I came across this TechCrunch piece about SoftBank and OpenAI’s new joint venture, and honestly, it made me pause. On the surface, “Crystal Intelligence” sounds like a typical international expansion play – OpenAI wants to sell enterprise AI tools in Japan, SoftBank knows the Japanese market, so they team up 50-50. Makes sense, right?
But here’s where it gets weird. SoftBank isn’t just OpenAI’s business partner here – they’re also a major investor in the company. So essentially, SoftBank is using money to fund a joint venture with a company they’ve already put money into. The article raises a pretty uncomfortable question: are these AI mega-deals actually creating economic value, or are we just watching money go in circles?
The Circular Money Problem
According to the article, this SoftBank-OpenAI deal is raising eyebrows precisely because of this circular investment pattern. When an investor funds a joint venture with their own portfolio company, you have to wonder – is this genuine business expansion or financial engineering?
Think about it from SoftBank’s perspective. They’ve already invested billions in OpenAI through various funding rounds. Now they’re putting up more money for a 50-50 joint venture. If Crystal Intelligence succeeds in Japan, SoftBank benefits twice – once as a JV partner and once as an OpenAI investor. If it fails, well, they’re essentially losing money they’re paying to themselves.
This reminds me of the WeWork saga, where SoftBank kept doubling down on their investment through increasingly complex deal structures. The fundamental question remains: are we seeing sustainable business growth or just creative accounting?
What This Says About AI Investment Sustainability
The TechCrunch Equity podcast discussion (featuring Kirsten Korosec, Anthony Ha, and Russell Brandom) apparently digs into what this signals about AI’s current investment model. I think they’re onto something important here.
The AI sector has seen unprecedented funding levels – OpenAI alone has raised over $13 billion across multiple rounds. But when investors start creating joint ventures with their own portfolio companies, it suggests the traditional expansion playbook might be running into limitations.
In the US market, OpenAI has enterprise customers, but expanding internationally requires local partnerships, regulatory navigation, and cultural adaptation. That’s expensive and risky. Rather than OpenAI funding this expansion independently (which would demonstrate true business sustainability), they’re essentially getting their existing investor to co-fund the effort.
Compare this to how other tech giants expanded internationally. When Google entered Asian markets, they funded it themselves. When Microsoft built Azure data centers globally, that came from their own cash flow. These AI deals feel different – more like financial engineering than organic business growth.
The Japan Market Context
To be fair, Japan represents a massive opportunity for enterprise AI tools. Japanese companies have been notoriously slow to adopt cloud technologies, but they’re increasingly interested in AI applications for productivity and automation. SoftBank’s local relationships and regulatory knowledge are genuinely valuable here.
But here’s what bothers me: if the Japanese market opportunity is so compelling, why couldn’t OpenAI fund this expansion from their existing capital base? They’ve raised enough money to fund several international expansions. The fact that they need SoftBank to co-invest suggests either capital constraints or risk management – neither of which screams “sustainable business model.”
Broader Implications for AI Economics
This circular investment pattern isn’t unique to SoftBank and OpenAI. We’re seeing similar structures across the AI ecosystem, where the same institutional investors appear in multiple layers of deal structures. Andreessen Horowitz, Sequoia, and other major VCs are investors in AI companies that then partner with other portfolio companies.
The article’s skepticism feels warranted. If AI technologies are as transformative as claimed, shouldn’t they be generating enough organic demand and cash flow to fund their own expansion? The reliance on complex investor-funded partnerships suggests the underlying unit economics might not be as strong as the hype suggests.
I’m also thinking about what happens when this music stops. If AI companies can’t eventually fund their own growth, and investors get tired of circular deal structures, who’s left holding the bag? Probably the enterprise customers who’ve built their operations around these AI tools.
My Take on Where This Goes
Honestly, I think we’re in the “financial engineering” phase of the AI boom. The technology is real and valuable, but the business models haven’t matured enough to support the valuations and expansion ambitions.
These circular investment structures remind me of the late-stage dot-com era, when companies started doing increasingly complex deals with their own investors and partners. It worked until it didn’t.
That said, I don’t think this means AI is a bubble waiting to burst. The underlying technology has genuine utility, especially in enterprise applications. But it does suggest we need to be more skeptical about AI deal announcements and focus on the actual economic fundamentals rather than the headline numbers.
The Crystal Intelligence venture might succeed in Japan – SoftBank knows how to navigate local business culture, and there’s real demand for AI productivity tools. But as an indicator of AI industry health, it’s more concerning than encouraging. When investors have to keep funding deals with their own portfolio companies, it suggests the path to self-sustaining growth is longer and more uncertain than anyone wants to admit.
This post was written after reading The circular money problem at the heart of AI’s biggest deals. I’ve added my own analysis and perspective.
Disclaimer: This blog is not a news outlet. The content represents the author’s personal views. Investment decisions are the sole responsibility of the investor, and we assume no liability for any losses incurred based on this content.
