Ah, yes, the unofficial creed of Deep Tech: If 20 engineers can finish it in a year, it’s just Tuesday in SaaS.
Deep Tech lives on the edge of the possible — usually the kind of edge where “five years,” “uncertain feasibility,” and “requires its own new physics” are standard phrases in project plans.
That’s why it can feel almost absurd to hear phrases like “move fast” or “fail early” in our world. After all, if building it were fast or simple, it wouldn’t be Deep Tech, right?
But here’s the thing: speed isn’t the enemy of depth — it’s a tool to get there faster. Companies like SpaceX, NVIDIA, and Rigetti are proving that you can tackle complex, world-changing problems without falling into the trap of over-planning and endless refinement.
Let’s talk about how.
Deep Tech lives on the edge of the possible — usually the kind of edge where “five years,” “uncertain feasibility,” and “requires its own new physics” are standard phrases in project plans.
That’s why it can feel almost absurd to hear phrases like “move fast” or “fail early” in our world. After all, if building it were fast or simple, it wouldn’t be Deep Tech, right?
But here’s the thing: speed isn’t the enemy of depth — it’s a tool to get there faster. Companies like SpaceX, NVIDIA, and Rigetti are proving that you can tackle complex, world-changing problems without falling into the trap of over-planning and endless refinement.
Let’s talk about how.
The Long Road Is Inevitable, but the Potholes Are Optional
In Deep Tech, the scope alone can make progress feel paralyzing. You’re not tweaking interfaces; you’re rewriting the laws of computation, energy, and biology. But complexity doesn’t mean you’re doomed to crawl. It just means you need to think differently.
Take SpaceX: they’re literally building reusable rockets. That’s not just hard — it’s unprecedented. But they’ve chosen speed over perfection because every failed test gets them closer to what works.
The SN8 prototype didn’t need a flawless landing; it needed to explode in just the right way to tell engineers, “Fix the valve.”
If they’d waited for a perfect first try, they’d still be in the design phase. Instead, they’re out here launching, crashing, and learning faster than anyone thought possible.
Steps to Apply:
• Build prototypes to uncover weaknesses.
• Analyze failures and resolve issues immediately.
• Conduct rapid improvement cycles.
Framework: Test → Analyze → Refine → Test Again
Take SpaceX: they’re literally building reusable rockets. That’s not just hard — it’s unprecedented. But they’ve chosen speed over perfection because every failed test gets them closer to what works.
The SN8 prototype didn’t need a flawless landing; it needed to explode in just the right way to tell engineers, “Fix the valve.”
If they’d waited for a perfect first try, they’d still be in the design phase. Instead, they’re out here launching, crashing, and learning faster than anyone thought possible.
Steps to Apply:
• Build prototypes to uncover weaknesses.
• Analyze failures and resolve issues immediately.
• Conduct rapid improvement cycles.
Framework: Test → Analyze → Refine → Test Again
Deep Tech Loves Its Drama — But That’s a Problem
There’s this unspoken belief in Deep Tech that big problems need to be tackled with big, sweeping solutions. But real progress? It’s almost always incremental.
Rigetti’s work in quantum computing is a perfect example. They didn’t aim for a fully functional, breakthrough quantum machine from day one. Instead, they released a cloud platform that let researchers test algorithms on their processors. Those experiments gave Rigetti the data they needed to refine hardware, one step at a time.
Not glamorous. Definitely not headline-worthy. But effective. Because when the mountain is too big to climb in one shot, you start with the first foothold.
Steps to Apply:
• Use minimal experiments to test ideas.
• Focus on high-risk challenges first.
• Base decisions on measurable outcomes.
Framework: Hypothesize → Test → Analyze → Improve
Rigetti’s work in quantum computing is a perfect example. They didn’t aim for a fully functional, breakthrough quantum machine from day one. Instead, they released a cloud platform that let researchers test algorithms on their processors. Those experiments gave Rigetti the data they needed to refine hardware, one step at a time.
Not glamorous. Definitely not headline-worthy. But effective. Because when the mountain is too big to climb in one shot, you start with the first foothold.
Steps to Apply:
• Use minimal experiments to test ideas.
• Focus on high-risk challenges first.
• Base decisions on measurable outcomes.
Framework: Hypothesize → Test → Analyze → Improve
Silos Are Where Good Ideas Go to Die
Deep Tech has another bad habit: working in silos. The biologists figure out one part, the engineers another, and the data scientists come in at the end wondering why no one told them about that massive bottleneck upstream. By the time it’s all pieced together, years have passed, and nothing works.
NVIDIA’s Omniverse flips this script entirely. It’s a shared platform where everyone — designers, developers, engineers — collaborates in real time.
Change one variable in a simulation, and everyone else sees the ripple effects instantly. It’s chaos, but it’s productive chaos.
Deep Tech doesn’t need more geniuses working alone. It needs more messy, vibrant collaboration.
NVIDIA’s Omniverse flips this script entirely. It’s a shared platform where everyone — designers, developers, engineers — collaborates in real time.
Change one variable in a simulation, and everyone else sees the ripple effects instantly. It’s chaos, but it’s productive chaos.
Deep Tech doesn’t need more geniuses working alone. It needs more messy, vibrant collaboration.
Steps to Apply:
• Release prototypes to gather user feedback.
• Integrate teams with diverse expertise.
• Use user insights to refine tools continuously.
Framework: Collaborate → Test → Refine → Iterate
• Integrate teams with diverse expertise.
• Use user insights to refine tools continuously.
Framework: Collaborate → Test → Refine → Iterate
The Punchline
The joke may be that if it’s simple enough for 20 engineers to finish in a year, it’s not Deep Tech. But the tragedy? That’s when the 20 engineers take 10 years because they’re too afraid to build, test, or fail.
Deep Tech can still move fast — it just requires us to be bolder about breaking things, quicker to involve the whole team, and less precious about our initial designs.
So, yes, you’re solving the hardest problems on Earth (or beyond), but that doesn’t mean you can’t borrow a page from the 20-engineer, one-year playbook when it comes to speed and agility.
The challenge isn’t in the complexity. It’s in how we respond to it.
Deep Tech can still move fast — it just requires us to be bolder about breaking things, quicker to involve the whole team, and less precious about our initial designs.
So, yes, you’re solving the hardest problems on Earth (or beyond), but that doesn’t mean you can’t borrow a page from the 20-engineer, one-year playbook when it comes to speed and agility.
The challenge isn’t in the complexity. It’s in how we respond to it.