Here’s a rather eye-watering bit of news from the world of big tech bets: SoftBank Group, under the ever-bold leadership of Masayoshi Son, is reportedly gearing up for another colossal splash, this time squarely aimed at the foundational layers of the AI revolution. Forget funding startups for a minute; Son wants SoftBank to become a leading Artificial Super Intelligence (ASI) platform provider. And he wants to do it by around 2035. Yes, around 2035.
That’s a deadline that might make even the most frenetic Silicon Valley founder gulp. Son, according to recent reports, laid out a vision to shareholders that isn’t just ambitious; it’s positively gargantuan. Reports indicate ambitions for massive investments into computing infrastructure, with plans including projects on the scale of the proposed $1 trillion AI computing complex. Where’s this money going? Largely focused on securing substantial compute capacity, involving investments in infrastructure like data centers (potentially via partnerships) and acquiring high-end semiconductors and systems essential for advanced AI – the golden tickets of the AI age.
Masayoshi Son’s Audacious ASI Ambition
Mr Son isn’t known for thinking small. The man who founded SoftBank and built the sprawling empire, including the dizzying, sometimes baffling, heights of the SoftBank Vision Fund, has a new fixation: Artificial Super Intelligence. Now, definitions of ASI vary, but generally, it implies AI far surpassing human cognitive abilities. Son isn’t just interested in investing in companies *using* AI; he wants SoftBank to provide the very bedrock – the computational power – that makes such advanced AI possible. He explicitly stated a goal to become a top provider of ASI compute platforms by around 2035.
This isn’t pocket change, nor is it a gradual ramp-up. Reports indicate massive investment into compute capacity, signifying an accelerated, intense period of SoftBank AI investment. Imagine the sheer scale of infrastructure required to power something approaching ASI. It needs data centres, power, cooling, and critically, vast quantities of high-end chips. And right now, chips capable of powering leading AI are highly sought after. Acquiring enough high-end computing systems to support this vision in the current supply-constrained market is a challenge in itself, even before you consider the cost.
Funding the Future of Compute: Vision Fund vs. Infrastructure
For years, the world watched the SoftBank Vision Fund spray billions across the tech landscape, investing in everything from ride-sharing to real estate platforms. That model was about placing bets on *applications* and *disruptors* powered by technology. This new move feels fundamentally different. It’s a direct investment in the underlying *infrastructure*. It’s like shifting from funding airlines to buying up all the airports and airspace.
Where will the immense capital come from? While details remain somewhat opaque, reports indicate SoftBank is exploring options, including potential collaborations. One significant possibility mentioned is a SoftBank joint venture to help fund the scale of this endeavor, with sources mentioning exploring various partnerships. A partnership like that could certainly bring the necessary financial firepower to the table for this scale of investing in AI servers and data centres. It makes strategic sense too; building global compute infrastructure requires vast capital and potentially global partners.
Why Bet Everything on ASI Compute?
Why this pivot? Why focus on being an ASI platform provider rather than just continuing to back AI startups? Perhaps Son sees the computational layer as the new bottleneck and, therefore, the most valuable part of the future AI stack. If you control the ultimate resource – the raw compute power needed for cutting-edge AI development – you hold immense leverage. Every company building advanced AI, whether for drug discovery, materials science, or something we haven’t even dreamt of yet, will need access to this kind of supercompute. SoftBank wants to be the landlord.
This strategy also hedges against picking individual AI winners. Instead of hoping one of your portfolio companies develops the killer ASI application, you simply provide the engine that *all* potential ASI developers will need. It’s a powerful position, assuming, of course, that Artificial Super Intelligence is indeed the next frontier and achievable within a reasonable timeframe, and that others don’t beat you to building the necessary infrastructure.
The Challenges and the Clock
Let’s be frank, a around 2035 target for becoming a *leading* ASI compute provider is incredibly aggressive, though perhaps more realistic than a 2025 goal. Building the necessary infrastructure, securing massive quantities of cutting-edge computing systems, and scaling operations to support world-leading ASI development isn’t trivial. It requires engineering expertise, operational excellence, and deep pockets. SoftBank certainly has the latter, but executing on this scale is a monumental task.
It also raises questions. What exactly does “leading ASI platform provider” mean? Is it measured by total compute capacity, the sophistication of the platform tools offered, or the calibre of the AI models being trained on their infrastructure? And how does this tie into Arm Holdings, the chip design firm SoftBank controls? Arm is foundational to many future computing architectures, including those being explored for AI, and its technology is key to providing the compute infrastructure needed, although the major investment discussed appears to be in acquiring systems rather than solely developing core chip architecture within Arm.
What Does This Mean for the AI Race?
SoftBank’s potential move to aggressively build out compute infrastructure throws another interesting variable into the global AI race. It underscores the increasing centralisation of power around those who control the most advanced hardware. While cloud providers like AWS, Azure, and Google Cloud are already major players here, SoftBank aiming for a dedicated “ASI platform” suggests a potentially different kind of offering, perhaps more focused purely on the bleeding edge of AI compute rather than general cloud services.
It also highlights the continued strategic importance of hardware manufacturers like Nvidia, among others producing high-end semiconductors. Demand remains insatiable, and companies like SoftBank are willing to make enormous, long-term capital commitments just to secure supply. This isn’t just about buying technology; it’s about securing a fundamental resource for the future economy.
So, while we await more concrete details on funding and execution, Son’s latest pronouncements serve as a stark reminder of the sheer scale of investment and ambition pouring into the pursuit of advanced AI. SoftBank is making a massive bet, shifting its focus from picking startup winners to building the very foundation upon which those winners might be built.
What do you make of SoftBank’s pivot to infrastructure? Is aiming for ASI leadership by around 2035 realistic? Share your thoughts in the comments below!
***
Disclaimer: This analysis is based on publicly available news reports and discussions surrounding SoftBank’s recent announcements and does not represent financial advice. As an AI expert analyst, I aim to provide insightful commentary on technological and business trends.
