Back in April, before anyone outside a small room could buy a share, I sat down with Chris Quilty to answer one question:

What is Starlink actually worth?

I kept that session private at the time. It was for our LPs only. With yesterday’s news it feels like the right moment to open it up.

A few things from that hour have stuck with me.

Chris and his team built their Starlink model from the ground up. They pulled every FCC gateway filing. They estimated the business component by component, headcount by function, state by state. When the filing went public, he told me their numbers had come in within a few percent.

And his opinion on the IPO surprised me. He called it almost a non-event. SpaceX never struggled to raise money, and going public will not change how they operate. The bigger story, the one worth your attention, is forming above the IPO. A nascent space economy is maturing. Debris removal. Orbital power. Manufacturing in orbit. Lunar resources. Refueling depots. Even data centers in space. Most of it early, some of it leaning on government as the first customer, all of it real enough to track.

We spent real time on the question everyone is asking:

Does a company this large actually have a moat?

Chris’s answer was that there is no single thing. It is the culture, and how they solve problems from first principles. The first satellites used solar panels from Home Depot and a vending machine cooling system. They prefab their ground gateways in a factory, pour the concrete pedestals, truck the whole thing out to a scraped patch of land, and light it up. Cheap where it can be, redundant where it matters, and it works.

We dove a lot deeper into SpaceX’s Culture as a Moat in our Ubiquity University module with former SpaceX engineer Brannon Jones (who is now a fellow deeptech VC). Read our summary or watch the full 17-minute video now:

For fifty years, military space ran on a short list of prime contractors. The joke in the room was that you couldn’t play unless you had a blue logo. That is finally changing, and the newcomers are the whole reason we find this sector so interesting.

The scale is hard to wrap your head around. Outside of the recent AI companies, Starlink reached ten billion in revenue faster than almost anyone. There was an eighty day stretch last year where they added a million subscribers. They now build around nine million user terminals a year and hand many of them out for free. It took fifty five years after Sputnik for the world to put its first thousand satellites in orbit. SpaceX is on track to launch roughly thirty five hundred this year alone.

I have backed early-stage space startups for almost 15 years, going back to leading an early round in Rocket Lab when it was valued under $50 million (today the company is worth over $50 billion). Space is software beyond the screen taken to its furthest edge. Smart hardware solving a high stakes problem in one of the most overlooked sectors there is. One foot in today and one foot in tomorrow. Watching a company go from a crazy idea to the largest offering in market history is a good reminder to stay nerdy and early.

Here are 2 parts of the Ubiquity Ventures event from April. Worth the watch if you want the analysis behind the number everyone is thinking about today.

Part 1: 9-minute Space 101 Primer with Sunil

In this brief video, we go into the basics of space: How we utilize space, the 3 major segments of the space economy, LEO vs. GEO orbits, launch vehicle dynamics, and more:

Part 2: Starlink Analysis and Resulting Space Opportunities with Chris & Sunil

In this video, Chris Quilty analyzes Starlink’s overall performance, the various Starlink customer segments (maritime, aviation, defense, mobile/direct-to-cell). Then Chris and Sunil explore the resulting opportunities and implications for the broader space economuy.

Huge thanks to Chris and the team at Quilty Space. If you want space research at this level, they are the source. You can find them at quiltyspace.com.

Ubiquity Ventures is a seed-stage venture capital firm investing in software beyond the screen. We back founders building AI, software, and smart hardware for the physical world — technology you can touch, hear, and feel.

If you know a founder working on something nerdy and early, we’d love an introduction.

Deep tech is having a moment.

What used to be a niche corner of venture now draws serious capital and attention, as investors hunt for the kind of durable moats that are getting harder to find anywhere else. When software is cheap to build and quick to copy, the advantage moves to the problems that stay hard.

Most of the investors pointed to the same shift. The action is moving from software on a screen to AI in the physical world, where the hard problems are energy, robotics, sensors, compute, and the infrastructure that lets any of it run. And almost everyone agreed that deep tech is harder for AI to flatten than SaaS, because you still need the robot, the grid, and the proprietary data.

A few pointed to where the next edge gets built, in regulated industries like energy, defense, and healthcare, and in the rush of generalist investors now crowding into the category.

Then the takes split: one investor thinks the biggest new risk to deep tech has nothing to do with technology at all.

Thanks to Lu Zhang, Adam Draper, Reilly Brennan, Julie Lein, and Rob Toews for sharing what they’re watching. Here’s where they landed.

Rob Toews, Partner, Radical Ventures

“In 2026, breathtakingly quickly, deep tech has gone from contrarian to consensus. The SaaSpocalypse narrative and the breadth of Anthropic and OpenAI’s competitive blast radius has pushed many VCs and entrepreneurs to focus on areas that feel less vulnerable to Claude’s near-term roadmap, including a lot of hardware. Space, semiconductors, robotics, bio, brain-computer interfaces, these have all seen a massive surge of interest and funding in recent months.”

For Rob, it’s a homecoming. “In a way, it is a return to venture capital’s most pure and true roots.”

Lu Zhang, Founder & Managing Partner, Fusion Fund

“We’re witnessing a shift from digital AI to physical AI. Today, the bottlenecks are increasingly in the physical world: energy, data centers, industrial systems, robotics, connectivity, sensors, and compute infrastructure.”

Her bigger claim is about who captures the value. “Many of the biggest winners of the AI era may not be AI model companies at all. They may be the companies solving the constraints around power, compute, data movement, and real-world deployment.”

On why deep tech holds up where SaaS strains, her logic is concrete. “A robotics company still needs robots. An energy company still needs grid infrastructure. An industrial AI platform still needs access to proprietary operational data and deep workflow integration. AI is becoming a force multiplier for deep tech.”

Adam Draper, Founder & Managing Director, Boost VC

“Bio is Deep Tech.” Adam’s takes come compact.

On what survives the AI wave: “Atoms-based businesses are more protected from AI’s software-killing nature, but all businesses will be transformed by AI. The adaptable survive in Software and Deep Tech.”

Julie Lein, Co-founder and Managing Partner, Urban Innovation Fund

“Everyone is a deep tech investor in 2026. As investors look for any startup with a shot at building a competitive moat, deep tech looks increasingly appealing to generalist VCs.”

Her second take points to where the next edge gets built. “2026 will be the year of startups working in regulated industries. As startups look for real competitive advantages, they’ll find an edge in building for highly regulated customers.” On AI, she doesn’t hedge. “I don’t see AI as a threat to deep tech. It’s unlocking transformative deep tech innovation. And that’s a good thing.”

Reilly Brennan, Partner, Trucks VC

“Automated driving is back to being the biggest opportunity in venture and 2026 feels like 2016 all over again. Even Travis is getting back in. We are seeing the ecosystem formation moment, much like we did with aviation in the 1930s.”

On AI, Reilly agrees deep tech is less threatened than SaaS, but points to a different risk. “Deep tech is now more threatened by entrepreneurial tourism than ever. If the costs of starting an automated driving company are down more than 10x by our measurements, you are going to have a lot more companies, which is good, and a lot lower switching costs, which can produce the TV remote control effect.”

Sunil Nagaraj, Founding Partner, Ubiquity Ventures

The most durable companies take software off the screen and into the physical world, where a model alone cannot win and the moat is the hardware, the proprietary data, and the domain expertise wrapped around it.

Nerdy and early, before the category is obvious.

Ubiquity Ventures is a seed-stage venture capital firm investing in software beyond the screen. We back founders building AI, software, and smart hardware for the physical world — technology you can touch, hear, and feel.

If you know a founder working on something nerdy and early, we’d love an introduction.

Reach Us

After hundreds of deep tech pitches, the ones that land all have one thing in common. The founder defines the evaluation criteria before presenting the results.

In the absence of a standardized framework, the smartest founders write their own rubric and walk investors through it before showing a single number.

Here’s what that looks like in practice.

Step 1: Define what success looks like in your space

Before you show any results, give investors the context to interpret them. What does the state of the art look like today? What milestone were you originally aiming for? What would a meaningful leap actually mean in your field?

This matters especially in physical AI and “software beyond the screen” categories, where the metrics are physical and unfamiliar: megapixels on an infrared sensor, cycle time on a robotic arm, latency on a hardware inference stack. A VC with a SaaS background has never thought about any of these. Your job is to make your progress legible before you present it.

For example: if you’re building a surgical robotics system, don’t open with your accuracy rate in isolation. Open with what the current standard of care looks like, what human error rates are in that procedure today, and what improvement would actually matter clinically. Now your number has a home.

Step 2: Get alignment before you present

Once you’ve framed the criteria, make sure your audience is with you before moving on. A simple gut check (”Does that baseline make sense to you?”) goes a long way. If investors don’t agree with your starting point, your results won’t land, no matter how strong they are.

For example: if you’re building in autonomous vehicles, anchor to the SAE autonomy levels before presenting your results. Once an investor understands that most commercial deployments are still at Level 2, your Level 4 milestone in a constrained environment carries real weight.

Step 3: Present your results inside that framework

Now show your numbers. With context established, milestones carry weight. Investors can see not just what you achieved, but how it compares to what was previously possible, and what it signals about your trajectory.

For example: instead of “we’ve completed 500 hours of field testing,” try “most systems at our stage have done 50 hours in controlled environments. We’ve done 500 in the field, across three different climate conditions.” Same number, completely different signal.

Use this in your pitch deck, in early investor conversations, and any time you’re asked to explain what you’ve built. The goal is the same every time: give your audience the criteria before you give them the results.

The infrared camera example

Founder, version one: “We built a 5-megapixel infrared camera.”

VC: blank stare.

Founder, version two: “The most advanced infrared camera available today is 1 megapixel. We raised our initial funding with the goal of hitting 3. We’ve already blown past that and reached 5.”

Same milestone. Completely different reaction. The second version works because the criteria come first. The investor can see not just a number, but a leapfrog over the state of the art, and a team that set a goal and exceeded it.

What deep tech investors wish more founders knew

Deep tech fundraising is fundamentally different from raising for a software company. You are often the only person in the room who understands what your numbers mean. Defining the scorecard before you show the score is how you close that gap.

The best deep tech founders treat this the same way they treat founder-market fit. Knowing the problem deeply is only half of it. The other half is being able to bring others inside that understanding. If you’re a VC who encounters deep tech deals outside your core thesis, this is the framework worth sharing with your founders before they pitch.

Ubiquity Ventures is a seed-stage venture capital firm investing in software beyond the screen. We back founders building AI, software, and smart hardware for the physical world — technology you can touch, hear, and feel.

If you know a founder working on something nerdy and early, we’d love an introduction.

Reach Us

How do you build real conviction… when there’s almost no data?

The full interview is linked at the bottom, but here are my top 5 takeaways :

1) Start with the product — not the market slide

“I’ve always liked to put my hands on tech, to immerse myself in technology… I realized that makes me a little bit of a quirk in the venture industry… I always start with product. I always start with the technology… not from the business potential, but from what’s actually being built.”

That instinct is what led me to Rocket Lab — with their unique rocket engine that looked like hardware, but was really software underneath. Today’s the startup is worth 1,000x more than when we initially invested while I was at Bessemer and is the foundation for our software beyond the screen thesis.

2) The best investments don’t take weeks — they click

“Without exception, every single one of my top investments has been a decision within about 24 hours… but when something resonates, it resonates pretty quickly… I stick to my knitting and when the signals line up, it just clicks.”

Speed here isn’t about rushing. It’s about having a prepared mind and recognizing patterns instantly when they show up.

3) Turn messy real-world problems into clean software abstractions

With Halter, I didn’t see agriculture — I saw code:

“My mind was blown… I was thinking, oh, so each of these cows is like an C++ object… with methods like “setters” and “getters”… there’s an SDK for a cow… you can write scripts so all the cows come to the milk shed at 6 a.m.… like a cron job in linux… accelerometer data can be referenced with if-then logic applied to 500 cows.”

That translation layer — from atoms to APIs — is where a lot of non-consensus insight lives.

4) “Moonshot culture” often removes the very discipline you need

“Moonshot culture says ignore all the rules… I’m not going to raise a $2 million seed or $5 million seed… I’m going to raise a $500 million seed because I’m doing a moonshot… I’m not going to have board meetings… you don’t understand, it’s a moonshot… and it gets more and more abrasive from there.”

That mindset sounds bold. But in practice, it often skips:

• technical diligence

• staged risk reduction

• real customer validation

The best deep tech companies still follow a disciplined, step-by-step path. I wrote about this more here:

5) Scarcity isn’t a constraint — it’s a forcing function

“There’s an argument that scarcity… drives a certain intensity, a certain creativity… it allows you to create focus… most entrepreneurs have 20 ideas they’d like to do, but because of the capital and runway, they focus on one or two… and if they did all 20, there’s a chance all 20 fail.”

When resources are abundant, it’s very hard to stay scrappy.

Focus drifts. Standards slip. Complexity creeps in.

Constraints, on the other hand, force clarity.

We also covered:

• why the new AI benchmarks like “$100M in year one” are often misleading (“stop the panic”)

• how to move at “founder speed” without cutting corners

• and where new deep tech alpha is emerging (including… space)

If you’re building or investing at the earliest stages, this one is worth a listen.

Check it out now:

• Apple Podcast

• Listen on Spotify

• Watch on YouTube

  • (0:00) Intro and welcome

  • (1:20) Origin story: the Rocket Lab Pelican case moment

  • (5:30) Why solo GP + deep tech was non-consensus in 2017

  • (13:08) How Halter got a yes in 45 minutes at Il Fornaio

  • (16:38) Moving at founder speed: the 24-hour clock

  • (27:35) Moonshot culture is strangling deep tech discipline

  • (40:06) Why Bessemer’s “Supernova” benchmarks can be misleading for founders

  • (46:50) Mistakes founders make when capital floods a category

  • (51:44) Ubiquity University and the DRY principle applied to VC

  • (56:30) The future: data centers in space and life beyond Earth

Ubiquity Ventures — led by Sunil Nagaraj — is a seed-stage venture capital firm focused on startups building software that reaches into the real world. In a screen-obsessed world, we focus on "software beyond the screen" startups, which include technology companies that apply AI, software, and smart hardware to physical problems and systems that you can touch, hear, and feel.

If your startup fits this description, reach out to us.

For our latest Ubiquity University module, I just sat down with Brannon Jones, a former SpaceX manufacturing engineer who worked on Falcon and Raptor and is now a deeptech VC at AlleyCorp, to unpack what actually makes SpaceX different from the inside. Not the mythology—the operating system. The consistent theme:

SpaceX has turned culture into a compounding execution engine.

1. Test Like You Fly. Test Often.

SpaceX prioritizes truth over elegance. Instead of trying to design the perfect system upfront, they push hardware into real-world conditions as quickly as possible. That includes tests that fail—sometimes visibly and dramatically—but internally, those are treated as progress because they collapse uncertainty.

The contrast with legacy aerospace is sharp:

• Design extensively to avoid failure

• Optimize before validating

versus:

• Build quickly

• Test in reality

• Iterate based on actual data

This approach compresses timelines and eliminates false confidence. In deep tech, especially in the physical world, you cannot simulate your way to certainty.

2. Ruthless Elimination of the Unnecessary

One phrase Brannon emphasized was avoiding anything “superfluous.” In practice, this is much harder than it sounds, but the mantra is “delete, delete, delete”. Most teams overbuild because they assume they understand the system early.

SpaceX assumes they don’t.

They start from first principles, keep designs as simple as possible, and let real-world testing determine what needs to exist. That discipline leads to systems that are cheaper, faster to build, and easier to improve.

A useful heuristic:

• If your roadmap feels fully specified, you’re probably overbuilding.

• If parts of it feel unresolved, you’re likely learning in the right places.

3. Manufacturing Is the Innovation Engine

A key insight from Brannon: SpaceX’s velocity is driven as much by manufacturing as by design. High-throughput production of complex components creates a rapid feedback loop between building and learning.

• Faster builds → faster tests → faster insights → faster redesigns

This flips the typical mental model. Manufacturing is not downstream of innovation—it enables it. The faster you can produce and iterate, the faster you converge on better systems.

This pattern is increasingly visible elsewhere:

• EV companies improving through production scale

• Robotics systems learning through deployment

• Hardware startups refining products in the field, not the lab

4. Extreme Clarity on Priorities

Another defining trait is how clearly priorities are communicated. At SpaceX, Elon regularly makes it explicit what the single most important thing at the company is, and that clarity propagates across the organization. Engineers, technicians, and new hires all operate with a shared understanding of the current objective.

This reduces friction in several ways:

• Less time spent aligning

• Faster decision-making

• Fewer conflicting efforts

It’s not always perfectly clean, but the organization consistently converges around what matters. In complex technical environments, that clarity is a force multiplier.

5. Ground Truth Over Abstraction

SpaceX also places a premium on direct, firsthand information. Brannon described how leaders often bypass layers to understand what is actually happening on the ground. The goal is not polished summaries—it’s reality.

That creates a system where:

• Issues surface earlier

• Individuals closer to the work have real influence

• Decisions are anchored in facts, not narratives

It also removes the ability to hide behind abstraction. The organization continuously orients itself toward what is actually happening.

Culture as a Moat

What ties all of this together is not any single tactic, but the system they create. Culture at SpaceX is not about stated values—it’s about how quickly the organization learns from reality and adapts.

• Fast feedback loops

• Direct information flow

• Clear priorities

• Continuous iteration

That combination produces something extremely difficult to replicate: a higher rate of learning.

In deep tech, that is the moat. Competitors can copy products, hire talent, and raise capital. But it is much harder to replicate how quickly a company improves.

When SpaceX IPOs, the market will try to value rockets. The harder—and more important—thing to value is a culture that learns faster than everyone else.

Watch the Full Breakdown on Ubiquity University

Brannon goes deeper on how these principles show up day-to-day inside SpaceX, and where they do—and don’t—translate to other startups.

Ubiquity Ventures — led by Sunil Nagaraj — is a seed-stage venture capital firm focused on startups building software that reaches into the real world. In a screen-obsessed world, we focus on "software beyond the screen" startups, which include technology companies that apply AI, software, and smart hardware to physical problems and systems that you can touch, hear, and feel.

If your startup fits this description, reach out to us.

Two 10-meter “eyeballs” sitting at over 13,600 feet atop a dormant Hawaiian volcano, peering deeper into the universe than anything on Earth. Last week, I had a very special opportunity to visit the Keck Observatory on Mauna Kea in Hawaii (thanks Loren!) and what stayed was the way physics, geography, engineering, and human effort all come together in one place.

1. The Single Best Spot on Earth for a Telescope

At nearly 14,000 feet on Mauna Kea, you’re above roughly 40% of Earth’s atmosphere, which meaningfully reduces distortion before the light even reaches the telescope. The mountain itself plays a role. Mauna Kea is a shield volcano, so its gradual slope keeps airflow relatively smooth rather than turbulent. Trade winds tend to push clouds below the summit, and the air is extremely dry, which is especially valuable for infrared observations.

There’s also a practical layer that’s easy to overlook. Despite being remote, the observatory is within driving distance of a skilled workforce—engineers, technicians, and support staff who keep everything running. Location, location, location.

2. The Eyeball Isn’t a Mirror. It’s a Network

Each Keck telescope uses 36 hexagonal mirror segments instead of one monolithic mirror. Each segment is about six feet across and weighs close to 900–1,000 pounds. Together, they are actively aligned to behave like a single 10-meter surface.

Large mirrors eventually run into physical limits—weight, deformation, manufacturability—so Keck solves this by coordinating many smaller elements in real time. The result feels closer to a distributed system than a single piece of hardware, with constant adjustments to maintain a unified optical surface.

3. Software Correcting Physics

Even at this altitude, the atmosphere still introduces distortion. Incoming starlight gets slightly blurred as it passes through turbulent air. Without correction, the performance gap between a backyard telescope and a much larger instrument would be narrower than you might expect.

Adaptive optics changes that. A deformable mirror adjusts its shape thousands of times per second to counteract atmospheric turbulence. The system measures the distortion and compensates for it in real time, restoring sharpness that would otherwise be lost.

This capability has enabled some of the most important scientific work done at Keck, including the observations that led to the Nobel Prize-winning discovery of the supermassive black hole at the center of the Milky Way. I wrote more about that here:

UBQT - The Ubiquity Blog

Startups, Black Holes, and Nobel Prizes

Exciting news today as Dr. Andrea Ghez has just won the Nobel Prize in Physics for her pioneering work on black holes. As a professor of astronomy at UCLA, she has spent the last 25 years trying to understand black holes and determine if one sits at the center of our own Milky Way galaxy. A few questions emerge - if black holes are black, how can you p…

Read more

6 years ago · Ubiquity Ventures

4. This Is Also a Factory

One of the more surprising aspects of Keck is how much ongoing work happens on-site. The mirrors are not static assets. Each segment is periodically removed, stripped, and recoated with a fresh aluminum layer only about 120 nanometers thick, then reinstalled and precisely aligned.

All of this happens at 14,000 feet. There’s a full operational ecosystem on the mountain—mechanical shops, technical staff, and processes designed for continuous upkeep. The observatory runs most nights of the year, and maintaining that level of performance requires constant attention.

5. The Journey Up (a.k.a. Oxygen Is Optional)

The experience of getting there makes the system more tangible. You stop first at Hale Pōhaku, around 9,000 feet, to acclimate. From there, the drive continues up to about 13,600 feet.

The change in atmosphere is noticeable. Barometric pressure drops from ~1,000 hPa at sea level (14.7 psi) to around 600 hPa at the summit—about a 40% reduction.

Movement slows a bit. You become more aware of each step. Then you walk inside and see the primary mirror up close.

It’s hard not to pause. Thirty feet of precision, perfectly aligned to enable us to see the other side of the universe.

6. Small Human, Big Machine

One moment that stuck with me was seeing my own reflection in the mirror—small, almost easy to miss. It’s a reminder of the scale of the instrument and the scale of what it’s designed to observe.

7. What Keck Has Actually Done

Keck is also one of the most scientifically productive observatories in the world. Over the past few decades, it has contributed to a wide range of discoveries, including:

• The supermassive black hole at the center of the Milky Way

  Precise tracking of stellar orbits (Andrea Ghez, Nobel Prize 2020) — see our UBQT blog post about this

• Exoplanet confirmation at scale

  Radial velocity measurements helping confirm thousands of Kepler candidates as real planets around other stars

• Characterization of exoplanet atmospheres

  Detecting chemical signatures such as water vapor

• Observations of distant galaxies in the early universe

  Extending our understanding of how galaxies form and evolve

The observatory has been operating since the 1990s and continues to play a central role in modern astronomy.

8. The Outside View

From the outside, the telescope is a dome on a mountain. Inside, it’s a tightly coordinated system built to extract signal from noise with extraordinary precision.

Final Thought: This Is What Deep Tech Looks Like

Keck has remained at the forefront for decades through a combination of thoughtful site selection, materials science, mechanical design, control systems, software, and ongoing operational discipline — something that became incredibly clear to me after spending the day with the scientists and engineers that run this amazing facility. It’s a reminder that progress often comes from integrating many layers of expertise rather than relying on a single breakthrough.

If you want a concrete example of software meeting the physical world in a meaningful way, Keck is a good place to start.

Ubiquity Ventures — led by Sunil Nagaraj — is a seed-stage venture capital firm focused on startups building software that reaches into the real world. In a screen-obsessed world, we focus on "software beyond the screen" startups, which include technology companies that apply AI, software, and smart hardware to physical problems and systems that you can touch, hear, and feel.

If your startup fits this description, reach out to us.

Not how we build models, but how we run them.

AI is not bottlenecked by models anymore. It’s bottlenecked by inference—the cost, efficiency, and infrastructure required to actually deploy these systems at scale.

And right now, that layer is breaking.

Every week, new ML techniques unlock meaningful gains—quantization, sparsity, better attention mechanisms, longer context. But those gains rarely show up in production. Not because the ideas are wrong, but because the hardware can’t keep up.

Hyperscalers made a reasonable bet: deploy GPUs at massive scale and refresh every few years. That worked when models evolved slowly. It breaks when model optimization evolve weekly.

Your hardware is already out of date.

We’re trying to run modern AI on hardware that acts more like a fixed-function calculator than a programmable system. And then we try to fix it by buying more GPUs. That’s not scaling. That’s scaling a broken system.

ElastixAI exists to fix this.

The Hidden Tax on Every Token

This mismatch shows up everywhere once you look for it.

New ML optimizations should deliver step-function improvements. In practice, they get watered down because existing hardware can’t support them natively. A technique like 4-bit quantization should unlock dramatic gains, but when you force it through hardware that wasn’t designed for it, most of the benefit disappears.

At the same time, the cost structure is moving in the wrong direction. Memory is becoming the dominant expense, with high-bandwidth memory costing an order of magnitude more than traditional DRAM. Compute is underutilized—large portions of modern GPUs sit idle during inference workloads. And power requirements are ballooning to levels that strain real-world data center constraints.

We are building increasingly powerful models on infrastructure that wastes most of its capability.

That’s not a marginal inefficiency. That’s a structural problem.

Why Elastix Exists

Elastix flips the stack.

Instead of forcing models to conform to hardware, Elastix makes hardware conform to models. The key is reconfigurable compute—specifically, FPGAs.

Unlike GPUs, which are fixed after manufacturing, FPGAs can be reconfigured at the silicon level. Elastix takes advantage of this by effectively compiling ML models into hardware, allowing each model to run on a system tailored to its exact needs.

That means new ML techniques can be implemented directly in hardware, without software workarounds. It means better utilization, lower memory overhead, and dramatically improved efficiency. And it means you’re no longer waiting years for the next chip cycle to benefit from advances in ML.

Early results are striking: significant cost reductions versus GPU-based systems and the potential for 10x-50x cost efficiency gains, over various user latencies, as models continue to evolve.

What Would You Do With 100x More Tokens?

The right way to think about Elastix is not “is this cheaper?”

It’s: what becomes possible if inference is dramatically more efficient?

If the cost of generating tokens drops by an order of magnitude or more, entirely new categories of applications open up. Fast reasoning becomes viable in places it isn’t today. Power-constrained environments can run meaningful models locally. Context windows can expand without blowing up cost. AI starts to look less like a premium feature and more like infrastructure.

This is the kind of shift that doesn’t just improve existing use cases—it creates new ones.

Why Now

This approach only works now because several trends have converged.

First, ML innovation is outpacing hardware cycles by a wide margin. Second, inference has become the dominant cost center for many AI applications. And third, the ecosystem has matured to the point where a handful of ML models are generating multiple billion dollars in revenue.

ElastixAI is not just a hardware play—it’s a tightly integrated stack spanning ML optimization, software development, and hardware reconfiguration. This is a real-time HW-SW-ML co-design. It means the hardware can evolve with the same pace as software and machine learning optimizations all of this without the need of taping out a new silicon.

That combination is what allows them to capture the gains that today’s systems leave on the table.

Why ElastixAI Matters

Every major AI player today is making the same implicit bet: that the current GPU-centric stack will hold.

Elastix is making the opposite bet—that the future belongs to adaptive compute systems that evolve at the pace of software.

If that’s right, this is not a niche optimization. It’s a new layer of the AI stack.

Lower costs lead to more usage. More usage leads to new applications. And new applications expand the market. This is the kind of compounding effect that defines platform shifts.

The Team That Should Be Building This

This is not a generic hardware startup.

The founding team previously built Xnor.ai, pioneering deep learning quantization and efficient computer vision algorithms such as Xnor-Net and YOLO, which was acquired by Apple. They then spent years inside Apple and Meta working on large-scale ML systems, including contributions to Apple Intelligence and LLaMA-class models.

They’ve worked at the exact intersection this problem requires: ML algorithms, systems, and hardware. And they’ve done it both in startups and inside some of the most advanced AI organizations in the world.

Where They Are Today

Elastix is still early, but the core system is working and demonstrating meaningful efficiency gains. The company is based in Seattle and backed by Ubiquity and a strong set of investors.

The focus now is moving from proving the technology to scaling it in real-world deployments. They are already generating tokens using a variety of LLM models in hundreds of billion parameters and running them on multiple off-the-shelf FPGAs which are available to scale in the market.

The Bottom Line

AI is going to be constrained by economics, not intelligence.

Whoever fundamentally improves inference efficiency doesn’t just win on cost—they unlock entirely new capabilities.

Elastix is attacking that problem at its root. Not by building slightly better models or slightly better chips, but by rethinking how models and hardware interact in the first place.

If they’re right, every token gets cheaper, every model gets faster, and a whole new class of applications becomes viable.

That’s the kind of shift worth paying attention to.

Ubiquity Ventures — led by Sunil Nagaraj — is a seed-stage venture capital firm focused on startups building software that reaches into the real world. In a screen-obsessed world, we focus on "software beyond the screen" startups, which include technology companies that apply AI, software, and smart hardware to physical problems and systems that you can touch, hear, and feel.

If your startup fits this description, reach out to us.

Ever since Claude 4.5 Opus and other similarly powerful models came out at the end of 2025, it’s old news that individuals can get 10x or more boosts in their output when they know how to leverage it.

Many draw the analogy to human computers in the 1960’s and how a modern spreadsheet on a laptop is faster than entire skyscrapers of human computers. And imagine if one cell of that spreadsheet had to be manually computed by a human: it would bottleneck the entire sheet by orders of magnitude, waiting for the one step that needs human attention. That’s how many agent-powered systems are today: incredibly fast and powerful, but still bottlenecked by the required human steps. This is even worse at the organization level.

So, in an organization, how do we remove the human bottlenecks on this spreadsheet?

This is the trillion dollar question, and we can’t answer it without next generation models. But we also need to change our organizations.

The first step is that everyone in the organization needs to be on the same page and using AI. This is easier said than done.

I see plenty of CEOs pounding the table and writing memos about how they demand AI usage. And many of them are good memos. But it’s not helpful to just be demanding: how is everyone in the org supposed to use AI? For some roles, it’s easy, but for others, it’s much harder. Put yourself in the shoes of an employee: this whole revolution is a bit terrifying in the first place, changing how you work is hard, and often it’s unclear how to get the real productivity gains. Making matters worse, even if you you are good at individually using AI, you are bottlenecked by all the human communication in the org, and the AI doesn’t have the tooling to navigate it.

I have an idea: let’s just fire everyone who isn’t AI-pilled and addicted to this stuff like us!!!

Those LOSERS are just going to get left behind!! Sucks to suck.

No, of course not. That would be a disaster while your org relies on load-bearing human positions, and it’s unnecessarily unfair to humans who are slower to change. Your senior security engineer might be less inclined to use AI because of how much security risk AI introduces.¹ And if you fire your security person who is incredibly smart and knowledgeable but just hasn’t seen the light on AI yet, and you haven’t properly invested in showing it to them in a fun, caring, compassionate way, you’re an idiot.

¹ True of any new tool, and proportional to the power of the tool.

The answer: Socially engineer your org to use AI

Step 1: Agents in the chat app

If you want your org to use AI better, it needs to be a social thing. The problem with current AI tooling is that it is incredibly individualistic. Humans are social creatures, we pick up on signals, we want to signal about things we care about, and we rely on the direction of the herd for 90% of our behavior. If you want AI use to spread organically, it has to be social.

Unfortunately robotics in the workplace is a ways behind LLMs. So, we need to get AI usage directly in your primary company chat app. For most startups, this is Slack. I personally hate Slack and use Zulip instead², but there are plenty of solutions for getting AI agents in Slack nowadays: go read up on it and find one that works.

Some people mistakenly think that the purpose of having agents in Slack is convenience. It’s lower friction to send a message in Slack than to open a code terminal or go to claude.com. But if anything Slack is a worse interface than a dedicated UI, so conversing with it there runs the risk of becoming a crutch.

But the value of agents in your chat app is the social experience. The goal is to build a group chat that is a symbiotic combination of humans and AI agents, conversing natively in a single town square.

The point of agents in the chat app is to build social dynamics, not convenience.

² See my post on zulipllm for a small taste of why Zulip is so much better.

Step 2: Keep important discussions public

The biggest hurdle that one runs into with agents is information security and privacy. If the agent has access to private conversations, and it has any form of memories, you are in for a bad time unless you are really careful. Being careful slows things down and makes it less fun, so you want to build things to be inherently safe.

The solution to this is to only give these social agents access to public channels. This also means that you have to actually be having useful conversations in public. WindBorne was already built this way, so it was easy for us. I hate the mental overhead of thinking about who can see what, so I’m just radically transparent except for the minimum necessary set of sensitive information. I believe part of what made this possible was Zulip’s core features of topics within channels.

Step 3: Put your money where your mouth is

You’re going to need to spend a lot on tokens. This has already been widely talked about. But I also mean that you need to assign some of your best engineers to building good AI infrastructure. You can’t just plug in a SaaS product and call it a day, and you can’t just hire someone new to do it for you: someone needs to deeply understand how your system works and build tooling around it. Plan to hire additional people to keep expanding things along the way, but it starts with an existing engineer.

At WindBorne, I reassigned one of our best engineers, Davy Ragland, to this full time. It was a real cost to what he was previously working on and required others to step up. But he built SHODAN, our primary internal AI agent, and it paid off enormously. Treat it like they quit their old job, and commit to the change.

Step 4: Make your org searchable by agents

In order for agents to be useful, they need context. Unless you give them the tools to find context the way you would onboard a human, they won’t seem that useful. This is at the root of a lot of the problem with individual AI use: every individual has to learn on their own how to give the AI context.

Today this looks like setting up a bunch of MCPs and command line tooling to give agents access to things. Most important of all is message search: if you want them to be able to act native in your chat app, they need to be able to find messages like a human can. At WindBorne, after doing step 3 in this guide, the first thing we did was build out and optimize vector search of all Zulip messages for both humans and agents to use.

Step 5: Lead by example in public

This one is pretty straightforward: you set up agents in your chat app. They have access to information, and infrastructure to support them. Now lead by example and ping the AI when relevant in conversation. Build the muscle and the habit. Even if it’s set up well, and even if you’re a big fan of AI, it takes some effort to change habits, and it’s important you demonstrate this yourself.

Step 6: Bake in usage

If you did steps 1-5 well, AI agent usage in your chat app should already be spreading through your org like wildfire. At WindBorne, we quickly grew to $20k / month spend with SHODAN (and then made some optimizations to manage spend, but even at that cost it’s well worth it).

But, there may be pockets it doesn’t reach as much, and if you’re investing less time and energy than I did, results may be more muted at first. So you want guaranteed entry points to social AI usage.

The best example I have is that we made SHODAN be the onboarding buddy for every new hire. Now, no matter what background you come from, even if you only played with a free-tier model 9 months ago and thought it sucked (they do), when you join WindBorne you get the sci-fi experience of interacting with a top-tier model, with all the right context loaded up, operating natively in a UI that was otherwise designed for humans. This sets you up for success from day 1.

A note on security and caution

Any time an agent can cross an information security boundary, there are risks. So in this post, I advocate for giving your agent access to all internal public-visibility messages. But this means the same bot can NOT have inbound information from people outside the org, and it shouldn’t be able to see DMs.

Consider the visibility levels of the following 3 things, ordered by level of privacy:

1. A DM between two people in your org

2. A public channel in your org’s chat app

3. The open internet

A current-day AI agent shouldn’t be able to write to a lower level if it has access to information from a higher level, unless there is meticulous infosec design and monitoring. If it can read your DM, then it shouldn’t be able to post in public in your org, because then it can leak information. And each set of DMs has individually mutually exclusive privacy, so a bot that has memory and is active across all DMs is a nightmare to manage, unless you know what you are doing.

Email is an interesting subject: it’s private read, but there is essentially public write access: anyone can just email you. This means that an AI that is reading all of your emails as they come in is susceptible to prompt injection attacks.

I don’t write this to scare anyone away from playing around. Early humanity had to literally play with fire to harness its power and get to where we are today. Take these notes on security as reason to learn more yourself, not reason to shy away.

Where from here?

The AI singularity is going to transform everything, that much is obvious. But it’s very hard to predict exactly how and when specific things will change.

If you are a CEO, you need to be plugged in and playing around every day. You cannot just buy some off-the-shelf SaaS wrapper for AI and expect it to plug in seamlessly. You need to re-wire your org with AI. Only the CEO has the vantage point and authority to execute on this: don’t let your team down.

This was a guest post from Ubiquity-backed WindBorne Systems Co-founder & CEO John Dean.

Ubiquity Ventures — led by Sunil Nagaraj — is a seed-stage venture capital firm focused on startups building software that reaches into the real world. In a screen-obsessed world, we focus on "software beyond the screen" startups, which include technology companies that apply AI, software, and smart hardware to physical problems and systems that you can touch, hear, and feel.

If your startup fits this description, reach out to us.

Today, we’re releasing a new Ubiquity University module featuring Halter’s VP of Product & Engineering, Andy McLaren, on one of the hardest problems in deeptech: shipping Gen 1 hardware. If you’re building “software beyond the screen,” this is where theory meets mud, rain, broken parts—and reality.

You can watch the full session, but below are Andy’s three hard-earned lessons after eight years of building collars that live on cows, outdoors, for 5–6 years straight.

Lesson 1: Get to the real thing—faster than feels comfortable

Most teams iterate safely. Hardware punishes that instinct.

Andy’s core idea is to pull risk forward as aggressively as possible. The slides capture the progression well: rapid prototyping for fit and form, real materials for durability and function, and then integration as early as possible . But the real insight is how quickly you move through those stages—and how willing you are to abandon work.

At Halter, this wasn’t theoretical:

• Early prototypes looked correct, but leaked water immediately

• Vacuum-cast parts appeared production-ready, but tore apart in the field

• Fully integrated systems exposed failure modes no one had even considered

Rather than polishing intermediate steps, they accelerated toward the “final” version. Within roughly two months, they scrapped an entire design path and jumped directly to real tooling and contract manufacturing.

The uncomfortable truth is that the fastest way to learn in hardware is often to build something that looks finished—and then let it fail. Manufacturing, in particular, is not a downstream step. It is the product. As Andy puts it, “prototypes are easy, production is hard” .

Lesson 2: Selectively innovate (or waste years)

Early-stage teams, especially technical ones, have a bias toward innovation everywhere. Hardware companies don’t have that luxury. Andy’s second lesson is a discipline: be ruthless about where you innovate—and equally deliberate about where you don’t.

At Halter, the team focused innovation on the core problem: training animals with a collar and enabling virtual fencing. That’s the breakthrough. Around that, they deliberately avoided reinventing the wheel:

• Used existing buckle designs rather than inventing a new attachment system

• Leveraged experienced contract manufacturers instead of building novel processes

• Bought off-the-shelf testing equipment instead of designing custom systems

One story captures this well. Early on, they built a beautifully simple “slip-on” collar—no buckle, seamless user experience, almost elegant in its minimalism. It also fell off the cow in about five minutes. Meanwhile, a standard buckle had been working reliably in the market for years.

They chose the buckle.

That decision sounds obvious in hindsight, but it reflects a deeper trade-off founders constantly face: ego versus progress. The winning teams innovate where it truly matters and show humility everywhere else.

Lesson 3: Track—and obsess over—field failures

Once hardware is in the field, the job changes. You’re no longer primarily building—you’re learning.

Andy emphasizes that many of the most important failure modes simply cannot be replicated in a lab. Real-world environments introduce edge cases you will never predict. The only way to improve is to instrument, observe, and debug relentlessly.

At Halter, this became a core competency:

• Every deployed unit is monitored for performance

• Failed devices are retrieved whenever possible

• Root causes are investigated deeply and systematically

• Failure modes are categorized and tracked over time

This image below shows a real distribution of failures—everything from corrosion to capacitor degradation to vent punctures . One particularly instructive example was a waterproof membrane vent that passed all lab testing. In the field, however, small blades of grass could work their way into the housing and puncture it. No simulation caught it. Only real-world exposure did.

Because they had tight feedback loops, they were able to identify the issue, understand the root cause, retrofit deployed units, and fix future designs. That cycle—deploy, fail, learn, fix—is what turns Gen 1 into something durable.

Andy also highlights a useful framework for thinking about failures:

• Infant mortality: early defects, often tied to manufacturing issues

• Random failures: unpredictable environmental events

• Wear-out failures: degradation over time

Each category behaves differently and requires a different response. Understanding which one you’re dealing with is half the battle.

Why this matters now

There’s a growing narrative that AI has made building companies easier. For software, maybe. For hardware, not even close.

If anything, expectations are higher. Customers don’t care how advanced your models are if your device leaks, breaks, disconnects, or fails in the real world. This is why we’re so focused on this category at Ubiquity. It’s where AI meets physics—and physics always wins.

Final thought

Halter didn’t reach a $2B valuation by getting Gen 1 right the first time. They got there by getting Gen 1 wrong quickly, learning faster than anyone else, and compounding those lessons in the real world.

If you’re building something that has to survive outside a data center—on a farm, in a factory, on a satellite, or in a hospital—you’re not just writing software. You’re shipping reality. And reality is an unforgiving customer.

If you want the full breakdown from Andy, the Ubiquity University session is worth your time. And if you’re building something that feels like software beyond the screen, we would love to hear from you.

Ubiquity Ventures — led by Sunil Nagaraj — is a seed-stage venture capital firm focused on startups building software that reaches into the real world. In a screen-obsessed world, we focus on "software beyond the screen" startups, which include technology companies that apply AI, software, and smart hardware to physical problems and systems that you can touch, hear, and feel.

If your startup fits this description, reach out to us.

It started with an idea that sounded… unusual: collars for cows.

We recently sat down with Craig for a Ubiquity University module on raising your first round as a deep tech founder. His advice is refreshingly practical—no jargon, no theatrics, just lessons earned the hard way. Watch our 16-minute Ubiquity University module now — “Raising your First Round for your Deeptech idea with Halter CEO Craig Piggott”:

Our journey with Halter

We first met Craig Piggott in a Palo Alto coffee shop in 2018. Ubiquity co-led Halter’s seed round just a few days thereafter. At the time, there was no playbook for what he was building—no obvious comps, no clear category. Just a sharp founder, a deeply personal connection to the problem, and a willingness to explain, over and over again, why this mattered.

If you want the full arc, we’ve written about it in the past:

• In 2021, when Halter raised $23M, we focused on early conviction: Backed by $23 million of fresh capital: Craig Piggott, Founder of Halter

• In 2025, when Halter became a unicorn, we unpacked how a “weird” idea became inevitable: Did You Just Say “Cows”? How Halter Became a Unicorn by Reimagining Livestock Management

Today’s announcement is another milestone on that journey. But the real lesson is how Halter raised its first round when the idea still sounded strange. That’s exactly what our new Ubiquity University module with Craig unpacks.

The Truth About Deeptech Fundraising

At Ubiquity, we live at the earliest stage. A third of the companies we back, we invest the day they’re incorporated—no revenue, no metrics, often not even a product.

And when the idea is off the beaten path—our angle is what we call “software beyond the screen”—the fundraising looks very different. You’re not just pitching a company. You’re teaching a new way of seeing the world.

That’s exactly what Craig had to do.

Ubiquity University: Craig’s Playbook for Raising Your First Round

We’ve summarized some of Craig’s best tips here, but we still recommend watching the full interview on Ubiquity University.

1) It’s your job to make people understand.

If your idea is novel or confusing, that’s not the investor’s problem—it’s yours. Early on, Craig saw investors misunderstand Halter constantly. The instinct is to blame the audience — but that’s wrong; the better move is to simplify the explanation, then simplify it again, until it clicks.

2) Start simple. Earn the right to get complex.

Deep tech founders often lead with sophistication and big vision. That’s a mistake. Craig learned to begin with something almost anyone could grasp—“we help farmers move cows without fences”—and then layer in complexity. Clarity isn’t dumbing it down; it’s widening the funnel.

3) You don’t have leverage (yet).

As Craig puts it, “you don’t have any leverage until you have a term sheet.” Early founders sometimes try to balance the power dynamic too soon. At the beginning, your job is to communicate clearly, be responsive, and build momentum. Leverage is earned, not assumed.

In a past post, we dove into this idea of when you do and when you don’t have leverage: “0, 1, and ∞ (Infinity): The Only Three Numbers That Matter to Entrepreneurs and Astronomers”

4) Pitch broadly. Iterate fast.

By the end of a round, Craig estimates they were on version 15 or 20 of their pitch. That’s not a failure—that’s the process. Every conversation sharpens the story: where people get confused, where they lean in, what questions repeat. Early on, you don’t get selective; you get better.

5) Show, don’t tell.

Especially for real-world systems, words aren’t enough. Photos, videos, demos—anything that helps someone see what you’re building—can unlock understanding instantly. For Halter, this was critical.

6) Investors are backing you, not just the idea.

At the earliest stage, everything will change—the product, the pitch, even the market. What endures is the founder. Craig’s authenticity and depth of understanding mattered far more than polish.

Why This Matters

It’s tempting to look at Halter today—over a million cows on the platform and a $2B valuation—and assume the path was obvious.

It wasn’t. It rarely is for ideas that matter.

The uncomfortable truth is that many off-the-beaten-path ideas fail not because they’re wrong, but because they’re poorly explained.

At Ubiquity, our job is to back those ideas before they’re obvious. But just as importantly, it’s to help founders tell their story in a way the world can actually understand—so those ideas have a real chance to reach their full potential.

We’re excited about today’s milestone. But the part we care about most is what comes next: the next founder with an idea that sounds just as strange as “cow collars” once did—and the chance to help them bring it to life.

Ubiquity Ventures — led by Sunil Nagaraj — is a seed-stage venture capital firm focused on startups building software that reaches into the real world. In a screen-obsessed world, we focus on "software beyond the screen" startups, which include technology companies that apply AI, software, and smart hardware to physical problems and systems that you can touch, hear, and feel.

If your startup fits this description, reach out to us.

That gap—massive real-world systems underserved by software—is exactly where we invest.

And only recently has it become tractable (thanks to mobile devices, real-time data, and AI).

Enter BusRight.

This morning, BusRight announces a new $30 million raise to modernize student transportation—and Ubiquity has a been proud partner since 2022.

Why BusRight Matters

Student transportation is one of the most complex logistics systems in the country.

• Routes change constantly

• Driver shortages are worsening

• Parents expect real-time visibility

• And there’s zero margin for error

When transportation breaks, students don’t get to school. Education itself breaks.

And yet ~13,000 transportation directors across the U.S. are asked to manage this with tools that haven’t meaningfully evolved in decades.

BusRight brings together dynamic routing, real-time communication, driver navigation support, and student tracking – turning analog chaos into digital coordination.

Just see how Brandon Caldwell, Transportation Director at Madison School District, explains why he loves BusRight in this 90-second video:

Customers value time, confidence, and peace of mind. Now the economic impact is becoming just as clear:

“BusRight saved our district $989,000 in the first year,” said Gregory Mott, Superintendent at Poughkeepsie City School District.

Inside the Product: From Chaos to Coordination

BusRight unifies a fragmented system, giving transportation teams the tools, visibility, and confidence they need to run their operation without the constant stress. Routing, communication, driver support, and real-time tracking all work together, turning what used to feel overwhelming into something manageable. Parents get real-time location of their child’s school bus, whether students safely badged onto the bus, and more.

For leaders looking at the bigger picture, it feels like the industry finally has something built for them.

“These tablets are our lifelines,” said Samantha Steed, Transportation Supervisor at Pocatello Schools. “They give us critical information in real time to make sure everyone on board is safe.”

BusRight is not just solving operational problems. It is doing it in a way that feels personal. The product is shaped by the people using it every day, built alongside the communities it serves.

Because at its core, this is not just about software. It is about making sure a student gets to school safely. It is about giving a transportation director their evenings back. It is about restoring trust in a system that millions of families rely on every single day.

What does the new $30 million of capital enable?

We are doubling down on two things: talent and product.

We believe there are only two roles that matter: you are either building or you are selling. AI is expanding what individuals can do, turning individual contributors into full-stack builders and sellers. We want BusRight to be the place where ambitious people come to do both, to own outcomes end to end, grow faster than they thought possible, and have a real impact on an industry that matters.

On the product side, we are building what we believe is the future of student transportation. We are developing the first student transportation agent, Joy AI, designed to act as a right hand for transportation leaders. Instead of reacting to problems, leaders will be able to get ahead of them, with real time insights that help them make better decisions every day. In many ways, Joy AI isn’t just a tool—it’s a step toward the first operating system for student transportation.

The Bigger Picture

1.5 billion students attend K-12 schools globally, and getting them to and from school is one of the most challenging aspects of educating the next generation. The U.S. represents only about 3% of that population, and BusRight today serves just a fraction of it.

The challenges of student transportation do not stop at district, state, or national borders. They are global, structural, and increasingly complex. BusRight is building the operating system for student mobility in the physical world.

The role of AI longer-term

In a system as dynamic as student transportation, the value of AI isn’t automation—it’s anticipation—predicting problems before they escalate.

For a long time, building something meaningful meant being constrained by time, tools, and bandwidth. Most people spent their days stuck in front of a screen, moving slowly from idea to execution.

AI changes that. It expands what one person can do. It shortens the distance between seeing a problem and actually solving it.

But the real unlock is not technical, it is human. When you remove the busywork, people can spend more time in the physical world, talking to customers, understanding nuance, and building things that actually resonate. That is where great companies come from.

Final Thought

This is what “software beyond the screen” looks like.

Not another SaaS tool.

But software that quietly, reliably moves millions of people through the real world—every single day.

That’s a category worth building—and backing.

If you enjoyed this post, read more in our June 2024 founder profile of Keith

Ubiquity Ventures — led by Sunil Nagaraj — is a seed-stage venture capital firm focused on startups solving real-world physical problems with "software beyond the screen", often using smart hardware or machine learning.

If your startup fits this description, reach out to us.

Set up a Pitch Meeting

Don’t compare your beginning to someone else’s middle.

You pick up tennis for three weeks and wonder if Serena Williams was just “natural.”

You record your first YouTube video and assume MrBeast was born with a ring light and viral instincts.

From afar, excellence looks effortless. Up close, it’s decades.

And here’s the corollary — one I feel strongly about:

Saying “they’re just natural” isn’t praise. It’s an erasure.

It erases the thousands of hours. The awkward early drafts. The trade-offs. The obsession. The lonely reps.

When we call someone “natural,” we flatten their story.

The Magic of YouTube: You Can Actually SEE the Beginning

In most fields, you can’t rewind history. You can’t easily watch a 12-year-old Serena miss her backhand. You can’t watch a first-time founder stumble through their first pitch deck.

But YouTube is different. It preserves the beginning — and it’s glorious.(Just search for a YouTuber, click on “Videos”, “Oldest”)

Below are three of the biggest and most famous creators we all think of as polished, confident, and “natural.” Let’s rewind to their beginnings:

Mark Rober

Early Mark Rober (October 29, 2011): Awkward camera angles. Basic editing. Finding his voice.

Today’s Mark Rober (70 million subscribers): Hollywood-level production. Precision storytelling. Masterful pacing. Tens of millions of subscribers.

Was he “natural”?

Or did a former NASA engineer decide to nerd out — relentlessly — for over a decade?

MrBeast

Early MrBeast (September 19, 2015): Grainy gaming videos. Quiet commentary. A teenager experimenting.

Today’s MrBeast (a decade later… 450 million subscribers!): Global productions. Massive logistics. Industrial-scale creativity.

He didn’t wake up one morning and know how to orchestrate million-dollar challenges.

He spent years studying thumbnails. Studying retention curves. Studying the craft.

Total nerd behavior.

Marques Brownlee (MKBHD)

Early MKBHD (Jan 1, 2009): Teenager in his bedroom. Webcam. Low lighting. Straight into the lens.

Today’s MKBHD (20 million subscribers): Cinematic studio. Clean delivery. Industry-defining reviews.

He has been publishing for over 15 years. Fifteen.

That’s not “natural.”

That’s reps.

The Middle Is Built on Invisible Trade-Offs

Here’s what the middle really means:

• Turning down distractions.

• Choosing practice over comfort.

• Making thousands of small refinements no one applauds.

• Repeating something long after the novelty wears off.

When you see someone “in their middle,” you’re seeing the compounding of obsession.

And this matters deeply in startups.

At Ubiquity, we back founders who are willing to nerd out for a decade on something unglamorous — perception stacks, robotics middleware, agricultural sensing, logistics optimization.

The outside world might say, “Oh, they’re just brilliant.”

No. They’re committed. There’s a difference.

And in venture, we bet on the difference.

Why This Is Encouraging

If excellence were “natural,” you’d either have it or you wouldn’t.

Game over.

But if excellence is built — slowly, awkwardly, imperfectly — then your messy beginning is not a verdict. It’s a starting line.

The most inspiring part of YouTube’s time capsule isn’t that creators improved. It’s that they kept going. And one day, without realizing it, they were no longer at the beginning.

They were in the middle — and now someone else is watching them, thinking, “Wow. They’re just natural.”

Don’t make that mistake. If you’re at the beginning — good. Lean in. Nerd out. Put in the reps.

One day, someone will compare their beginning to your middle.

And you’ll know exactly how much work that illusion required.

Ubiquity Ventures — led by Sunil Nagaraj — is a seed-stage venture capital firm focused on startups solving real-world physical problems with "software beyond the screen", often using smart hardware or machine learning.

 If your startup fits this description, reach out to us.

That new car smell? All downhill from there.

A microwave, a watch, a book, a factory machine—on day 1 they were perfect, pristine, complete. From that moment forward, entropy took over. Parts wore out. Interfaces aged. Capabilities stayed frozen in time while the world moved on.

That mental model is now obsolete. Permanently.

The day you buy a product is no longer its peak. It’s the starting line.

If you buy the right thing today, it will be better next week. Smarter next month. Meaningfully more capable a year from now—without you lifting a finger.

This shift is subtle but profound, and it’s the backbone of why we focus so obsessively on software beyond the screen at Ubiquity Ventures. (Watch our 5-minute explainer)

Consider a modern car—say a Tesla. You don’t just buy hardware. You buy into a roadmap. New driver-assist features, performance improvements, safety upgrades—delivered over the air, long after the keys are in your pocket.

The same is true for your Apple devices. Your iPhone and Apple Watch are not static objects. They’re living systems, quietly improving while you sleep.

Or take one of our portfolio companies, Halter. Their cow collars—now deployed on over half a million cows—don’t just sit there doing the same thing forever. They receive regular software updates that unlock new capabilities after installation. Farmers aren’t buying a fixed feature set. They’re buying progress. (Read more in our UBQT post “Did You Just Say “Cows”? How Halter Became a Unicorn by Reimagining Livestock Management.

Unlocking nine extra months of software development on the way to Mars

This phenomenon is so powerful that we’re now doing it on other planets.

NASA’s Mars 2020 mission launched the Perseverance rover toward Mars on a nine-month journey through space—while engineers kept writing and improving its software back on Earth. Just before arrival, Perseverance got a brain upgrade. It literally landed better than it launched.

What this means: the good and the bad

That has never happened before. Not once. In any era.

But this brave new world cuts both ways.

When software can improve hardware, it can also be withdrawn. Belkin recently ended support for its WeMo smart switches—stranding millions of perfectly functional devices (including about a dozen that I own). The risk is real: when you buy a product, you’re also betting on the company behind it.

This leads to the most important implication of all.

You’re no longer buying a spec sheet.

You’re buying a trajectory.

You’re locking hands with a team, a roadmap, and a vision for how that product evolves over time. That creates enormous upside—continuous improvement, new features, expanding value. It also demands more discernment from customers and enterprises alike. Who are you trusting with your future?

This is the world we’re building toward: real-world products that solve physical problems—and get better after you own them.

Buy it now.

It gets better next week.

That’s not a tagline.

It’s a permanent change in how progress works.

Ubiquity Ventures — led by Sunil Nagaraj — is a seed-stage venture capital firm focused on startups solving real-world physical problems with "software beyond the screen", often using smart hardware or machine learning.

 If your startup fits this description, reach out to us.

Every once in a while, you get something rarer.

In 2014, while I was at Bessemer Venture Partners, I made a $1 million seed investment in a then-tiny company called Auth0. Here’s the actual investment memo from July 2014 to lead their seed round at which point I joined Auth0’s board. Less than 12 months later, I went on to lead their $7 million Series A. For over 2 years, serving as the only investor on their board, I got to work shoulder-to-shoulder with the founders as we worked to build an enduring company around their vision of a developer-centric identity platform.

Seven years later, Auth0 sold to Okta for $6.5 billion.

That outcome has been written about before—including in my 2021 blog post, The Power of “Nerdy and Early” Investing.

But next week, that story gets something it almost never gets in venture: a live, unfiltered sequel.

A rare rewind—live in Seattle

On the evening of Monday, February 2, I’ll be hosting a fireside chat in Seattle with Auth0’s co-founder and CEO, Eugenio Pace.

👉 Event details & RSVP: https://luma.com/hz8rtzdp

We’ll talk candidly about:

• What Auth0 looked like at seed—before the logo, before the narrative, before product-market fit

• How culture and leadership building morphed at each stage of the business.

• How trust between founders and early investors really gets built—and tested

• And what seed-stage founders routinely underestimate on the road from “interesting” to enduring

If you’re building from zero—or backing companies that are—that’s where the real signal lives.

Why this still matters

It’s easy to compress success into a headline: “Startup X sells for billions.” What gets lost is the long middle—the years of ambiguity, product rewrites, pricing debates, hiring mistakes, existential doubt, and slow, compounding progress.

What made Auth0 special wasn’t just the outcome. It was the consistency of the partnership from the earliest days. I left Bessemer in mid-2017 to start Ubiquity Ventures, but that relationship endured—because it was built on shared conviction, not convenience.

That arc—from first check to final chapter—is vanishingly rare in venture. Which is exactly why it’s worth studying.

Watch the 2021 interview (now public)

In December 2021, Eugenio and I sat down for a long-form conversation with TechCrunch about Auth0’s journey. Until recently, that interview lived behind a paywall. It’s now fully public—and well worth watching ahead of the event:

Final note

If you’re a founder in Seattle—or an operator thinking seriously about building something enduring—I’d love to see you there. Stories like this don’t repeat often. When they do, it’s worth slowing down and asking why.

From seed to $6.5B doesn’t happen by accident. Let’s talk about how it actually happens.

Ubiquity Ventures — led by Sunil Nagaraj — is a seed-stage venture capital firm focused on startups solving real-world physical problems with "software beyond the screen", often using smart hardware or machine learning.

 If your startup fits this description, reach out to us.

And sometimes the straight line is real… but your business still gets kneecapped by a deeper trend hiding underneath.

Two examples from my own recent life:

1) The “slam dunk” bet that wasn’t: cell phone loss insurance

If you were sitting in 1995 watching cellphones go mainstream, the obvious overlay business was cell phone insurance—especially loss. More phones → more losses → bigger market. Easy.

Except that’s not what happened.

As phones became more essential, we started treating them like a vital organ.

The adoption curve was real. Cell phone penetration in the U.S. went from 12% in 1995 to 69% in 2005 and eventually 100% by 2020 (source).

The mistake wasn’t misreading adoption—it was assuming human behavior would stay constant as importance exploded.

“Everyone has a phone” doesn’t mean “everyone loses a phone.”

Usage intensity changed behavior. Americans now check their phones 100+ times per day. You don’t casually lose something you touch a hundred times daily. Loss became rarer than 1990s intuition suggested, and insurance demand shifted toward damage and theft—not “oops, left it in a taxi.”

Today, only about one-third of U.S. smartphone owners buy insurance at all.

Penetration created a market—but behavior capped it.

2) The travel adapter that died quietly

I just traveled internationally, realized mid-flight I’d forgotten my plug adapter, and braced for the usual airport-gift-shop penance.

And then… nothing. I never needed one.

The obvious bet was straightforward: more global travel plus more electronics should mean more travel adapters.

The miss was assuming the accessory would survive the standardization.

USB-C won. Standardization crept upward—from devices, to chargers, to outlets themselves—and quietly made the adapter irrelevant. Every hotel room had a USB-C port in the wall. The category didn’t stall. It got erased.

In the phone insurance example, importance changed behavior. Here, standardization deleted the middleman.

The investing lesson: bet on what’s underneath, not what sits on top

When a trend looks obvious, your job isn’t to cheer. It’s to ask:

“What adjacent trend might make the obvious overlay unnecessary?”

Penetration creates opportunity—until importance changes behavior, or standardization removes the need entirely.

The future is rarely a straight line.
It’s a straight line—until the ground shifts and the overlay disappears.

Ubiquity Ventures — led by Sunil Nagaraj — is a seed-stage venture capital firm focused on startups solving real-world physical problems with "software beyond the screen", often using smart hardware or machine learning.

 If your startup fits this description, reach out to us.

This playbook has become known to me over my 15 years as VC investor, many of those inside a large VC firm. This is not a hit piece. It’s a love letter—with footnotes—to entrepreneurs, written by someone who’s watched the machinery up close.

This post is not actually about how to take advantage of entrepreneurs. It’s about how the system quietly does it—and how well-meaning people sometimes participate without even realizing it.

Let’s begin the playbook:

1. Make Everyone a “Partner” (Except in the Ways That Matter)

If you really want to confuse entrepreneurs, teach them a faulty lesson early: “Only talk to partners.”

This advice gets passed around accelerators, founder Slack groups, and warm-intro folklore like it’s gospel. And it’s… wrong.

Any good salesperson knows to map the decision-making unit. Who can be a champion? Who can guide you? Who understands the internal dynamics? In venture, that person is very often an associate or senior associate.

Speaking from experience: I spent several years as an associate at a large VC firm, working alongside more senior folks like Principals, Partners, and General Partners. Within this larger construct, I helped several first-time entrepreneurs navigate partnership dynamics, timing, objections, and internal politics to secure top-tier funding on their path to massive outcomes. I was useful. This includes deals like Rocket Lab’s 2014 Series B (now a $40 billion public company - see our 2014 investment memo), Auth0’s seed (later a $6.5 billion sale - see our 2014 investment memo) — serving as a sort of “inside man” for Pete at Rocket Lab and Jon/Eugenio at Auth0’s (despite not being a partner at the time).

However, once entrepreneurs are trained to say “I only want to talk to a partner,” the incentives flip.

So VC firms respond rationally: promote everyone to partner—in title only.

Over time, most large VC firms went ahead and re-titled associates and principals as “Partner”. Voilà. Every meeting is now with a “partner.”

The catch?

Partner no longer means decision-maker, shared upside, or the ability to truly sponsor a deal.

Entrepreneurs weren’t actually seeking a title. They were asking for agency but now must work even harder to decipher what is really going inside a large VC firm’s organization.

2. Use Friendly, Minimal Term Sheets (And Decide Later)

The final milestone of an entrepreneur’s fundraising process is receiving an investment offer from a VC in the form of a term sheet. And nothing says “we’re aligned” like a one-page term sheet.

It feels clean. Founder-friendly. Modern. Almost European.

It’s also an excellent way to defer all the decisions that actually matter.

Control terms. Protective provisions. Board dynamics. Information rights. Future financings. All of that gets punted into the definitive documents—written after the founder has signed exclusivity.

Here’s the part most entrepreneurs learn too late: The only consistently binding clause in a term sheet is no-shop.

For ~30 days, the founder halts all conversations with other investors. The VC firm? Free to keep thinking. Or not thinking. Or renegotiating.

If your goal were to take advantage of entrepreneurs, this asymmetry would be a feature, not a bug. Vagueness in a term sheet usually benefits the VC, not the entrepreneur, so it’s prudent for entrepreneurs to hammer out more deal terms and push for a more fleshed out term sheet before signing.

A Kinder Ending (Because That’s the Point)

Most VCs are not trying to take advantage of entrepreneurs. Many are former founders themselves. Many genuinely care.

But systems have gravity. Titles drift. Incentives calcify. And entrepreneurs—especially first-time ones—pay the tuition.

If you’re a founder: talk to the whole firm. Associates included. Ask who can truly champion you. Read past the page count of a term sheet. Understand when exclusivity starts—and what you get in return.

If you’re a VC: be precise with titles, clear with intent, and honest about power. Kindness scales better than cleverness.

I’ve written before about what you’ll never hear from a solo GP, and why we chose this high-alignment model for Ubiquity Ventures. This post is the companion piece—the one that says the quiet parts out loud, with respect.

Because the best way not to take advantage of entrepreneurs is to first understand how it happens.

Ubiquity Ventures — led by Sunil Nagaraj — is a seed-stage venture capital firm focused on startups solving real-world physical problems with "software beyond the screen", often using smart hardware or machine learning.

 If your startup fits this description, reach out to us.

It’s like a medium-sized city relocated to 35,000 feet and collectively asked, “Would you like ginger ale or sparkling water?”

That’s just the start of the wild geography of where humans actually are.

The Farthest We’ve Ever Been

In 1968, Apollo 8 became the first mission to orbit the Moon, and in doing so astronauts Frank Borman, Jim Lovell, and Bill Anders set humanity’s initial record for farthest distance from Earth. That mark stood until the ill-fated Apollo 13. When an oxygen tank exploded en route, the mission flipped from exploration to survival — and in the course of swinging around the Moon on a free-return trajectory, the crew accidentally set the all-time distance record: 248,655 miles from Earth. No one has gone farther since, despite reusable rockets and billionaire space races.

There’s another, very different record: being farthest from everyone else. In 1969, during Apollo 11, Michael Collins orbited the Moon alone in the Command Module while Neil Armstrong and Buzz Aldrin made history on the surface. Each time Collins slipped behind the Moon, radio contact vanished entirely.

For roughly 48 minutes per orbit, Michael Collins became:

👉 The most physically isolated human in history.

Every other member of our species was on the far side of a 2,000-mile-wide rock.

The Deepest We’ve Ever Been

Going the opposite direction, filmmaker and submarine nerd James Cameron piloted the Deepsea Challenger almost 36,000 feet down into Challenger Deep — the deepest known point in the ocean and part of the Mariana Trench. That’s seven miles straight down, with pressure so intense it could pancake a car like a tortilla. It works out to 16,000 psi (pounds per square inch), or 1,100 times our normal atmospheric pressure of 14.7 psi. Check out this UCTV talk with James Cameron.

He stayed on the bottom of the Mariana Trench by himself exploring for 3 hours with his custom-designed vehicle withstanding all that pressure (as did an experimental Rolex watch attached on the outside that inspired the Rolex Deepsea Challenge watch).

A Fun Reality Check

If you plot human presence vertically, the picture looks sci-fi:

• +35,000 ft: ~1,000,000 people airborne

• +250 miles: A handful of astronauts on the ISS

• +240,000 miles: Apollo 13’s all-time distance record

• –36,000 ft: Challenger Deep’s tiny visitor log

We casually occupy a vertical spread of nearly half a million miles, and most of us barely think about it. Just a delightful reminder that the real world is already pretty weird — and occasionally, beautifully science-fictional.

Ubiquity Ventures — led by Sunil Nagaraj — is a seed-stage venture capital firm focused on startups solving real-world physical problems with "software beyond the screen", often using smart hardware or machine learning.

 If your startup fits this description, reach out to us.

And then something changes.

As conductor Benjamin Zander explains in one of my all-time favorite TED talks, children slowly stop living note-to-note. They learn the phrase. Then the line. Then the emotional arc. Eventually, they’re so inside the music that their body can’t help but sway. They’re lifted—literally—onto one buttock at a time.

This is the transition from competence to flow.

From “I know the notes” to “I am the music.”

It’s the moment expertise becomes joy.

Where this shows up outside classical music

I’ve seen this in startup founders for years. At first, an entrepreneur is a two-buttock operator: gripping the keyboard, managing tasks, sweating every impulse. Necessary, noble, and usually awkward.

But given enough time—and enough honest practice—they relax into the craft. They start hearing the long line of the company. They shift weight. Decisions flow through them instead of being forced by them.

You can see it in engineering, too. And design. And policy. And deep-tech research. And honestly, in anything people do with enough repetition to move from conscious control to embodied instinct.

It’s the same idea behind our past UBQT post on 1000 hours of deliberate practice. Practice isn’t about polishing each note. It’s about reducing impulses until the phrase plays you.

A nerd’s version of enlightenment

When you meet a person operating in one-buttock mode, you feel it. Their enthusiasm leaks. They’re contagious. They’re unmistakably alive in what they’re doing—and very hard to compete with.

Zander captures this perfectly near the end of his talk:

“I realized my job as a conductor was to awaken possibility in other people… How do you know if you’re doing it? You look at their eyes. If their eyes are shining, you know you’re doing it.”

And his definition of success:

“It’s not about wealth and fame and power. It’s about how many shining eyes I have around me.”

To me, shining eyes = people deep in their craft, working from passion, not posture. People playing on one buttock. Nerds living in their obsession.

Build a life full of one-buttock players

Being around these people—founders, scientists, engineers, artists—changes the way you operate. They’re expansive. They pull you forward. They’re immune to cynicism because they’re too busy building things that matter.

The Ubiquity ecosystem is full of them: the nerds whose eyes light up when they talk about lidar waveforms, autonomous cargo drones, microbes as data sensors, FPGAs as inference engines. They shift in their seats when describing the long line of their company. They don’t play each note—they play the phrase.

That’s the energy I chase as a VC. It’s the culture I try to build. And it’s the kind of expertise worth celebrating for its own sake.

Here’s the simple formula:

• Two-buttock playing: mastery of the mechanics

• One-buttock playing: mastery of the meaning

And if you’re lucky enough, over a long enough line, you end up surrounded by shining eyes—people animated by their craft, by possibility, by music only they can hear.

That’s the whole game.

Ubiquity Ventures — led by Sunil Nagaraj — is a seed-stage venture capital firm focused on startups solving real-world physical problems with "software beyond the screen", often using smart hardware or machine learning.

 If your startup fits this description, reach out to us.

Unsurprisingly, pre-seed and seed founders are now asking me if they’re “already behind.” They’re not. The framing is. It makes for great VC content marketing and punchy headlines—big numbers, big momentum, big clicks—but it also has real downstream consequences. Early-stage founders absorb these benchmarks as gospel, reshape their expectations around them, and quietly panic when they don’t see a path to $40M ARR twelve months after starting their company. That’s the part nobody writing these reports ever seems to wrestle with.

The misleading bit

“Year one of commercialization” is a clever escape hatch. It lets any founder or VC pick when the clock starts, so you can be three years post-incorporation, still developing product, declare “year one commercialization” when you finally sell something—and then you’re measured. In short: the benchmark is malleable.

Despite this, the report lays out “top takeaways for founders” that the 2 winning patterns are either “hit ~$100M ARR in 1.5 years” or “$3M to $100M over 4 years”, implying that seed-stage founders must aim for those numbers or pack up their things and go home. That subtly scares entrepreneurs into believing: fall short = failure.

The real impact

For seed founders—especially in the “software beyond the screen” category that we love at Ubiquity—this framing is demoralizing and counterproductive. Many of the most durable deeptech and real-world AI companies succeed by doing the opposite of sprinting: they nail customer development, build patiently, test deeply, integrate carefully, and then scale fast once the foundation is solid.

And this approach works. Halter (cows), Loft Orbital (satellites), Esper (devices), BusRight (school buses), Eyebot (optical prescriptions) and others in the Ubiquity portfolio are scaling beautifully—but only because they first invested in unglamorous and patient work.

The early years of these Ubiquity companies didn’t look like “Supernova” charts. They looked like thoughtful, disciplined company-building.

Chasing someone else’s clock leads to forced scaling, thin margins, and artificial pressure—exactly the opposite of building a durable, defensible business.

The Ubiquity view

• Don’t panic. You do not need $100M ARR within 4 years of incorporation to be “on track.”

• Define your own start line. Be explicit about when commercialization actually begins—not when a benchmark report suggests it does.

• Prioritize durability. Real businesses have happy, engaged customers. This customer love shows up in retention, expanding usage, and margin improvement. Customer love can also mean a customer tattooing your company’s logo on their leg as a BusRight customer did.

• Use catchy VC benchmarks with caution. They’re outer bounds, not moral imperatives.

Final word: “Don’t compare your beginning to someone else’s middle.”

Founders: Most “year zero” stories come from companies that have quietly been building for years. If you’re newly incorporated, comparing your literal Day 1 to someone else’s multi-year “year one of commercialization” is not just unfair—it’s delusional. Don’t let glossy charts dictate your self-worth.

Pace yourself, build something defensible, and let your velocity be earned rather than forced.

Ubiquity Ventures — led by Sunil Nagaraj — is a seed-stage venture capital firm focused on startups solving real-world physical problems with "software beyond the screen", often using smart hardware or machine learning.

 If your startup fits this description, reach out to us.

Enter: the Tesla Cybertruck.

Elon Musk had already learned this lesson once. The falcon-wing doors on the Model X SUV? They were tech theater — clever, expensive, and headache-inducing. Musk later admitted:

“I do think there is some hubris there with the X… I’m not sure Tesla would make a car like this again.”

And then, somehow, the Cybertruck arrived to say: hold my stainless steel.

⚙️ Over-Engineered Ego

The Cybertruck wasn’t just unconventional — it felt like Musk dared the automotive industry to laugh at him. It was his billion-dollar passion project:

“To be frank, there is always some chance that Cybertruck will flop, because it is so unlike anything else. I don’t care,” Musk tweeted. “I love it so much even if others don’t.”

Mission accomplished.

• Stainless steel exoskeleton → almost impossible to form, costly to weld, questionable crash performance

• Sharper edges than a Minecraft sword → pedestrian safety nightmare

• A truck bed that’s smaller and harder to access than a crossover SUV

• Promised base price: $39,900 (never materialized)

• Top-tier trim: quietly scrubbed from Tesla’s website

Reservations flooded in — over 1 million people sent Tesla $100 for their place in line. But a deposit is enthusiasm, not demand. When production began, the conversion rate evaporated, with only 2.5% of reservation holders actually making the purchase.

🎤 The Night the Kool-Aid Overflowed

November 2019. Musk’s “Cybertruck Reveal” in Hawthorne, California — streamed live to hundreds of thousands of people. He boasted the truck had:

“Ultra-hard 30X cold-rolled stainless steel… you want your truck to be tough.”

Then:

“Armor glass. Bulletproof.”

We all know what happened next. Tesla Chief Designer Franz von Holzhausen winds up with a solid steel ball. Thud. Spiderweb of cracks. Gasps. Musk forcing a laugh:

“Well… maybe that was a little too hard.”

For reasons known only to the 100%-ego-side of his brain, they tried again on the rear window. Crack. Again.

Musk, visibly shaken but still clinging to the bit:

“Room for improvement.”

Iconic technology launch → accidental slapstick. It’s worth watching the 60-second video of this moment to fully understand the cringe factor.

Even the memes recognized the moment: a truck designed to survive bullets couldn’t survive its own marketing pitch.

Beta products can still have kinks but even after delivery of Cybertrucks to paying customers, very serious issues emerged in the field:

• a stuck accelerator pedal (because Tesla decided last minute to use soap to slip the pedal cover on in the factory)

• body panels falling off the truck (because they used the wrong adhesive on stainless steel)

🧠 The Real Lesson for Founders

This isn’t about dunking on Elon (credit where due: Tesla reinvented the car industry). It’s about respecting the physics of markets:

• Break rules to unlock value — not to unlock applause
  Novelty is only healthy when tethered to usefulness.

• Customer > Cool
  A truck that can’t do truck things is modern art, not a product.

• Ego compounds engineering risk

  Every “just for fun” detail is a tax on manufacturing and on user patience.

The Cybertruck isn’t a failure — but it is a cautionary tale:

Disruption for the sake of disruption is just rebellion cosplay.

At Ubiquity Ventures, we back founders with bold visions — but also bold customer humility. The future should be innovative and wonderful… but it must also be wanted.

If you liked this post, read more about this topic in our past blog post: “Look What I Can Do” vs. “Look Who Cares”: Apple misses the mark with its new “Thinpossible” iPad Pro

Ubiquity Ventures — led by Sunil Nagaraj — is a seed-stage venture capital firm focused on startups solving real-world physical problems with "software beyond the screen", often using smart hardware or machine learning.

 If your startup fits this description, reach out to us.