Rafael Brown’s Post

“David Cahn, an analyst with Sequoia Capital, believes that AI companies will have to earn about $600 billion per year to pay for their AI infrastructure, such as datacenters.” But current AI revenue is likely in the single digit billions, not the triple digit billions that are needed. “Cahn's math is relatively simple. First, he doubles Nvidia's run-rate revenue forecast to cover the total AI data center hashtag #datacenter costs (GPUs are half; the rest includes energy, buildings, and backup generators). Then, he doubles that amount again to account for a 50% gross margin for end-users, such as startups or businesses buying AI compute from companies like AWS or Microsoft Azure, which must make money, too.” As the figure for this post shows, these rough calculations give us big numbers. Nvidia has estimated annual revenues for data center customers of $150 billion based on the current quarter. Doubling that number gives us $300 billion in revenue needed for AI cloud providers to justify their $150 billion purchases of Nvidia processors. Doubling it again gives us $600 billion in revenue needed for AI software companies to justify their $300 purchases of AI cloud services. The problem is that the revenues for the AI industry are still very small. The biggest revenues are likely for OpenAI, which uses Microsoft's Azure infrastructure. It “has seen a substantial increase in revenue, from $1.6 billion in late 2023 to $3.4 billion in 2024. This growth underscores OpenAI's dominant position in the market, far outpacing other startups that are still struggling to reach a $100 million revenue mark. Yet those several billion are no way near what is needed, even if other AI software companies or end users are included. “Even optimistic projections for major tech companies' AI revenues fall short, Cahn says. Assuming Google, Microsoft, Apple, and Meta each generate $10 billion annually from hashtag #AI and other companies like Oracle, ByteDance, Alibaba, Tencent, X, and Tesla generate $5 billion each, there remains a $500 billion gap.” Now that is a big number. The analysis is rough, but it captures the order of magnitude differences between what is needed for AI software revenues and what currently exists. hashtag #technology hashtag #innovation hashtag #hype hashtag #ethics hashtag #startups hashtag #artificialintelligence hashtag #shareprices hashtag #marketbubble https://lnkd.in/gW-tgQTX

“David Cahn, an analyst with Sequoia Capital, believes that AI companies will have to earn about $600 billion per year to pay for their AI infrastructure, such as datacenters.” But current AI revenue is likely in the single digit billions, not the triple digit billions that are needed. “Cahn's math is relatively simple. First, he doubles Nvidia's run-rate revenue forecast to cover the total AI data center hashtag #datacenter costs (GPUs are half; the rest includes energy, buildings, and backup generators). Then, he doubles that amount again to account for a 50% gross margin for end-users, such as startups or businesses buying AI compute from companies like AWS or Microsoft Azure, which must make money, too.” As the figure for this post shows, these rough calculations give us big numbers. Nvidia has estimated annual revenues for data center customers of $150 billion based on the current quarter. Doubling that number gives us $300 billion in revenue needed for AI cloud providers to justify their $150 billion purchases of Nvidia processors. Doubling it again gives us $600 billion in revenue needed for AI software companies to justify their $300 purchases of AI cloud services. The problem is that the revenues for the AI industry are still very small. The biggest revenues are likely for OpenAI, which uses Microsoft's Azure infrastructure. It “has seen a substantial increase in revenue, from $1.6 billion in late 2023 to $3.4 billion in 2024. This growth underscores OpenAI's dominant position in the market, far outpacing other startups that are still struggling to reach a $100 million revenue mark. Yet those several billion are no way near what is needed, even if other AI software companies or end users are included. “Even optimistic projections for major tech companies' AI revenues fall short, Cahn says. Assuming Google, Microsoft, Apple, and Meta each generate $10 billion annually from hashtag #AI and other companies like Oracle, ByteDance, Alibaba, Tencent, X, and Tesla generate $5 billion each, there remains a $500 billion gap.” Now that is a big number. The analysis is rough, but it captures the order of magnitude differences between what is needed for AI software revenues and what currently exists. hashtag #technology hashtag #innovation hashtag #hype hashtag #ethics hashtag #startups hashtag #artificialintelligence hashtag #shareprices hashtag #marketbubble https://lnkd.in/gW-tgQTX

Really interesting to see these number emerge from this post of @Dr. Jeffrey Funk and Sequoia. Whatever the real numbers may prove to be, this is only one side of the value generated and lost. It is also important to note, that all this flurry of new tools and assistants will need to be repaid, sooner rather than later. The marketing from these tech champs is about helping you to improve your life, but might it just be, that we are again the product (personal data transfer, very personalised ads) and many of these investors in search of returns will skip the ethical part of AI? Thst they may resort to data brokerage or surveillance? Who knows, but I am bracing myself for another wave of unethical usage of my personal data. #ethicalai

“David Cahn, an analyst with Sequoia Capital, believes that AI companies will have to earn about $600 billion per year to pay for their AI infrastructure, such as datacenters.” But current AI revenue is likely in the single digit billions, not the triple digit billions that are needed. “Cahn's math is relatively simple. First, he doubles Nvidia's run-rate revenue forecast to cover the total AI data center #datacenter costs (GPUs are half; the rest includes energy, buildings, and backup generators). Then, he doubles that amount again to account for a 50% gross margin for end-users, such as startups or businesses buying AI compute from companies like AWS or Microsoft Azure, which must make money, too.” As the figure for this post shows, these rough calculations give us big numbers. Nvidia has estimated annual revenues for data center customers of $150 billion based on the current quarter. Doubling that number gives us $300 billion in revenue needed for AI cloud providers to justify their $150 billion purchases of Nvidia processors. Doubling it again gives us $600 billion in revenue needed for AI software companies to justify their $300 purchases of AI cloud services. The problem is that the revenues for the AI industry are still very small. The biggest revenues are likely for OpenAI, which uses Microsoft's Azure infrastructure. It “has seen a substantial increase in revenue, from $1.6 billion in late 2023 to $3.4 billion in 2024. This growth underscores OpenAI's dominant position in the market, far outpacing other startups that are still struggling to reach a $100 million revenue mark. Yet those several billion are no way near what is needed, even if other AI software companies or end users are included. “Even optimistic projections for major tech companies' AI revenues fall short, Cahn says. Assuming Google, Microsoft, Apple, and Meta each generate $10 billion annually from #AI and other companies like Oracle, ByteDance, Alibaba, Tencent, X, and Tesla generate $5 billion each, there remains a $500 billion gap.” Now that is a big number. The analysis is rough, but it captures the order of magnitude differences between what is needed for AI software revenues and what currently exists. #technology #innovation #hype #ethics #startups #artificialintelligence #shareprices #marketbubble https://lnkd.in/gW-tgQTX

◾Tech-Talks - 04◾ Title: Unveiling the AI Investment Landscape: Big Tech's Strategic Moves and Revenue Dynamics 🚀💡 🚀 Exciting Developments in AI Investments by Big Tech! 🌐💡 In the fast-paced world of AI startups, major players like Amazon, Microsoft, Google, and NVIDIA have been making significant waves. Over the past year, more than 65 investments have been recorded, with eye-catching numbers and strategic considerations. 📈💰 Key Highlights: Microsoft's whopping $10B investment in OpenAI AWS and Google's combined investment of up to $4B in Anthropic NVIDIA, Microsoft, and Google contributing to a $1.3B round in Inflection NVIDIA's $400M investment in Coreweave Google and NVIDIA backing Runway with a $141M round The rationale behind these investments is diverse and strategic, touching on aspects like differentiated models, influencing market structure, acquiring scaled users of infrastructure, and potentially locking in certain customers. 💻🔄 🤔 Controversial Topic: Are these investments a form of "buying revenue"? The post delves into this, drawing parallels with Palantir's unique approach in the SPAC boom and exploring how cloud providers may be influencing revenue growth through these investments. 📊💡 💼 Cloud providers are leveraging partnerships to drive revenue, but the specifics on accounting and the actual impact remain undisclosed. The post emphasizes the need for better reporting and transparency as the trend of such investments continues to grow. If you're interested in the intersection of technology, business, and strategic investments, this post is a must-read! 🚀📖 #AI #TechInvestments #BigTech #CloudProviders #StrategicInvestments #BusinessInsights Article nourished by; 👤Devarsh Mehta

A $500 billion hole in the AI market. So what should we do about it? Corporations and startups are investing huge amounts of money in infrastructure to develop their AI services. It is a fact. How much should these AI services earn to recoup their costs? After all, this is not all done out of love for art; business must bring in money. David Kahn from the Sequoia Foundation tried to answer this question. He made his first calculations in September last year (https://lnkd.in/dUa5QtMk). You can read the details, as he believed, at the link. In short, he took the publicly known costs of chips, added electricity costs, threw in a pinch of company margins, and the result was... $200 billion. Yes, that’s how much all AI companies and AI services of corporations, like Google, should earn per year and Microsoft to recoup the investment. The problem is that he only counted $75 billion in revenue. Where is the other $125 billion? This is the main mystery. About nine months passed and David decided to repeat his calculations (https://lnkd.in/dHFF7R8u). And this time it turned out that in the fourth quarter of this year, the annual revenue of AI companies and services should already be $600 billion. This is ideal. But here, as you understand, we have a problem again. Yes, OpenAI itself has significantly increased its revenue, others have caught up. But Kahn counted only $100 billion in revenue. That is, the hole has reached $500 billion! One could argue that they are investing in infrastructure for a long time, so they can fight back for a long time. Kahn writes that he partly agrees. But there are nuances. Chips lose value quite quickly because new, more efficient ones appear. Plus, great competition forces services to reduce prices to the limit. The conclusion is simple - not all companies will survive this hype around AI. And investors will simply burn their money by betting on the wrong player. But this is a necessary evil with such large-scale changes in the market. But for the founders, as Kahn writes, everything is not so bad. They will accumulate experience and create better AI products. #productmanagement #management #marketing #customerservice #hr #hiring #UX #startup #web3 #tech #growth #digital #Growthhacking #projectmanagement #blockchain #UI #startup #fintech #agile #edtech #business #leadership #analytics #Productdevelopment #design #iot #ai #researchanddevelopment

Artificial intelligence & machine learning startups are still red hot, ending Q4 with $22.3 billion in deal value across 1,665 deals, according to PitchBook’s latest Emerging Tech Research report. Powering these startups are AI-specific datacenters, whose specialized infrastructure and operations are designed to meet the rigors of AI workloads. A few upstarts in the AI datacenter space are emerging, looking to challenge the likes of hyperscalers such as Microsoft, Google, and AWS. The Q4 2023 AI & ML report spotlights the startups taking on Big Tech and other emerging opportunities in the space that investors need to know.

This image from Alvin Foo , and this is what GPt4 said about it This image is an infographic titled "AI Investing Ecosystem." It outlines the various layers and components involved in the AI industry, which are critical for investors to understand. The infographic is divided into four main layers: 1. Application Layer: This emphasizes that venture capitalists (VCs) can invest in AI applications and that public investors can look for existing companies releasing new products. 2. Models Layer: It lists OpenAI and other companies, noting that most models are developed by large tech firms or private companies, and that open-source models will be important. 3. Data Layer: This includes companies like AWS, Google Cloud, and Palantir, and highlights the significance of data processing, storage, and management in AI. 4. Compute Layer: It features companies like AWS, NVIDIA, and Azure, mentioning that most AI computation is done on GPUs, and that public cloud and datacenters house the majority of this compute. 5. Hardware Layer: This features companies such as AMD, Cisco, and Intel, pointing out the importance of hardware in enabling computing power for AI. The bottom of the infographic includes a watermark for "@AITechInvesting" and the website "aitechinvesting.com," suggesting that the source is an entity focused on AI technology investments. - #AIInvesting - #VentureCapital - #ArtificialIntelligence - #TechStartups - #CloudComputing - #DataManagement - #MachineLearningModels - #AIHardware - #SemiconductorIndustry

Things That Make You Go Hmm🤔: OpenAI CEO Sam Altman seeks as much as $7 trillion for new AI chip project! OpenAI CEO Sam Altman is seeking trillions of dollars in investments to overhaul the global semiconductor industry, The Wall Street Journal reported. Altman has long talked of the supply-and-demand problem with AI chips — many AI giants want them, but there aren't enough to go around — and that it limits OpenAI's growth. He's considering a project that would increase global chip-building capacity, according to a Thursday evening report in The Wall Street Journal, and is reportedly in talks with different investors, including the government of the United Arab Emirates. Altman could need to raise between $5 trillion and $7 trillion for the endeavor, The Wall Street Journal reported, citing one source. CNBC could not confirm the number. OpenAI did not respond to a request for comment. On Wednesday, Altman posted on X that OpenAI believes "the world needs more ai infrastructure--fab capacity, energy, datacenters, etc--than people are currently planning to build." He added that "building massive-scale AI infrastructure, and a resilient supply chain, is crucial to economic competitiveness" and that OpenAI would try to help. $7 trillion? Hmm🤔. Reminds me of a Kanye West line from a song "no one man should have all that power". The fact that chipmaker Nvidia surpassed Amazon and Google owner Alphabet, in market capitalisation and catapulted it to fourth place on the list of the most valuable companies in the US makes Sam Altman's argument even more interesting. Nvidia's market capitalization came to $1.83 trillion, narrowly surpassing Alphabet’s $1.82 trillion and Amazon’s $1.8 trillion. However, Nvidia Corp. Chief Executive Officer Jensen Huang anticipates that advances in computing over the next few years will keep the cost of developing artificial intelligence well below the $7 trillion that Sam Altman is said to be fundraising. So will the new battle be between Nvidia and Sam's audacious capital raise? Double Hmm... #ai #chips #samaltman #openai https://lnkd.in/dScc_FYN

𝐈𝐬 𝐭𝐡𝐢𝐬 𝐀𝐈 𝐦𝐚𝐝𝐧𝐞𝐬𝐬, 𝐡𝐲𝐩𝐞 𝐨𝐫 𝐰𝐡𝐚𝐭…… Startup CoreWeave, specialized in GPU computing in the cloud has secured a 💲1.1 billion in new funding. According to Crunchbase data they have now raised a total of 4.6 billion💰💰💰💰 According to a news release: “Organizations of every size and industry are racing to deploy larger-scale AI solutions and capabilities – driving explosive demand for powerful and scalable infrastructure. “ 👉 https://lnkd.in/dzvedqTB I am not so sure about that statement, I see many organizations still struggling with there basic data. Getting proper product master data or customer data to even start with #datascience, euuhm I mean #AI. Or having issues to deploy a simple random forest model. Will coreweave ever become profitable? Investors have put 4.6 billion into the company. You can spin up GPU’s for around $ 2.50 an hour. That would mean companies have to spent more than 210.000 years on Coreweaves’s GPU’s in the cloud to earn back that investment, ignoring all other costs…… #AIbubble…. #TulipMania