view or add a comment, sign in
-
➡ 𝐆𝐞𝐧𝐞𝐫𝐚𝐭𝐢𝐯𝐞 𝐀𝐈 𝐢𝐧 𝐁𝐢𝐠 𝐓𝐞𝐜𝐡: Companies like Google (with Gemini) and OpenAI are rapidly advancing generative AI models, known as large language models (LLMs). ➡ 𝐈𝐧𝐟𝐫𝐚𝐬𝐭𝐫𝐮𝐜𝐭𝐮𝐫𝐞 𝐃𝐞𝐦𝐚𝐧𝐝: The rise of GenAI is driving a massive demand for advanced data centers and semiconductor chips, essential for supporting the computational needs of LLMs. ➡ 𝐈𝐧𝐝𝐮𝐬𝐭𝐫𝐲 𝐋𝐞𝐚𝐝𝐞𝐫𝐬: NVIDIA, Microsoft, and Meta are leading efforts to expand AI capabilities, pushing the boundaries of current tech infrastructure. ➡ 𝐓𝐞𝐜𝐡 𝐈𝐧𝐧𝐨𝐯𝐚𝐭𝐢𝐨𝐧: GenAI is accelerating innovation, reshaping the landscape of the technology sector. For more details, you can read the full article at LiveMint https://lnkd.in/gXST4iFC
Who’s making money from GenAI? Big Tech, consultants or data centres?
livemint.com
To view or add a comment, sign in
-
Nvidia: Boss says AI at 'tipping point' as revenues soar The boss of the world's most valuable chip maker Nvidia said artificial intelligence (AI) is at a "tipping point" as it announced record sales. The technology giant reported that revenues surged by 265% to $22bn (£17.4bn) in the three months to 28 January, compared to a year earlier. For the year as a whole, turnover more than doubled to $60.9bn. "Accelerated computing and generative AI have hit the tipping point," said Nvidia chief executive Jensen Huang. "Demand is surging worldwide across companies, industries and nations." Nvidia also forecast a 233% jump in its quarterly revenues for the current quarter, beating analysts' estimates. Nvidia: The chip maker that became an AI superpower "There was a lot riding on this last quarter and they actually absolutely knocked it out of the park," Bob O'Donnell of Technalysis Research told the BBC. "We're starting to see mainstream usage of AI," he added, highlighting that AI is not longer only used by specialised technology companies. In addition to its AI chips, sales at the firm's data centres have grown rapidly. Its data centre business contributed the vast majority of its revenues in the most recent quarter after growing more than than five-fold over the last year. However, the company said it faced several challenges including constraints on its supply chains. The US has also tightened its restrictions on trade with China, the world's second largest economy. Ipek Ozkardeskaya, a senior analyst at Swissquote, told the BBC that Nvidia's results had been "unusually amazing". But she added that Nvidia could face difficulties in addition to restrictions in China. "Nvidia... will see challenges on the way up because first the revenue growth will likely stabilise and the euphoria regarding these growth and growth perceptions will level out," she said. The company is also likely to face competition and regulation issues, Ms Ozkardeskaya added, and it could be "constrained by their own capacity to respond to this fast-surging demand". AI's public profile has risen sharply since the launch in 2022 of ChatGPT, which was developed by Microsoft-backed OpenAI. ChatGPT and other similar systems use huge amounts of data to create convincing human-like responses to user queries. They are expected to dramatically change the way people search for information online. Nvidia's stock market value has soared by 225% over the last year, making it one of the most valuable companies in the US. Its share price jumped by more than 9% in extended New York trading. #news #uknews #technews #latestnews #dailynews #newsupdates #dailyupdates
To view or add a comment, sign in
-
Headline: Reflecting on AI's Collaborative Success: NVIDIA's Whales and Likely Candidates As I was reading a Fortune article on NVIDIA's CEO Jensen Huang, I couldn't help but appreciate the spirit of collaboration that's driving AI innovation. Despite nearly half of NVIDIA's revenue coming from just four anonymous "whale" customers, Huang emphasizes their role as infrastructure providers, not competitors. This attitude fosters partnerships with tech giants like Amazon, Microsoft, and Google. While the whales' identities remain undisclosed, it's reasonable to speculate that these high-spending customers could be major tech companies, research institutions, or governments with significant investments in AI and deep learning technologies. Potential candidates may include the likes of Amazon, Microsoft, Google, Facebook (Meta), Apple, or even large research organizations such as OpenAI or DeepMind. The article highlights how collaboration and knowledge sharing can fuel the growth and advancement of AI technology. NVIDIA's willingness to work with, rather than against, other industry leaders serves as an example of how cooperation can propel the entire field forward. In a time where competition is the norm, it's refreshing to see collaboration and shared progress driving the success of AI. It's a valuable lesson for all of us in the industry. What are your thoughts on the importance of collaboration for AI's future? Do you have any predictions on who NVIDIA's mystery whales could be?
Nearly half of Nvidia’s revenue comes from just four mystery whales each buying $3 billion-plus
fortune.com
To view or add a comment, sign in
-
Did you know that NVIDIA controls over 90% of the #AI chip market? However, this dominance could soon be challenged as Meta and Google have announced the development of their own AI chips. Sharing his insights with Fortune, Senior Principal Analyst Edward Wilford discusses what this move brings to the highly competitive AI #semiconductors landscape. Learn more here (subscription required): https://lnkd.in/gUnsw-Sy
Meta and Google announce new in-house AI chips, creating a “trillion-dollar question” for Nvidia
fortune.com
To view or add a comment, sign in
-
Finally a good article about A.I. "Tech giants like Alphabet Inc., Amazon.com Inc. and Microsoft have warped expectations for generative AI’s contribution to profits. They will pay for that if they don’t temper expectations... If you were to measure the malaise with the Gartner Hype Cycle, AI would be deep in the “trough of disillusionment:” In this fools' gold rush the only one making money is the one selling the pick and axes, again, here... the chips: Nvidia. https://lnkd.in/eRd3qxw3
Nvidia’s Explosive Growth Masks AI Disillusionment
bloomberg.com
To view or add a comment, sign in
-
Is there a $600 billion hole in the AI business model? Sequoia recently claimed that “the AI bubble is reaching tipping point”. (If bubbles can tip – but you get their meaning). Sequoia used a simple metric which I’ll summarise here: Every $150B capex that Nvidia sells to data centres (DCs) costs a further $150B capex in DC build. So, we’re now at $300B. Then software needs to add value for customers at a 50% margin, so we get to double the $300B to $600B. This calculation was picked up by media and bloggers and repeated. Someone has to pay for all that capex, goes the argument. Well, yes, but… The doubling of the Nvidia cost for DC’s is simple but speculative – reality is that a DC cost is more around 50%-75% of the kit in the DC. For very expensive kit like AI, the DC build percentage will, potentially, be even lower. Also, Nvidia doesn’t just supply chips, it supplies complete DC solutions in racks that can be “dropped” into existing DC’s - Nvidia effectively builds part of the DC. Perhaps more important is the percentage of the Nvidia revenue that is going to clients that don’t need to make a profit, per se. Nvidia doesn’t release this number separately, so I’m speculating here, but I reckon this could be as high as 25% of their revenue. Jensen Huang is on the record as saying that a number of countries are building sovereign state AI factories, including: Japan, US, UK, India, Canada, France, Italy, Singapore and Malaysia (source: Bloomberg YouTube interview with Jensen). Going forward, all countries will have sovereign DC’s. Data is a national asset, like land, sea and the air above, maintains Jensen. Makes sense. Add into this mix the kit being sold into security services (CIA, FBI, MI5 etc). Then add the military spend. Then add in healthcare and other institutions that will use AI to increase efficiency and save money. Then add industry and enterprise AI that is not aiming for a 50% margin. My sense is that Sequoia may be out by a factor of two. That leaves us with 'only' $300B in revenue to find, mostly spread across the Mag-7, plus Oracle, Alibaba, Bytedance, X, Tesla et al. Just the Mag-7 do about 1.5 trillion in revenue, so an additional $300B should be findable. FInal point: not all margin from AI DC investment will be direct and may be lower that 50%. Apple, for example, will use AI to drive iPhone sales. Is there a feeding frenzy going on? Sure, and not all AI kit buyers will make money. Is the AI bubble about to ‘tip’? Hard to call. It’s easier to join up the dots looking back. #AI #artificalintelligence #NVIDIA #DC #Sequoia #thedigitalexecutive
To view or add a comment, sign in
-
If you can’t make the NVIDIA Global Conference this week in San Jose, you may want to at least read the Economist’s pre-conference piece. They give context and perspective to the players who are not just participating in but defining this era. At the heart of this transformational shift is Nvidia, whose valuation has soared to an eye-watering $2.3 trillion, thanks to its unparalleled AI chips. It was so nice to buy in at $500B, btw. Yet, it's not a solo race. Competitors like AMD and Intel are hot on its heels, vying for a piece of the AI pie. But there's more to this narrative. The Economist casts a spotlight on the rising leaders of AI model-making, such as OpenAI and @Anthropic, Mistral AI and Inflection AI, hinting at their burgeoning value. Meanwhile, it forecasts a promising horizon for cloud behemoths Amazon, Google, and Microsoft, suggesting they might emerge as the ultimate victors in this digital gold rush. This conference promises to be more than just an event; it's a glimpse into “now” and the near future of AI. If you can make the show, hit me up and I’ll fill you in. Rosebud Communications Fabian Baier https://lnkd.in/ep-2uhZQ
Just how rich are businesses getting in the AI gold rush?
economist.com
To view or add a comment, sign in
-
Nvidia: Dominating the AI Chip Market and Shaping the Future of Technology In the world of artificial intelligence, Nvidia stands as a titan. CEO Jensen Huang confidently assures investors that Nvidia will remain the gold standard for AI training chips, even as rivals strive to cut into their market share. With the much-anticipated rollout of the Blackwell system later this year, Nvidia’s lead in the AI chip industry seems unshakable. Why Nvidia Stays Ahead: - Dependable Tech for Big Players: From OpenAI to Tesla, industry giants rely on Nvidia semiconductors to power their large language and computer vision models. - Next-Gen Innovation: The Blackwell architecture, set to follow the renowned Hopper line of H100 chips, is expected to be a game-changer. Huang believes it could be the most successful product in computer history. - AI Ecosystem (CUDA): Nvidia’s Compute Unified Device Architecture (CUDA) has created a loyal base of developers. This ecosystem is as sticky as the Apple iOS, making it difficult for users to switch to other platforms. Economic Powerhouse: - Nvidia briefly surpassed Microsoft and Apple, becoming the world’s most valuable company, highlighting its impact not just on technology but on the global economy. Despite a recent dip, the long-term outlook remains robust due to its unparalleled AI training capabilities. Discussion Points: - Innovation vs. Competition: Can any company realistically challenge Nvidia’s dominance in the AI chip market? What would it take? - Future of AI: With the advancements in AI training chips, how do you see the future of technology evolving? What new possibilities excite you the most? - AI Ecosystem: How important is the CUDA ecosystem in maintaining Nvidia's lead? Could another platform ever compete? Join the Conversation! 💬 - Developers: Have you worked with Nvidia's CUDA? What has your experience been like? - Tech Enthusiasts: What are your thoughts on the future impact of AI training chips on industries like healthcare, finance, and automotive? - Investors: How do you view Nvidia’s market position in the long term? Are you optimistic about their continued growth? At LeadShutter.ai we are at the forefront of ai driven transformation. Learn how we utilise AI to transform our clients’ businesses and unlock new opportunities for growth. Join us as we navigate the impact of AI on business dynamics and shape the future of commerce. Explore more at LEADSHUTTER.ai | AI Automation Agency ( https://lnkd.in/eBNtSK_f) Nvidia's influence in AI isn't just about hardware—it's about an entire ecosystem that drives innovation and sets industry standards. Let's discuss how this tech giant continues to shape our world. 🚀 #Nvidia #AI #TechInnovation #FutureOfAI #CUDA #AIChips Article Link : https://lnkd.in/eDb_-zuu
To view or add a comment, sign in
Technical Team Lead - Data Hub | Mathephysineer 🇵🇸
5moMost companies in the tech industry are becoming more hungry for more computing just because they see huge "AI" models as the way forward. However, it will take some time before they realize that another approach is needed where we improve our algorithms, given the minimum computing possible. When they know that, it will mean going back to fundamentals, by then no one will be ready to start again.