AI: Weekly Summary. RTZ #528
…week ending November 1, 2024
Big Tech Reiterates AI Capex Commitments in Earnings Calls: Earnings season kicked off in earnest, with Big Tech companies continuing ‘pedal to the metal’ on AI capex spending. More AI GPUs and data centers as far as the eye can see. With huge helpings of Power from all sources, including the previously taboo Nuclear option. Even the private equity industry is jumping in with both feet, as KKR announces a $50 billion commitment. Not to mention Blackstone and Blackrock earlier. Softbank’s Masayoshi San announced a $9 trillion figure for global AI data center and power spend over the next decade, with Nvidia being a key beneficiary. It was an obvious attempt to outdo OpenAI Sam Altman’s $7 trillion number a few months ago, which has since been adjusted down to a few hundred billion for the industry. Nevertheless, investors for now are more or less taking the big public tech companies’ AI capex in stride. More here on 100,000+ AI GPU Data Center race.
OpenAI widens ChatGPT Search Launch: OpenAI broadened its ChatGPT Search feature with more features and across more of their product tiers. Their strategy is to insert the Search capability directly into ChatGPT used by hundreds of millions both directly via paid and free tiers, as well as via APIs. This as Google also widens AI Search Overviews with its Gemini AI models aggressively across 100+ countries. This highlights Google’s AI Infrastructure and Compute advantage vs OpenAI and Microsoft having to share and allocate their AI GPU resources. The broader point is that AI Search is being expanded by the two major players, as well as Meta with Meta AI, Perplexity, Elon’ xAI Grok, and a host of other AI companies. Of course the concurrent challenge and opportunity here is how to make these AI Search capabilities also attractive to advertisers. More here.
Google holds its own in ‘AI Race’: Despite Wall Street concerns over AI impacting Google’s core Search businesses, Alphabet reported strong results at Google Cloud, Search, YouTube and other businesses. Google in particular has reduced the cost of AI answers in Search queries, leveraging its immense AI Compute infrastructure. This has been a long-time concern for investors. In addition, CEO Sundar Pichai emphasized that Google’s perceived AI expertise is drawing more enterprise customers to Google Cloud relative to its peers. The deeper dives were generally positive, especially with Google Cloud. Also, there is growing recognition that Google has the same opportunities as Meta on using AI Tech to leverage Advertising capabilities for Advertisers of all types. More here.
Meta leans in on AI & ‘Meta AI’ strategy: Meta on the other hand did raise Wall Street concerns with its results, especially the company’s rising AI capex plans. In particular, The company particularly highlighted its focus on bigger AI data centers to support Meta AI and other services over its 4 billion plus user base across Instagram, Facebook, Messenger and WhatsApp. Ironically, the company is better positioned on monetizing its AI investments than peers given its laser focus on developing AI tools and capabilities for advertisers writ large. Also, as explained above, the company is focused on developing an AI Search Engine to lessen reliance on peers like Google, Microsoft and others. Meta also continues to benefit from its open source leadership with Llama 3+ LLM AI models that give it opportunities with AI Cloud services providers. More here.
Apple launches AI (Apple Intelligence) optimized Mac hardware: As expected last week, Apple introduced a host of Mac computers to leverage its expanding Apple Intelligence driven applications and services, including its partnership with OpenAI & ChatGPT. A worthwhile video watch is this WSJ interview with Apple Software head Craig Federighi for a lot of the underlying Apple strategy, and how they want to differentiate Apple in AI vs its peers. The early Apple Intelligence features are capable, but not yet world changing. But there’s a specific roadmap of additional features to come. The beefier AI optimized hardware in Macs, iPhones and iPads, as well as the Apple wearables will make these Ai services more mainstream over time. More here.
Other AI Readings for weekend:
Taiwan Semi (TSMC), the company that actually makes most of the world’s chips (AI and otherwise), is feeling geopolitical stresses. More here.
Notable AI video startup Decart apparently has a very differentiated tech approach to real time AI video generation vs OpenAI Sora et al. Especially in video games.
Up next, the Sunday ‘The Bigger Picture’ tomorrow. Stay tuned.
(NOTE: The discussions here are for information purposes only, and not meant as investment advice at any time. Thanks for joining us here)