Day: July 29, 2026

Across Japan, corporations are grappling with a familiar tension: promising innovation initiatives that stall before they scale, not for lack of merit, but because rigid organisational structures move slower than the market opportunities in front of them. For a growing number of Japanese enterprises, the response has been to look outward, toward ecosystems agile enough to move at startup speed while stable enough to be trusted with real business problems.

Singapore has become one of the clearest answers to that need. It combines advanced technological capability with a predictable regulatory environment and an English-language business culture, an unusual combination that lets cross-border collaboration happen without the friction that typically slows first engagements between Japanese corporates and international startups.

This is the backdrop to the Boundary Spanning Service, a structured programme developed by Sony Acceleration Platform will connect Japanese corporations with Singapore-based startups capable of addressing specific innovation challenges. For founders in Singapore, understanding this shift, and how to engage with it, is becoming increasingly relevant.

It is also a shift worth watching closely, because it says something about where Singapore’s startup ecosystem sits in the region’s broader innovation map. Japanese corporates are not simply adding Singapore to a list of markets to monitor; they are treating it as a working partner capable of solving problems their own internal teams cannot move on fast enough. That is a different, and arguably more durable, kind of attention than a scouting trip or a pilot programme with no clear next step.

Why Japan is looking to Singapore

Japanese corporations rarely approach business matching service for passive technology scouting. As Sony Acceleration Platform has described in earlier conversations with e27, they typically arrive with clearly defined operational bottlenecks, seeking practical, market-ready solutions across digital transformation, automation, and niche technical capabilities that can be fast-tracked into commercial deployment.

Singapore’s appeal lies in the specifics: a stable and predictable regulatory environment, strong technological capability, and business conducted seamlessly in English. For Japanese corporates trying to manage risk in cross-border ventures, that combination offers a stable, efficient starting point. Singapore’s tech community, meanwhile, brings a level of global compliance and agility that makes it a natural counterpart for enterprises seeking fast, reliable co-creation. Boundary Spanning Service exists to structure that meeting point.

Rather than functioning as a one-off transactional exercise, it is designed around mutual trust and reciprocal value creation, with the Japanese corporate contributing operational resources, industry expertise, and market access, and the Singapore startup contributing agility, speed, and disruptive technology. The goal is to lower the practical cost of survival for good ideas, and to get them into real-world deployment before the window closes.

What this looks like for a Singapore founder

For founders, engagement begins with a streamlined application process supported by e27, designed to be low-friction rather than bureaucratic. From there, Sony Acceleration Platform shares and recommends qualifying startup profiles to the Japanese corporations it works with, meaning founders can expect a structured introduction rather than a cold pitch. What follows is worth entering with clear eyes. Japanese corporate decision-making tends to involve extensive internal consensus-building, a structural characteristic of how these organisations operate, not a signal of disinterest or a personal hurdle. That front-loaded alignment process takes time, patience, and a willingness to meet exacting standards around quality and operational stability.

The payoff for founders who can meet that bar is a relationship that, once trust is established, tends to be exceptionally stable and deeply committed over the long term, backed by the distribution networks and credibility of a major corporate partner. Founders considering Boundary Spanning Service should honestly assess whether they are ready for this type of long-game partnership, rather than a fast transactional deal.

In practice, that means being candid with yourself about a few things: whether your product is genuinely ready for enterprise-grade scrutiny, whether your team has the bandwidth to sustain a longer sales and alignment cycle, and whether you are prepared to treat quality and operational stability as a foundation rather than a formality. Founders who approach Boundary Spanning Service with that mindset tend to be the ones who get the most out of it.

Why Singapore, and why now

Singapore’s position as the starting point for this corridor is not incidental. It reflects a deliberate choice by Sony Acceleration Platform to begin Japan-Singapore collaboration where the ecosystem conditions- technological, regulatory, and cultural- are most conducive to building the kind of trust these partnerships require.

For Singapore founders, it is a rare opportunity to meet Japanese corporate organisations directly, a meaningful step in a process that, by Sony Acceleration Platform’s own account, rewards founders who invest early in the relationship rather than those looking for a quick win.

Founders who engage with Boundary Spanning Service now are positioning themselves at the front of a collaboration corridor that both governments and leading corporations have signalled a long-term commitment to developing. As that corridor matures, early movers are likely to be the best placed to benefit from it.

Register your interest

Singapore founders curious about what this collaboration could mean for their own business have immediate ways to engage: submit an application through the Boundary Spanning Service at https://tally.so/r/815k1k?source=article . The region’s startup ecosystem is evolving quickly, and this corridor between Japan and Singapore offers a concrete, structured way for founders to be part of what comes next.

The region is evolving quickly, and e27 and this collaboration offer the right place at the right moment to be part of what comes next. Register here to join the conversation.

—

Want updates like this delivered directly?

Join our WhatsApp channel and stay in the loop.

The e27 team produced this article.

We can share your story at e27 too! Engage the Southeast Asian tech ecosystem by bringing your story to the world.

You can reach out to us here to get started.

The post Why Sony Acceleration Platform is turning to Singapore for innovation, and what it means for your startup appeared first on e27.

For much of the past two years, crypto investors have been forced to separate signal from noise. Token prices recovered, regulators tightened their grip, and the industry’s loudest claims about remaking finance gave way to a quieter question: which parts of blockchain are actually useful?

Singapore-based digital asset firm Psalion is betting that the answer lies less in speculative trading and more in the plumbing beneath everyday business. The firm has launched its third and largest venture vehicle, a US$50 million fund targeting early-stage startups using blockchain infrastructure in real-world markets.

Also Read: Malaysia, Singapore investors rebalance portfolios without ditching crypto

The fund, Psalion VC Fund III, is structured as a Singapore-domiciled Variable Capital Company, or VCC, a flexible investment fund structure widely used by asset managers in the city-state. It is managed by Conduit Asset Management, which is licensed and regulated by the Monetary Authority of Singapore (MAS).

The choice of Singapore is not incidental. Over the past few years, the country has tried to draw a firm line between crypto speculation and regulated digital asset innovation. The MAS has tightened rules around retail crypto promotion while continuing to support institutional experiments in tokenisation, stablecoins and digital money through initiatives such as Project Guardian. For fund managers like Psalion, that regulatory posture offers both credibility and constraints.

“Crypto was the thesis. The traditional financial system was the antithesis. What we’re investing in is the synthesis, where web2 businesses run on web3 rails,” said Tim Enneking, Managing Partner of Psalion.

That framing captures where much of the venture market around blockchain has moved. The first wave of crypto investing was largely about native protocols, tokens and decentralised finance platforms. The next wave, if it materialises, may be less visible to end users: stablecoins for payments, tokenised real-world assets, blockchain-based trade finance, middleware for digital ownership, and consumer products where the underlying rails are decentralised but the interface feels familiar.

Backing builders before the market turns

Psalion said the new fund will focus on pre-seed and seed-stage companies across infrastructure, middleware, trade finance, real-world assets, stablecoins and decentralised finance. It will also look at consumer applications where web3 technology changes how people own, trade or interact with assets without requiring them to behave like crypto-native users.

The timing is notable. Psalion’s previous two funds were also launched during weaker market cycles, and the firm appears to see the current environment as a feature rather than a drawback.

“Some of the best opportunities arise in down markets. Valuations are more reasonable, founders are more focused, and the builders who show up are in it for the long term,” Enneking said.

That view is common among venture investors, but it carries particular weight in crypto, where bull markets often inflate valuations before products have found real users. The collapse of several high-profile crypto businesses in 2022 and the subsequent regulatory clean-up changed the funding landscape. Startups now face harder questions around revenue, compliance and utility.

For Southeast Asia, those questions are especially relevant. The region has a large underbanked population, strong cross-border trade flows, high mobile adoption and fragmented financial infrastructure. These conditions have long made it attractive for fintech founders. Blockchain companies now need to prove they can solve similar problems without adding unnecessary complexity.

Also Read: How to use blockchain to fund and create a greener future

Stablecoins, for example, have gained attention as a tool for faster and cheaper cross-border settlement. Tokenised real-world assets have attracted banks and asset managers seeking more efficient ways to issue, trade and settle financial products. Trade finance remains a persistent pain point for small businesses across the region, where paperwork, trust gaps and slow settlement can limit access to working capital.

The opportunity is clear. The challenge is distribution, regulation and trust.

Beezie becomes an early test case

Alongside the fund launch, Psalion disclosed that it led a US$4 million round in Beezie, a commerce platform that blends digital ownership, discovery and liquidity into consumer transactions.

Beezie says it has processed more than US$170 million in gross merchandise value since launch and generated more than US$85 million in revenue year-to-date. It also claims to have attracted over 30,000 active users since January 2026. The company’s recent community raise on Echo, initially targeting US$250,000, reportedly sold out in 10 minutes and was capped at US$1 million within 24 hours.

The platform operates across collectibles, luxury and entertainment markets — categories where scarcity, authenticity and resale value matter. These are also areas where blockchain has often been pitched as useful, though consumer adoption has been uneven. Many users care about whether an item is authentic, tradable and liquid; fewer care about whether the backend uses blockchain.

That is precisely the point Psalion is making with the investment.

“Beezie is exactly the kind of company our thesis is built around — a real consumer product, real revenue, and blockchain quietly doing the heavy lifting underneath,” Enneking said. “This is what web2 meeting web3 actually looks like in practice.”

Beezie plans to use the capital for inventory acquisition, geographic expansion and growth across collectibles, luxury and entertainment. Founder and CEO Andrea Miele said the company wants to make transactions feel more engaging by combining ownership, anticipation, discovery and liquidity.

The broader question is whether that emotional layer can translate into durable commerce behaviour. Collectibles and luxury resale markets are already competitive, and user trust is hard won. Beezie will need to show that blockchain improves the experience rather than becoming a technical wrapper around familiar marketplace mechanics.

A crowded field for digital ownership

Beezie is entering a market with several established and adjacent rivals. Globally, platforms such as StockX, Whatnot, eBay, Sotheby’s and Christie’s already serve different parts of the collectibles, luxury and resale economy. In Southeast Asia, Singapore-founded Carousell and sneaker marketplace Novelship have built regional consumer bases around resale and authenticated goods. On the crypto-native side, platforms experimenting with tokenised collectibles and digital ownership have struggled to move beyond early adopters.

Psalion, meanwhile, is competing for deals with a global group of crypto and web3 investors, including Animoca Brands, Spartan Group, Hashed, Dragonfly, Pantera Capital and a16z crypto. The difference will not simply be cheque size, but whether the firm can help portfolio companies navigate compliance-heavy markets and reach users outside the crypto bubble.

Singapore’s role as a base for such funds may strengthen if institutional blockchain adoption continues to move from pilot projects to production systems. The city-state has already positioned itself as a hub for digital asset regulation, tokenisation trials and wealth management. But the region’s startup markets remain uneven. Indonesia, Vietnam, the Philippines, Thailand and Malaysia each present different regulatory and consumer realities.

Also Read: Singapore crypto adoption hits new high as 61 per cent now hold digital assets

For founders, that means a product that works in one market may need significant adaptation elsewhere. Payments, identity, gaming, commerce and financial services all carry local rules and behaviours. A regional web3 startup cannot rely only on the promise of decentralisation; it must solve very specific market problems.

Psalion’s new fund lands at a moment when blockchain has become less fashionable but potentially more useful. The speculative cycle has not disappeared, and neither have the risks. But investors are increasingly looking for companies that can hide technical complexity behind products people actually use.

If Psalion’s thesis proves right, the next major blockchain companies in Southeast Asia may not look like crypto companies at all. They may look like commerce platforms, trade tools, payment networks or financial infrastructure providers, with web3 working quietly in the background.

The post Singapore’s Psalion raises US$50M fund for Web3’s next practical phase appeared first on e27.

Across Southeast Asia, AI is reshaping how workers operate, but not in a uniform way. A new analysis reveals a striking cultural divergence: Filipino professionals are turning to AI primarily as a burnout shield, using it to offload repetitive, draining tasks and protect their mental bandwidth. Singaporean workers, by contrast, are deploying AI as a deep work enabler, a tool to carve out uninterrupted focus time and accelerate high-value cognitive output.

The difference is not merely stylistic. It reflects each country’s distinct labour market pressures and workplace cultures. The Philippines, home to one of the world’s largest business process outsourcing industries, faces chronic worker fatigue at scale. AI is emerging as a structural response to that exhaustion. Singapore, with its knowledge-economy positioning and productivity-first ethos, is using the same tools to raise the ceiling on individual performance.

For founders and operators across the region, the implications are significant. A one-size-fits-all AI adoption strategy will underperform. Localising how AI tools are introduced, framed, and incentivised — by country, by sector, by workforce profile — may determine whether AI delivers measurable gains or simply adds another layer of complexity to an already stretched team.

—–

REGIONAL

Psalion raises US$50M fund for Web3’s next practical phase: Singapore-based Psalion has closed a US$50M fund targeting what it calls Web3’s practical phase, focusing on real-world utility over speculation, one of the larger crypto-native fund closes in SEA this year.

Sunwah Innovations and Brinc launch Vietnam JV for university spinouts: The joint venture will channel university research in Vietnam into commercialised startups, targeting a gap between academic output and venture-ready companies across the country.

Singapore’s Tikva targets solid cancer barrier with US$8M Series A: Tikva is using its Series A raise to advance therapies that address tumour microenvironments, a persistent obstacle in treating solid cancers, positioning Singapore as a node in precision oncology.

Why investors back Vietnamese startups more aggressively than Thai peers: Structural factors, including Vietnam’s younger demographic, higher risk appetite, and stronger technical talent density, help explain why Vietnam consistently attracts more aggressive VC bets than Thailand.

Malaysia and Singapore investors rebalance portfolios without ditching crypto: Institutional and retail investors across both markets are adjusting allocations amid macro uncertainty, trimming risk exposure while retaining crypto positions — suggesting a maturing, not retreating, digital asset stance.

The 27 SEA biotech firms betting on cells, fermentation, and code: A deep-dive list of 27 Southeast Asian biotech companies working across synthetic biology, cultured meat, and computational biology signals the sector’s growing depth beyond Singapore.

Triple-A says own digital assets hit by unauthorised access: Singapore-based crypto payments firm Triple-A confirmed a security breach affecting its own holdings — a significant incident for a firm that processes crypto transactions for major enterprise clients.

SEA’s regulatory patchwork demands a local-first approach: Fragmented licensing regimes across Southeast Asia’s fintech and tech sectors are forcing companies to build compliance infrastructure country by country rather than regionally — raising costs and slowing expansion.

Vietnam’s tech talent market is broken and hiring habits won’t fix it: Most companies are sourcing Vietnamese tech talent through outdated channels that prioritise credentials over capability, exacerbating a structural mismatch between employer needs and available skills.

---

INTERVIEWS & FEATURES

What BEYOND Expo 2026 revealed about Asia’s hardware edge: Observations from BEYOND Expo 2026 point to a regional shift; Asian hardware founders are moving faster from prototype to production than Western counterparts, with supply chain proximity as a structural advantage.

How do you finance a first nuclear reactor for a data centre: A detailed breakdown of emerging deal structures for nuclear-powered data centres shows how project finance, offtake agreements, and government backing are converging to make first-of-kind builds viable.

---

INTERNATIONAL

Anthropic’s Dario Amodei flags risk from Chinese AI, not open-weight models: Amodei clarified he does not oppose open-weight AI development but warned that Chinese AI advancement poses a more material geopolitical risk — a framing with direct implications for SEA’s AI policy debates.

Satya Nadella warns against over-reliance on a single AI system: Microsoft’s CEO cautioned that companies trusting one AI for everything risk structural fragility, a message directed at enterprise AI buyers accelerating consolidation of their vendor stacks.

Lyft and Baidu begin robotaxi testing in London: The partnership marks Baidu’s first Western robotaxi deployment, testing Apollo Go technology in a heavily regulated market, a move that signals Baidu’s intent to internationalise its autonomous driving stack beyond China.

Waymo reportedly weighing a break with Uber: A reported split between Waymo and Uber would reshape the autonomous vehicle partnership landscape, with implications for how robotaxi services are distributed globally, including in markets where Uber dominates ride-hailing.

Microchip Technology acquires Israeli AI chipmaker Hailo: The acquisition gives Microchip a dedicated edge AI inference chip portfolio, strengthening its position in embedded AI hardware used in automotive, industrial, and smart device markets across Asia.

DeepSeek pauses fundraising amid viral post scrutiny: DeepSeek’s decision to halt funding discussions follows a wave of social media attention that drew regulatory and investor scrutiny, a rare instance of viral exposure creating friction rather than momentum for a high-profile AI lab.

Apple sued over US$1.8M App Store crypto scam: A lawsuit alleges Apple failed to remove a fraudulent crypto app despite user complaints, raising platform liability questions relevant to SEA’s growing base of mobile-first crypto users.

Swiggy names Nandita Sinha to head Instamart: The appointment puts an experienced consumer leader in charge of Swiggy’s quick-commerce arm as it battles Blinkit and Zepto in India’s intensifying 10-minute delivery war.

---

CYBERSECURITY

Microsoft launches its first cyber model and agentic security system: The new agentic cybersecurity system uses AI agents to detect and respond to threats autonomously, a significant product move that could reset enterprise security procurement benchmarks across Asia.

Hugging Face CEO calls for transparency after OpenAI hack: Following what he described as an unprecedented breach at OpenAI, Hugging Face’s CEO argues the AI industry must adopt radical transparency on security practices, or risk systemic trust collapse.

Singapore’s agentic AI ambitions hinge on code trustworthiness: Deploying agentic AI systems at scale in Singapore requires a level of software auditability and verification that current frameworks do not yet adequately address, according to a new analysis.

---

SEMICONDUCTOR

Quantum sovereignty in Asia: computing, AI, and emerging ventures: A detailed look at Asia’s quantum computing landscape maps how governments and startups across the region are racing to establish independent quantum capabilities as a strategic national asset.

Are brain waves the next unlock for physical AI?: Researchers exploring brain-computer interfaces as an AI input layer argue that neural signals could give physical AI systems a more precise and low-latency human control mechanism than current interfaces allow.

---

AI

AI voice is scaling across APAC, but listening hasn’t kept pace: Voice AI deployment across Asia-Pacific is outrunning the comprehension capabilities of underlying models, particularly in multilingual and dialect-heavy markets — creating accuracy gaps that affect trust and adoption.

Indonesia’s AI hiring gap is real, just not 28x: Claims of a 28-fold AI talent shortage in Indonesiaare overstated, but the underlying deficit is genuine — driven by a mismatch between university output and the practical skills employers need to deploy AI systems.

AI is making SEA’s startups faster not richer, yet: AI tools are compressing execution timelinesfor SEA startups, cutting weeks off product cycles and reducing headcount needs, but revenue impact remains limited as monetisation models lag behind adoption.

Enterprise AI adoption in SEA accelerates despite internal friction: Southeast Asian enterprises are pushing ahead with digital transformation despite unresolved questions around data governance, integration costs, and workforce readiness, a paradox driven by competitive pressure rather than strategic clarity.

Asia’s AI race will be won by whoever keeps the lights on: Energy infrastructure, not capital or talent, is emerging as the decisive constraint in Asia’s AI buildout, with power availability and grid reliability determining which markets can scale compute at speed.

How AI and blockchain could make commerce decisions more accountable: Combining AI decision-making with blockchain audit trails could create verifiable, tamper-resistant records of automated commercial choices, an approach gaining traction among compliance-focused fintechs in SEA.

AI repriced SEA’s marketing agencies; it didn’t replace them: AI tools have fundamentally altered agency pricing models across Southeast Asia — compressing margins on execution work while shifting value toward strategy, creativity, and client relationships.

Philippines use AI to avoid burnout; Singapore for deep work: A comparative study of AI adoption behaviours across the two markets reveals that workforce culture, industry structure, and labour market pressures shape how and why employees actually use AI tools at work.

---

THOUGHT LEADERSHIP

How the next-generation neobank should be built for the agentic economy: Agentic AI will fundamentally change banking UX, shifting from app-based interaction to autonomous financial agents that act on behalf of users, requiring neobanks to redesign core infrastructure rather than just the interface.

Bitcoin lost US$65,500 support three times before the Fed spoke: BTC’s repeated failure to hold US$65,500 signals weakening buyer conviction at a key technical level — and the pattern emerged before any Federal Reserve guidance, amplifying concern about near-term price direction.

Volatility is here to stay; here is what founders should do: Persistent market volatility is no longer a cyclical event but a structural condition, and founders who treat it as temporary will be outmanoeuvred by those who build resilience into their operating models from day one.

Is ether quietly stealing Bitcoin’s throne as institutional favourite: On-chain data and institutional flow metrics suggest Ethereum is gaining ground over Bitcoin as the preferred vehicle for sophisticated investors, driven by yield potential, programmability, and growing ETF momentum.

SEA’s fintech apps don’t have a literacy problem — they have a fear problem: Low fintech adoption in underserved SEA markets is less about users not understanding the products and more about a deep-seated distrust of digital financial systems rooted in past fraud and institutional failure.

The 3Cs+1 framework for founders navigating geopolitical fragmentation: A practical strategic framework helps founders map their exposure to geopolitical risk across customers, capital, code, and compliance — and decide where to localise, hedge, or exit.

The real workforce challenge: bridging the credential-capability gap: Credential inflation is masking a widening gap between what workers are certified to do and what employers actually need, a structural problem that AI adoption is making more urgent, not less.

How Singapore SMEs should choose a payment provider: A practical checklist for SME payment provider selection covers fee structures, settlement timelines, FX handling, and integration depth, useful for founders navigating Singapore’s crowded but inconsistent payments landscape.

The future of clinical trials: Decentralised clinical trial models using remote monitoring, real-world data, and AI-driven patient matching are accelerating drug development timelines and expanding access beyond traditional trial sites in Asia.

The evolution of trust: from early adoption to institutional maturity: Digital asset markets are moving through a structural trust transition, from speculative retail participation to regulated institutional frameworks, with SEA markets at varying stages of that journey.

The post How the Philippines and Singapore use AI differently at work appeared first on e27.

A few months ago, I was invited to a private university to deliver a keynote on C-Suite skills for 2026 and beyond. The audience was a room full of MBA students. Smart, ambitious, mid-career. People who had paid real money to think seriously about their next decade.

After the talk, one of them raised his hand and asked the question I knew was coming. The question he probably could not ask his own boss. The question that no senior leader at his company wanted to hear out loud.

“Will AI replace senior leaders too?”

The room went quiet. Not because the question was clever. Because it was honest.

Every executive in the world is privately asking themselves the same thing. Almost none of them say it in public. Saying it would look weak. It would unsettle their teams. It would invite the same question right back at them.

But the student had no such constraint. He had paid to learn. He wanted the truth.

Here is what I told him.

I have been studying AI since 2017. Even back then I was telling people that this technology was going to replace a lot of jobs. Including the roles people assumed were untouchable. Doctors. Lawyers. Consultants. Senior managers. Anyone whose work was mainly about applying expertise was going to feel the wave.

But the question most people ask is the wrong one.

AI is a tool. And a tool only performs at its best when someone behind it takes responsibility for the outcome.

So the real question is not “will AI replace you.”

The real question is “are you the one taking responsibility for what AI produces?”

Also Read: The role of thought leadership in scaling beyond your first market

That single shift changes everything. The moment you stop seeing yourself as the person executing a task that AI can do, and start seeing yourself as the person who owns the outcome AI helps produce, you stop being a candidate for replacement. You become the operator of the tool.

This is the part most senior leaders refuse to confront. They built their careers on being the best executor in the room. The most experienced. The most certified. The most credentialed. AI is taking that ground away. And instead of moving up to the layer AI cannot reach, many of them are doubling down on the layer it is colonising.

The student in front of me understood this immediately. He asked the follow up question that most senior leaders never get around to.

“How do I become that person?”

I told him three things.

First, take ownership of outcomes, not tasks. A task is what you do. An outcome is what your work produces in the world. AI can do tasks. AI cannot want an outcome. The person who owns the outcome cannot be automated out of the loop because the outcome is the reason the loop exists.

Second, build judgement in places AI cannot reach. AI is excellent at pattern recognition. It is much weaker at situations where the patterns are new. Reading a customer who has never bought anything like your product before. Navigating a partnership that has never been structured this way. Making a call when the data is incomplete and the stakes are high. These are the moments where human judgement compounds. Spend your career deliberately collecting these moments.

Third, become responsible for results that involve other humans. AI is improving rapidly at producing content. It is not improving at being trusted by a CFO who needs to make a hard decision. Or being chosen by a customer who has options. Or being followed by a team through a difficult quarter. Trust, choice, and followership are the human-only layer. Senior leaders who operate in that layer remain irreplaceable.

This is what I have been doing personally since 2019. I mentor startups. I invest in founders. I help them accelerate beyond their home country. I connect them with local partners in new markets. I help them generate revenue and lift their valuations toward exits.

Can AI do parts of this work? A lot of it, actually. AI can draft introductions, analyse markets, summarise a business plan, generate first-draft strategy.

But who takes responsibility when the call goes wrong?

Who reads the market that does not show up in any dataset?

Who carries the relationships that took years to build?

That part still belongs to a human. And that human is the one who deserves to be in the senior seat.

Also Read: 7 leadership skills every manager needs in a monitored workplace

This is also why I built Future 500 to operate the way it does. Every founder we back gets direct access to people who have actually built businesses from zero. Not because AI cannot do the analysis. It can. But because the founder needs someone willing to take responsibility for the outcome alongside them. That layer of accountability is what AI cannot replicate. It is the layer where senior leadership lives.

When I evaluate a founder for investment, I am looking for exactly this signal. Does this person take responsibility for outcomes, or do they deflect to circumstances? Do they own the numbers, good or bad? Do they propose the next move before being asked?

The founders who answer those questions clearly are the ones I back. The ones who cannot, get a polite pass. AI did not create this filter. It just made the filter sharper, because the candidates who pass it are now even more rare.

If you are a senior leader reading this and feeling shaky about your future, the question to ask yourself is not “will AI replace me?”

The question is “am I the one taking responsibility for what my work produces?”

If you are, AI will multiply you. It will sharpen your judgement, accelerate your output, and free you to focus on the layer of work that only humans can do.

If you are not, you were already replaceable. AI just sped up the timeline.

The MBA student walked out of that hall with a clearer answer than most senior leaders will ever give themselves.

The question is whether you are willing to ask it now, while you still have time to act on the answer.

—

Editor’s note: e27 aims to foster thought leadership by publishing views from the community. You can also share your perspective by submitting an article, video, podcast, or infographic.

The views expressed in this article are those of the author and do not necessarily reflect the official policy or position of e27.

Join us on WhatsApp, Instagram, Facebook, X, and LinkedIn to stay connected.

The post The future of leadership: Owning outcomes in the age of AI appeared first on e27.

Hyperscalers are preparing to spend roughly US$700 billion in 2026 to make artificial intelligence more capable and cheaper to run. That sounds like a straightforward growth story. More models, more compute, more demand, more data centres.

The risk is that the spending is buying the very forces that weaken part of the demand case.

AI build-outs are usually debated through the depreciation question: will the chips become obsolete too fast to earn back their cost? That is a real issue, but it is not the only one. A more important pressure may come from below. Once a model is good enough to complete a defined task, the buyer no longer needs the newest frontier model for that job. And once the cost of running that level of performance keeps falling, more of the work can move onto hardware the buyer already owns.

The problem is not that AI demand disappears. It probably does not. Cheaper intelligence creates more use. The problem is price. The share of work that can be done by a good-enough model on local hardware sets a ceiling on what a data centre can charge for that work.

That ceiling is falling.

The four largest US technology platforms have guided toward extraordinary capital spending. Amazon has pointed to about US$200 billion, Microsoft to about US$190 billion for the calendar year, Alphabet to US$175 billion to US$185 billion, and Meta to US$115 billion to US$135 billion. McKinsey’s widely cited estimate puts worldwide data-centre capital needs near US$6.7 trillion by 2030, with about US$5.2 trillion AI-specific. Morgan Stanley’s estimate is lower, near US$3 trillion through 2029, but still leaves a financing gap it puts around US$1.5 trillion.

These sums create fixed obligations: debt, power contracts, long-dated capacity deals, and depreciation schedules. Those obligations must be serviced regardless of what price inference commands.

The spending buys two things at once. It buys capability, so models can do more. It also buys efficiency, so a fixed level of AI performance becomes cheaper. Both are desirable. Both also undercut scarcity pricing.

Also Read: How AI and blockchain could make commerce decisions more accountable

Capability has a bound for each task. A tax return, a legal draft, a customer-support exchange, or a code review requires a model good enough to finish the job to a competent standard. Once that threshold is crossed, a better model does not make the completed task more complete. The buyer then moves from “best available model” to “cheapest model that clears the bar.”

Efficiency then moves the same work toward commodity pricing. Epoch AI has found the price of reaching a fixed performance milestone falling between ninefold and several-hundredfold a year in some cases. Andreessen Horowitz has tracked inference at constant capability falling at roughly tenfold a year. Gartner expects inference on a trillion-parameter model to cost more than 90 per cent less in 2030 than in 2025, with on-device and edge inference among the drivers.

That matters because the edge is no longer theoretical. Gartner puts AI PCs near 31 per cent of the market in 2025. Counterpoint puts penetration closer to two-fifths. Canalys expects more than 200 million AI PCs shipped annually by 2028, and IDC expects neural processing units to be near-universal in new PCs by then.

Capability is moving with the hardware. A 120-billion-parameter open model now runs on a single desktop appliance at roughly 32 tokens per second. A 70-billion-parameter model runs on a mini-PC costing about US$1,500 at 12 to 15 tokens per second. Apple’s unified-memory Macs and Nvidia and AMD desktop systems put comparable capability on millions of desks. The best open-weight models now trail the leading closed systems by low single-digit points on neutral capability indexes, with several available under permissive licenses.

This does not mean the edge replaces the frontier. It does not. Frontier training remains capital-bound. The largest facilities have minimum efficient scale that owned devices cannot touch. Workloads needing very long context, low latency at high concurrency, proprietary cloud data, always-on orchestration, or the newest reasoning models still belong in centralised infrastructure.

But that is exactly the distinction. The defensible data-centre market is the work that structurally needs the data centre. The fragile slice is ordinary inference that once lived in the cloud only because a capable model could not run anywhere else.

Also Read: The barrier to AI adoption was never budget, it was knowing where to start

For that slice, the buyer has a standing choice: rent compute from a data centre, or run the workload on owned hardware. The amortised cost of the owned option becomes the maximum sustainable cloud price before work defects. A chip can be fully utilised and still earn a thinning margin if the price it commands is set by what the same job costs on a machine the customer already bought.

This is why the depreciation debate misses part of the issue. Obsolescence asks whether a chip’s useful life is shorter than the accounting schedule. Edge substitution asks whether the work the chip serves is still scarce enough to command the assumed price. Those are different risks.

The price signals are already mixed, which is what should make the issue interesting rather than dismissible. Hourly rates for prior-generation accelerators fell sharply from around US$8 in early 2024 into a US$1.50 to US$3.50 band by late 2025, as newer chips arrived and new providers entered the market. That looked like commodity pressure. Then rates reversed: one-year rental contracts rose about 40 per cent from an October 2025 low of US$1.70 to US$2.35 by March 2026, with on-demand capacity sold out. The rebound shows that demand is still strong enough to absorb capacity. It does not prove the ceiling has vanished.

McKinsey’s own estimate captures the uncertainty. It expects roughly 60 per cent to 65 per cent of US and European AI workloads to sit on hyperscaler infrastructure by 2030. That still leaves a third or more elsewhere, and the cloud-versus-edge split is a live swing factor.

The strongest objection is simple: total demand may outrun the whole problem. Agentic workflows can use five to thirty times more tokens than a chatbot exchange. Older chips can flow down into inference rather than strand. One Bernstein estimate says a five-year-old accelerator can still earn about US$0.93 an hour against US$0.28 of cash cost, implying a contribution margin above 70 per cent. If AI-generated productivity gains keep arriving, demand expansion may swamp pricing pressure.

That objection is serious. The likely outcome is not collapse. It is segmentation. Frontier work remains centralised and expensive. Commodity work gets cheaper. Some of that commodity work stays in data centres because management, integration, security, and convenience matter. Some moves to owned hardware because the economics become too obvious to ignore.

The watch points are clear. If open-weight models keep closing the gap, the local option strengthens. If memory shortages keep edge hardware expensive, the ceiling falls more slowly. If frontier capability re-widens, centralised infrastructure keeps more pricing power. If commodity accelerator rental rates soften while AI PC penetration rises, the pressure is binding.

The AI build-out may still work. But the risk is not just that racks sit idle. The sharper risk is that the racks are busy carrying work whose price is being set somewhere else.

—

Editor’s note: e27 aims to foster thought leadership by publishing views from the community. You can also share your perspective by submitting an article, video, podcast, or infographic.

The views expressed in this article are those of the author and do not necessarily reflect the official policy or position of e27.

Join us on WhatsApp, Instagram, Facebook, X, and LinkedIn to stay connected.

The post The US$5 trillion AI data-centre buildout unleashes the paradox that limits its returns appeared first on e27.