BlackRock along with Microsoft, Global Infrastructure Partners, and Abu Dhabi-backed investment outfit MGX are launching a fund to invest in artificial intelligence (AI) infrastructure. The fund will initially seek $30 billion of private equity capital and eventually plans to raise as much as $100 billion.
While companies have been cutting down on their capex as well as headcount, AI has been an exception where not only are tech giants still investing billions of dollars but are also expanding their teams. Meanwhile, BlackRock is launching the mega-fund at a time when analysts are starting to get worried about a potential bubble building in AI. Here we’ll discuss more about the fund and analyze the landscape for AI companies after the Fed’s first rate cut since March 2020.
BlackRock Is Raising Billions of Dollars for AI Infrastructure
Earlier this week, BlackRock along with the above-mentioned partners announced the launch of the Global AI Infrastructure Investment Partnership (GAIIP) which will “make investments in new and expanded data centers to meet growing demand for computing power, as well as energy infrastructure to create new sources of power for these facilities.”
The bulk of these investments would made in the US while the remaining will be in “US partner countries.” Nvidia would also support the GAIIP by offering its expertise in data centers.
The release said, “This partnership will support an open architecture and broad ecosystem, providing full access on a non-exclusive basis for a diverse range of partners and companies.”
AI Infrastructure Would Need a Lot of Money
Notably, developing data centers requires a lot of money, including buying graphic processing units (GPUs.) Nvidia has been minting money selling these chips and the company is expected to post revenues of over $125 billion in the current fiscal year – over 11x what it did in the fiscal year 2021.
The electricity demand from these data centers has also skyrocketed and is expected to continue rise even further. In its report, Rystad Energy estimates that US data center power demand which is currently at 130 terawatt hours (TWh) will more than double to 300 TWh by 2030. A lot of new energy capacity would need to come up to meet that demand and the GAIIP fund is looking to back these projects.
AI companies, including ChatGPT’s parent OpenAI, have raised billions of dollars to develop AI infrastructure. OpenAI has raised around $10 billion so far with the bulk of it coming from Microsoft which last year committed to invest billions of dollars in the company spread over multiple years.
OpenAI, which was valued at $80 billion earlier this year, is said to be seeking fresh funds that could value the company at $150 billion. The company reportedly hit an annualized revenue run rate of $3.4 billion in May which was up 580% YoY.
Sam Altman Was Seeking Up to $7 Trillion
OpenAI’s CEO Sam Altman believes that the company would need to raise $100 billion as it strives to achieve “artificial general intelligence” – which means an AI model with human-like intelligence that can learn and improve itself. If that number sounds outrageous, Altman was also reportedly in talks with global investors to raise $7 trillion to expand the chipmaking facility to meet the growing demand from AI.
Here it is worth noting that the money GAIIP is raising is separate from OpenAI’s capital raise even though the two could collaborate in the future.
xAI is now valued at ~1/3rd the valuation of OpenAI, while only being 15 months old
Elon Musk and xAI just announced the largest series B raise of all time at $6B, to train new versions of Grok
And they told investors that they're building a massive supercomputer with…… https://t.co/8xx7rcKgqz pic.twitter.com/cojC9dV0K3
— Rowan Cheung (@rowancheung) May 27, 2024
Meanwhile, the valuations of AI companies might appear a bit frothy as OpenAI is now seeking funds at almost twice its February valuation and over five times its valuation in Q1 2023. But then, earlier this year, Elon Musk’s AI startup xAI raised $6 billion at a mammoth $24 billion valuation which would appear even more frothy given that it is a much newer entrant in the AI space that hasn’t proved itself yet.
Is AI a Bubble?
Naturally, analysts are looking at these valuations and comparing them with returns on AI products and most don’t like what they see, ringing alarm bells about a potential bubble. To be sure, barring chip companies – predominantly Nvidia and its supplier Taiwan Semiconductor Manufacturing Company – not many companies have much tangible benefits to show from AI.
Much of the current spending towards AI is going towards building LLMs (large language models) and companies – big and small – are lining up to buy Nvidia’s high-power chips. Nvidia’s CFO Colette Kress dispelled fears that AI investments were not transforming into realized revenues for its customers. During the fiscal Q1 2025 earnings call earlier this year, she said, “For every $1 spent on NVIDIA AI infrastructure, cloud providers have an opportunity to earn $5 in GPU instant hosting revenue over four years.” Nvidia CEO Jensen Huang has also echoed similar views.
However, revenues of companies creating AI products like Microsoft and Alphabet aren’t taking off in a big way despite the massive investments that they are making in AI. While tech companies sound overtly optimistic about AI during the earnings calls (without divulging hard financial data), The Information reported that not even 1% of Microsoft 365 subscribers opted for its AI-powered “Copilot” products. It’s hard to imagine that it will be able to make back the billions it has spent on developing its AI tech.
At some point, big tech/LLM companies like OpenAI must charge the actual cost of delivering generative AI – and I believe there is a subprime AI crisis brewing where the entire industry is dependent on discounted and unsustainable prices.https://t.co/nzxKRVwf4v pic.twitter.com/b62kSpMgKh
— Ed Zitron (@edzitron) September 16, 2024
Apple Launched Ai-Powered iPhone 16
On a similar note, the initial response to Apple’s Apple Intelligence-powered iPhone 16 hasn’t been as positive as some experts expected. To be fair, Apple Intelligence isn’t even available yet so it could be likely that some customers might be waiting for more details on that feature before jumping to buy the latest iPhone.
That said, while there is significant euphoria among companies towards AI, so far there isn’t much to suggest that consumers also share the same optimism.
While some observers believe the AI industry is forming a bubble, drawing parallels to the dot-com days of the 1990s, others argue that the pivot toward AI looks real. AI is not only helping companies cut down on costs structurally and serve customers more efficiently but is also an opportunity to increase revenues. However, it would take time for companies to monetize the investments and if investors get impatient, they might need to rethink their burgeoning AI capex.
An AI Price War Cannot Be Ruled Out
AI companies are starting to worry that a price war on LLM products and services is likely. Earlier this year, OpenAI launched a cost-efficient GPT-4o mini which is 60% cheaper than the GPT-3.5 Turbo
The AI price war may have already begun in China where Alibaba slashed the prices for its Qwen LLM models by as much as 97%. Rival Baidu went a step further and said it would offer enterprise customers two of its Ernie LLM models for free.
The price war was initiated by TikTok’s parent company ByteDance by offering access to its LLMs at less than 1% of GPT-4.
We saw something similar in the electric vehicle industry where slowing growth led to a brutal price war which eventually led to many bankruptcies in the sector.
Fed’s Rate Cut Would Make Capital Raising Cheaper
Earlier this week, the Fed lowered rates by 50 basis points – a quantum leap that it usually reserves for severe downturns. However, as the Fed embarks on its rate cut cycle, capital raising should become cheaper and easier. While raising money wasn’t anyways much of an issue for AI companies despite the startup funding winter, things could get even rosier with the Fed’s rate cuts.
The rate cuts also bode well for the mega investment that GAIIP is seeking. However, it remains to be seen whether the fund can generate good long-term returns for investors given the uncertainty over how much return on investments can companies make on their AI capex.