“The interim phase to autonomous vehicles is harder than the end goal. The end goal is all cars are autonomous. The interim is humans and machines sharing the road.” - Kevin
Hey guys,
Something strange just happened in Singapore.
In September 2025, Grab announced plans to launch Ai.R, its first autonomous vehicle service for consumers in Singapore, in partnership with WeRide. Days later, Pony.ai announced its entry into the Singapore market, partnering with ComfortDelGro, the country’s largest transportation service provider, to deploy autonomous vehicles.
Two massive AV deployments. Same city. Same month. Different playbook than anything Silicon Valley tried.
Here’s what nobody’s connecting: This isn’t just about self-driving cars. It’s about how deep tech actually scales when you stop pretending regulation doesn’t exist.
P.S. Share this with one founder who needs to rethink their regulatory strategy. That’s how we grow.
The Partnership Puzzle
Think about this for a second. Grab partnered with WeRide. ComfortDelGro partnered with Pony.ai. Not acquisitions. Not licensing deals. Strategic partnerships where neither side could win alone.
Grab brings:
Millions of ride records
App infrastructure that works across 8 countries
15 years of regulatory relationships in Singapore
User trust built ride by ride
WeRide brings:
Autonomous driving technology
Over 50 million kilometres of autonomous driving experience
Two vehicle models that cleared Singapore’s safety assessment
Global AV credibility
Neither company walks into Singapore and “disrupts” anything. They both recognize that moving fast and breaking things doesn’t work when things include pedestrians.
What Actually Happened in the US
Let me get the Silicon Valley story straight, because it matters.
Waymo didn’t fail. They’ve been running commercial robotaxi service in San Francisco, Phoenix, Los Angeles, and Austin. Waymo leads the robotaxi market with over 4 million rides in 2024 and 96 million rider-only miles projected by mid-2025.
But here’s what’s true: They fought every inch of the way. Years of regulatory battles. Public skepticism. Constant scrutiny over every incident.
Cruise? Different story. They did get shut down in San Francisco after incidents that included one where their vehicle was involved in dragging a pedestrian. That’s not just a PR problem—that’s the entire autonomous vehicle future hanging by a thread.
The US approach: Build the tech, launch the service, negotiate with regulators as problems emerge, hope public opinion doesn’t turn against you.
The Singapore approach: In 2025 the Land Transport Authority announced partnerships with several AV developers for pilots covering daily commutes, with the aim of deploying a limited number of driverless buses and shuttles on public roads by 2025.
See the difference? The government didn’t approve pilots. They orchestrated them. They selected partners. They set the timeline.
The Question Every Deep Tech Founder Needs to Ask Right Now
Stop asking: “Can we build this technology?”
Start asking: “Who already owns the relationship we need?”
Not “might buy our product.” Not “could be a customer.”
Who has distribution, trust, and regulatory access we cannot replicate—and a strategic problem they cannot solve without us?
That’s your entry point. That’s your leverage. That’s how you survive the next 24 months in deep tech.
Because here’s the uncomfortable truth: If you’re selling to an established player, you’re a vendor. Vendors get squeezed on price and replaced when something cheaper comes along.
If you’re partnering with an established player to solve something existential for them, you’re a strategic partner. Strategic partners get equity, long-term commitments, and a seat at the table.
Pony.ai’s collaboration with ComfortDelGro began with a Memorandum of Understanding in July 2024, establishing a strategic partnership aimed at accelerating large-scale commercial deployment of autonomous vehicles. That’s not a vendor relationship. That’s co-building the future of transportation.
Now Let’s Talk About The Fundraising Desert
While autonomous vehicles get all the headlines, here’s the reality most founders are living:
Series A dollars deployed: Down 23% year-over-year Deal volume: Down 18% Median time from Seed to Series A: 20 months
But that median hides the brutal truth. A significant chunk of companies are taking 3.5 years or more to get from Seed to Series A.
If you raised an 18-month seed round expecting to raise Series A in 12-15 months? You’re probably scrambling for a bridge round right now.
This isn’t a temporary dip. This is the new normal. The easy money era of 2020-2021 ended abruptly, and it’s not coming back.
What’s Actually Getting Funded
Hot sectors:
AI infrastructure (chips, hardware, not the wrapper apps)
AI-native SaaS (built from scratch, not retrofitted)
Cybersecurity (always evergreen)
FinTech (surprisingly resilient)
Web3 (modest comeback in valuations, not volume)
Cold sectors:
Food & beverage
Personal products
Ad tech
Medical devices
Translation: If you’re in a hot sector, money still exists but standards are higher. If you’re in a cold sector, you need 2-3x better metrics to raise the same round, and it’ll take 50% longer.
The Growth Playbook Nobody Wants Anymore
Remember triple-triple-double-double-double? Hit $1M ARR, then grow 3x, 3x, 2x, 2x, 2x to reach ~$100M in five years.
That playbook is dead.
Now the bar is: 10-20x revenue growth in the first 18 months. $100M ARR before Series B. Path to profitability visible from seed stage.
If that sounds insane, congratulations—you’re paying attention. But that’s also what’s getting funded in 2025.
What Southeast Asia Founders Must Internalize
1. Fundraising takes 2x longer than you think
If you think it takes 6 months, plan for 12. Your competitors who don’t plan for this will run out of money mid-process.
2. Extended runway isn’t optional
Raise 24-30 months of runway, not 18. If you can’t, have a detailed burn reduction plan before you need it, not when you’re three months from zero.
3. Government isn’t your enemy—it’s your accelerant
Singapore proves something critical: In emerging markets, government backing is the fastest path to scale.
If you’re still trying to “disrupt” your way around regulators, you’re fighting yesterday’s war.
4. Strategic partnerships beat solo execution
The era of “we’ll build everything ourselves” is over. Find the established player who needs you as badly as you need them. That’s your Series A pitch.
5. Stop copying Silicon Valley playbooks
What works in San Francisco doesn’t work in Jakarta. What works in Jakarta doesn’t work in Manila.
Build for the market you’re actually in, not the market you wish you were in.
The AI Reality Check Nobody Wants to Hear
Everyone’s adding “AI-powered” to their pitch deck like it’s 2021 and we’re throwing “blockchain” at everything.
Here’s what investors actually ask:
“Is AI core to your value prop, or did you just wrap an LLM?”
“What’s your moat when the next model drops?”
“Why is this a company and not a feature?”
If you can’t answer these authentically, don’t force the AI angle. Investors can smell AI-washing from across the table.
Two Futures, One Winner
We’re watching two parallel approaches to innovation unfold:
Approach A: Move fast, break things, fight regulators, hope public opinion stays positive, burn billions proving a concept.
Approach B: Partner strategically, work with regulators, test in controlled environments, scale with infrastructure, build trust systematically.
Singapore just demonstrated Approach B at scale. Pony.ai will deploy autonomous vehicles in the Punggol area with fixed routes, pending regulatory approval. Controlled rollout. Fixed routes. Regulatory partnership.
Not as exciting as “we’re launching tomorrow and regulators can deal with it.” But far more likely to actually work.
What This Means for You
If you’re building deep tech in Southeast Asia right now, you’re navigating the hardest fundraising environment in a decade.
But here’s the upside: If you can build something real, something sustainable, something that solves an actual problem with proper unit economics and strategic partnerships—you’ll be one of the few left standing when this correction ends.
And corrections always end.
The founders who win won’t have the flashiest demos or the biggest seed rounds.
They’ll be the ones who figured out how to:
Partner with established players strategically
Navigate regulation as a competitive advantage
Build for 30-month runways in a capital-constrained market
Focus on fundamentals while everyone chases whatever’s trending on Twitter
One Last Thing
The autonomous vehicle story isn’t really about autonomous vehicles.
It’s about what happens when you stop pretending you can ignore the existing infrastructure and start figuring out how to become part of it.
WeRide joined Singapore’s Steering Committee on Autonomous Vehicles as the only international member alongside Waymo. That’s not disruption. That’s integration.
And integration, it turns out, scales faster than disruption when you’re building things that touch millions of people’s daily lives.
Worth thinking about.
What resonates? What did we miss? Reply with what you’re seeing in your market—these conversations shape how we cover the ecosystem.
Listen to the full episode:
🎧 Spotify | Apple Podcasts | YouTube
Sources:
Sea of Startups delivers weekly reality checks for founders building in Southeast Asia. No fluff. Just what actually matters.
Subscribe for insights on what’s really happening in Southeast Asian tech.
