Hello and welcome to Eye on AI. In this category… Big Tech is increasing the use of AI… OpenAI is reducing its basic ambitions… Wall Street is developing AI infrastructure… and AI models are providing inaccurate voting information.
Quarterly earnings are at Microsoft, Meta, and Google, and Amazon is on track tonight. Those who have reported so far have made one thing clear: they are about to increase spending on AI even more. Of course, most of this investment will go to infrastructure such as Nvidia chips and data centers. Nvidia has yet to announce its Q3 earnings for another month, but it will be another telling one to watch for the AI boom in the AI boom.
Meta
Meta said it expects capital spending to continue to grow significantly in 2025 as it accelerates infrastructure investment. The company raised its 2024 spending guidance to between $38 billion and $40 billion, from $37 billion to $40 billion.
“Our AI investments continue to require critical infrastructure,” CEO Mark Zuckerberg said on an earnings call yesterday.
Meta is currently working on the next version of its Llama model and continues to position itself as an AI-all-in-one company while relying on ad revenue from social media to foot the bill. To close the gap, Zuckerberg pitched AI as a winner for its core ad business. He said its advances in AI make it possible to better predict what content users want to see, leading to an 8% and 6% increase in time spent on Facebook and Instagram, respectively.
Microsoft
Microsoft has posted positive growth in its AI business, reporting that it is on track to surpass $10 billion in annual revenue in the current quarter and will be its fastest product category to reach that milestone yet. Additionally, the company disclosed details of its investment in OpenAI for the first time in a quarterly regulatory filing, painting a picture of big losses on the other side of its AI bet.
In the filing, Microsoft said it has made $13 billion in total commitments to OpenAI. This, however, did not include an additional $750 million that the company reportedly invested in OpenAI’s latest funding round in early October (the filing quarter ended on September 30).
Apart from its overall investment, Microsoft also disclosed the amount it is paying to bankroll OpenAI’s profits. The company said losses from its stake in OpenAI were part of the reason it had a $683 million charge on its equity investment in the quarter.
In AI, Google revealed how the boom has boosted its cloud business. But more interestingly, CEO Sundar Photosi said that AI writes more than 25% of new code in the company, but engineers still review the work before using it.
While some AI-related earnings statements from companies were somewhat predictable, I would have liked to hear more details about the AI program. Google didn’t answer my questions about what this looks like going forward, including what types of coding projects AI has found to be most (or least) suitable and if it’s shipping anything coded with AI yet. AI is not limited to the product, but also makes the products too. In future earnings calls, will we hear that the increased efficiency of AI coding (or the layoffs of expensive engineers) increases profits?
And with that, here’s some AI news.
Sage Lazzaro
[email protected]
sagellazzaro.com
How will your business adapt in the face of a changing world? Join global executives and policy leaders at the Fortune Global Forum in New York City Nov. 11-12 to discuss AI, the future of democracy, growing international tensions, and more. Request your invitation here.
AI IN NEWS
OpenAI is working with Broadcom and TSMC to develop its own AI chips. Reuters reported that OpenAI abandoned its ambitions to create its own infrastructure—the name of the semiconductor industry—because of the enormous cost and time it would have required. Instead, OpenAI has been working with Broadcom to build its first AI chip focused on thinking for several months and, through Broadcom, has secured manufacturing capacity with Taiwan Semiconductor Manufacturing Company (TSMC). The AI company plans to make its first customized chip in 2026, but sources told Reuters that timeline could change. OpenAI is also tapping commercially available chips from AMD, continuing its plan to reduce its reliance on Nvidia.
AI and data center projects get a $50 billion boost from Wall Street. Investment giant KKR and private equity group Energy Capital Partners are working with tech companies to accelerate their access to the energy needed for AI and will make a $50 billion investment over the next four years. Doug Kimmelman, founder and senior partner at Energy Capital Partners, said the wall A road journal that “[natural] gas will be at the forefront of this.” The insatiable energy needs of AI continue to be one of the biggest problems with the technology. In the past few weeks, Amazon, Google, and Microsoft have all announced investments in nuclear power as they look for an environmentally friendly way to power their AI.
An AI designed to identify people who need help paying their water bills is offering a huge discount to one of the city’s wealthiest customers. The Portland, Ore., water office has tested a machine learning algorithm designed to predict customers’ ability to pay and offer discounts to those who struggle. One of the first customers the algorithm targeted, however, was Columbia Sportswear CEO Tim Boyle, who said he received a notice that he was being offered a 40% discount even though he had no problem paying his water bill. Boyle earned more than $1.6 million last year and owns more than $1.7 billion worth of Columbia stock. “I don’t want a discount – they should give it to someone who needs it,” Boyle told the media. Willamette Weekly. The Portland Water Bureau has approved a $350,000 pilot contract with contractor SERVUS.
GOOD LUCK WITH AI
Super Micro’s shares rose 3000% on the AI wave – then the auditor quit, saying he didn’t trust the management. —by Amanda Gerut and Sharon Goldman
Elon Musk says there is a 10% to 20% chance that AI ‘goes wrong,’ even as he raises billions to launch his xAI. — by Marco Quiroz-Gutierrez
A Chanel boss went to Microsoft HQ and asked ChatGPT to show him a picture of his company’s leadership. They were all men in suits — by Sasha Rogelberg
AI CALENDAR
Nov. 19-22: Microsoft Ignite, Chicago
December 2-6: AWS re:Invent, Las Vegas
December 8-12: Neural Information Processing Systems (Neurips) 2024, Vancouver, British Columbia
December 9-10: Fortune Brainstorm AI, San Francisco (register here)
EYE NUMBERS IN AI
50%
That’s about how often leading AI models answer questions about voting with incorrect information, according to assessment from evidentiary mattersa non-profit news agency that tested Anthropic’s Claude 3 Opus, Google’s Gemini 1.5 Pro, OpenAI’s GPT-4, Meta’s Llama 3, and Mistral’s Mixtral 8x7B v0.1. Proof News also found a gap in accuracy by language: 52% of answers to questions asked in Spanish contained incorrect information compared to 43% of answers to questions asked in English. Folks, just get your voting information from official sources.
