Bruce Sinclair

3 red flags that are killing your fundraise according to a $100M VC fund. Warning signs that Juan Carlos Pacheco and IDEA Fund Partners screen for in the first conversation: • Arrogance (kills partnership) • Weak command of numbers (signals chaos) • Saying "no competition" (shows market naivety) Here’s how to avoid them to start showing up as a credible, coachable operator that investors actually want to partner with. #1 Arrogance Juan Carlos said it plainly: "If you're a jerk, that's a very quick red flag. Because it's a partnership when you invest." This isn't about being nice for the sake of being nice. It's about what arrogance signals. If you dismiss questions, you'll dismiss your customers If you can't take feedback in a pitch, you won't take it from your board. .  If you think you know everything, you won't adapt when the market shifts. Investors are buying into years of partnership. Board meetings. Hard conversations. Pivots. Bad quarters. If you're arrogant now, you'll be impossible later. #2 Weak command of your numbers Juan Carlos said: "If you can't speak with some intelligence about revenue, burn rate, how you're gonna use the capital... that's a huge red flag." Weak grasp of numbers signal operational chaos. Vague on how you'll use capital? ↳ You'll waste it. Don't know your burn rate? ↳ You're guessing at runway. Can't explain how you arrived at your valuation? ↳ You don't understand your business model. Investors aren't expecting perfect projections. They know things will change. But they need to know you're running your business with discipline. That you're making data-driven decisions. That you can be trusted with their capital. If you're fuzzy on the numbers now, you'll be reckless with their money later. #3 "We have no competition" This one kills more deals than founders realize. Juan Carlos said: "Competition is typically a good thing. It shows there's a market there." Here's what "no competition" actually tells an investor: Either you haven't done your homework, or there's no market and customers don't care enough to pay. And if there *is* competition and you just don't see it? You're naive. You're not paying attention. You don't understand your landscape. Juan Carlos used a sports analogy: "If you're going into a tournament, you scout your opponents. You see what they're strong at. You see what you're strong at. And you adapt." Competitors aren't threats. They're validation. They prove there's a market. They show you what works and what doesn't. They force you to get sharper. If you can't articulate your competition and your edge, you're not ready. At the early-stage, there is uncertainty and risk.  Investors don’t expect flawlessness. They do expect coachability, self-awareness, and market intelligence. Know your numbers cold, stay humble, and study your competition. Do that, and you’re already ahead of most pitches.

21

13 Comments

AI success that scales comes from systems design. Specifically it’s how systems learn and how they develop expertise over time. As we’ve built our AI Operating System (AIOS), we’ve made deliberate architectural choices that don’t show up in demos, but compound into learning advantages. It helps us scale over firms that have built individual agents. Three aspects of what Abdul and I found out over the past two years: 1) Memory as the foundation of learning. Most AI systems treat context as disposable. We treat it as cumulative. Our architecture centers on persistent, evolving memory that enables continuous learning across interactions, personalization at the OS level and compounding context that adapts performance over time. Without memory, there is no real learning and the system needs to repeat many steps. 2) Model layers as pathways to expertise. We’ve moved away from single agents on monolithic models toward modular, layered systems: - Routing to specialized models, fine tuned to develop domain-specific expertise - Orchestration layers that route tasks to the right “expert”, human or AI - Continuous evaluation and learning capabilities that upgrade without breaking the system Expertise emerges not from one agent doing everything, but from systems that know which intelligence to apply, when. 3) Multiplayer AI and reinforcement learning from experts. The next frontier is smarter learning loops. We’re building toward a multiplayer paradigm where: - Networks of experts (human + AI) provide reinforcement signals - Systems learn from real-world judgment, not just static data - Expertise is encoded, refined, and scaled across the network Most teams are still optimizing for benchmarks and surface-level capabilities. But durable advantage in AI will come from systems that learn continuously, architectures that compound expertise, and feedback loops that improve with scale The companies that win will build systems that get smarter, more specialized, and more valuable with every interaction.

9

3 Comments

Free markets won't function optimally and will do great harm if costs of negative externalities aren't reflected in the price of products produced, but it's difficult to figure out those costs and create mechanisms that force producers to pay for them. https://lnkd.in/gRDhY3i

11

1 Comment

The current-cycle AI infrastructure boom is unprecedented. The next cycle must be defined by AI creating real economic value. In barely a year, AI infra has become a huge part of US GDP growth: → Hyperscalers racing to build data-center capacity → NVIDIA racks pulling 120kW per unit → Hundreds of billions in CapEx committed over the next 24–36 months Soon, the market will start pricing in this fact: None of this infrastructure pays back unless AI delivers measurable ROI in the real world. Not “assistants” or "co-pilots". Not enterprise tire-kicking. Not flashy demoware that falls short at the hardest 10% of the job. Not productivity theater. Hard, real-world, financial outcomes. And this is why the next decade of AI alpha won’t be found in broad enterprise AI, it'll be in vertical AI with direct cost removal and / or revenue lift. This week on VERTICALS, when we spoke with Mike Kopko (CEO, Pearl Health), the pattern became obvious: → Avoiding a $30k hospitalization = immediate ROI → Aligning with value-based care = massive opportunity + society problem → Government-backed payment models = durable revenue Healthcare AI has potential to do more than "improve workflow." It has promise to make our HC system more efficient, profitable, and effective. That's the kind of economic impact that absorbs AI infra, rather than speculating on it. The 1st Generation of AI companies benefitted from novelty. But now, the question is shifting from: + “What's the biggest TAM we can slap AI onto" to: + "Which real-world workflows have clear AI-addressable cost removal / revenue uplift opportunity?" Healthcare is emerging as one of the strongest answers. 🎧 Full episode with Mike Kopko / Pearl Health here: https://lnkd.in/gMZnVvgq 📖 And our write-up on the AI infra cycle: https://lnkd.in/gPk9wb-q TLDR: Data-center CapEx won't be justified by copilots, it'll be justified by Vertical AI driving tangible ROI in risk, revenue, and cost.

24

3 Comments

Here's a problem we see all the time in early studios: they copy-paste a VC fund's operating cadence and wonder why nothing feels right. Weekly deal flow reviews. Portfolio monitoring calls. Valuation-focused LP reports. All borrowed wholesale from a fund playbook. But a VC fund's primary activity is deal selection -- source, evaluate, check, board, wait. A studio's primary activity is venture creation -- generate, validate, recruit, build, launch. Those are fundamentally different jobs. When you run a studio like a fund, things break in specific ways: Your meetings are wrong. You're reviewing deal flow instead of venture build progress -- milestones, blockers, resource allocation. The work is building, not sourcing. Your team is wrong. Studios need builders at every level -- operators, designers, engineers. Not just investment professionals. You can't build companies with a team designed to pick them. 📉 Your metrics are wrong. Entry valuations and markups don't tell you anything useful. Cost-to-build-a-venture, time-to-market, and venture survival at 18 months -- those are the numbers that actually measure whether the studio is working. Your LP reporting is wrong. Report operational progress -- what's building, what's launching, what's been killed and why. Not just portfolio movements. Here's the lesson -- the studios that struggle most are the ones that adopted VC templates without modifying them for fundamentally different work. The portfolio companies don't exist until the studio creates them. Not a fund that happens to build things. A builder that happens to have a fund. The operating rhythm should reflect which one is primary. 🎯

10

1 Comment

Something I’m pondering. We are becoming an AI driven economy. It’s good because the usage as measured by CPU cycles is growing exponentially and the US leads internationally by a fair margin. However, concentration has its obvious downsides. Data center investments drove 92% of US GDP growth in the first half of 2025, according to analyses from economists at Harvard and Renaissance Macro Research. Without this capital spending, growth would have been nearly flat. From Bloomberg: The amount of debt tied to artificial intelligence has ballooned to $1.2 trillion, making it the largest segment in the investment-grade market, according to JPMorgan Chase & Co. (JPM). AI companies now make up 14% of the high-grade market from 11.5% in 2020, surpassing US banks, the largest sector on the JPMorgan US Liquid Index (JULI) index at 11.7%, JPMorgan analysts including Nathaniel Rosenbaum and Erica Spear wrote in a note Monday.

35

5 Comments

Last week, my partner Rob Go laid out a provocative take on why seed VC is facing an existential crisis. This week, he explores the path forward. There’s no magic "answer"—but Rob lays out a thoughtful framework for how seed investors can adapt in this new era, and why there’s real reason for optimism in this AI Supercycle.

48

4 Comments

The world's largest AI companies are bypassing TSMC and Samsung to build their own semiconductor fabs, not because they want to, but because they have to. Foundry capacity has become the binding constraint on AI deployment at massive scale. While chip design gets the headlines, the real bottleneck is manufacturing capacity—specifically the ability to produce cutting-edge processors in the volumes that AI training and inference require. This happens because existing foundries face hard physical limits on how quickly they can expand production, driven by the complexity of advanced node processes and the 2-3 year lead times for new fab construction. The economics force a choice: wait in line behind other customers for limited foundry slots, or build dedicated manufacturing capacity. Companies pursuing AI at trillion-parameter scale cannot afford to let semiconductor production scheduling determine their deployment timelines. This shift raises fundamental questions about vertical integration in the AI stack. Which companies have the capital and technical capability to build fabs? How does owning production capacity change competitive positioning when training runs require thousands of chips? What happens to companies that remain dependent on third-party foundries when capacity becomes scarce? The analysis examines which AI leaders are positioned to build, buy, or partner their way to manufacturing independence—and which face structural disadvantages in the race for computational scale. Link in comments 👇

1

1 Comment

Data-center infrastructure-equipment sales may rise at a 12% CAGR and jump to $73 billion by 2028, based on BI's update to its market outlook of six months ago and after our February report on AI infrastructure. This report examines the evolving opportunities for industrial companies as rising manufacturing capacity eases constraints and AI-driven investment lifts sales of electricals like Schneider and Vertiv. We also consider how efficiency gains from new large language models like DeepSeek could ultimately boost AI adoption, which means growth in equipment needs will continue.

2

1 Comment

😧It appears the public markets fear of AI has done a 180 - from fear of excessive investment to AI's ability to obliterate long incumbent software companies. I have a feeling we are going to hear the new riff "software is eating software" more times than the famous 2011 op-ed statement from Andreessen that "software is eating the world". Another 180 - long hardware, short (trad) software 🤯. AI infrastructure is literally eating memory. The global memory-ship shortage is expected to last another few years! 💽 Shout out to our pals at Inovia Capital for their State of Canadian Software Report. Thanks for the Miovision mention regarding the massive partnership potential for Canadian companies in the Middle East. 👉 https://lnkd.in/dyB8JkEh

46

4 Comments

Big milestone for Ayar Labs — $500M Series E at a $3.75B valuation. This raise underscores just how critical optical I/O and co-packaged optics (CPO) have become in the AI and HPC era. As compute scales, the bottleneck is no longer just processing — it’s moving massive amounts of data efficiently, with lower power and latency. That’s where Ayar’s photonics platform is positioned to fundamentally reshape system architecture. Notably, this round combines top-tier institutional investors (Neuberger Berman, ARK, Insight, QIA, Sequoia Capital Global Equities, 1789 Capital) alongside strategic leaders like AMD, MediaTek, NVIDIA, and Alchip. That mix of financial and commercial backers signals both capital confidence and ecosystem pull. The focus now: scaling high-volume production and test capacity to accelerate deployment of their CPO solution. Moving from breakthrough technology to manufacturing scale is the real unlock. Congrats to the Ayar Labs team on continuing to push the semiconductor and photonics frontier forward. #optical #photonics #semiconductor #AI #HPC

8

An interesting read on #AImanagement: The Rise of #AI Corporate Citizens: which AI models to follow - what’s at stake - where humans lead: “#Leaders need to invest in operating-model redesigns, build new talent archetypes, and establish trust mechanisms that enable humans and AI to #collaborate safely and effectively at #scale. 👉🏻Humans will be custodians, who ensure the integrity of #data, model performance, and customer outcomes. 👉🏻Judgment holders who handle ambiguous or high-stakes decisions where context, nuance, and trust are essential. 👉🏻Approvers and auditors who review exceptions, manage escalations, and reinforce compliance boundaries. 💡Rethinking decision-making with 🧠 ‘smart ops’ To unlock the full potential of AI in service operations, organizations need to do more than deploy technology. They need to rearchitect how decisions are made and how work is done—by building a “smart ops” structure where humans and AI agents operate in coordinated, complementary roles. 📍#Governance and oversight - Humans operate under #policies and cultural norms. AI agents need the same guardrails: ethical #frameworks, transparency, auditability, and fail-safes for sensitive decisions. Especially in regulated sectors, this isn’t optional—it’s existential.” Source: McKinsey #ai #aimanagement #aiefficiency #llm #agenticai #aimodels #aiframework #infra #aipolicies #government #aisafety #impact #scale #generalpurpose #data #mckinsey

Everyone keeps asking: "Won't better chips solve the power problem?" No. And here's why. Chip efficiency is improving. That's real. NVIDIA's chips improved performance per watt by about 20%. Liquid cooling is reducing energy waste. Better architecture is squeezing more compute out of every kilowatt. But here's what people miss: AI compute demand is growing faster than efficiency improvements. Much faster. I've seen forecasts where AI compute demand doubles every 3 months. And I believe it... Chip efficiency is improving maybe 20-30% per year. Do the math: If demand doubles every 3 months, that's 16x growth in a year. If efficiency improves 30% in a year, you're using 0.7x the power per computation. 16x demand × 0.7x efficiency = 11x more power needed. Efficiency helps. But it doesn't come close to keeping up. This is why Google is betting $40 billion on Texas transmission upgrades. This is why Microsoft is restarting Three Mile Island. This is why Anthropic is spending $50 billion on infrastructure. They've done the math. And the math says: chip efficiency buys you time, but it doesn't solve the problem. The only solution is more power infrastructure. A lot more. So when someone tells you "better chips will reduce power demand," they're technically right. But they're practically wrong. Because the demand curve is steeper than the efficiency curve. And that gap is the entire power infrastructure build-out for the next decade. If you're betting on chip efficiency to slow down power demand, you're going to lose. The hyperscalers aren't betting on it. They're building power infrastructure as fast as they can. That should tell you everything you need to know.

41

25 Comments

Lithography the process of putting an image onto the surface of the silicon. Pretty much like the way a silkscreen puts  “Team Building Exercise 1999” on a T-shirt. Except that this image has to be the highest resolution, with the smallest microscopic features, of anything humans produce. “Moore’s Law” usually refers to increasing transistor density. Basically, how can we make transistors half as big as they were 18 months ago? Every time we figure that out, computers get twice as powerful. The state of the art uses Extreme Ultraviolet (EUV) light to do the lithography. The machine that can do this cost $50 billion to develop. It has 500,000 parts. Only the Large Hadron Collider is more complicated. To buy one costs $250 million and you’ll be stuck on a waiting list that is $40 billion long. The machine comes from ASML in the Netherlands and they don’t have a single competitor, in the entire world. That machine shoots a tiny ball of molten tin into a vacuum and blasts it with two lasers. This produces a flash of 13.5 nanometer ultraviolet light that gets aimed at the surface of a silicon wafer. ASML advanced from 193nm to 13.5nm light to make this possible, but there’s a problem. The diffraction limit of 13.5 nanometer light was set by either God or Issac Newton and there’s nothing we can do about it. We can’t print features smaller than that and there’s no practical way to do lithography with a shorter wavelength. When people say that Moore’s Law is over, this is why. The semiconductor industry knows this, so they’ve tried to solve the problem by handing it over to the marketing department where the laws of physics don’t apply. You’ve seen them progress from 45nm to 30nm to 20nm over the last decade, then all of the sudden, 12nm, 7nm, 5nm & soon 3nm chips are coming. This is all just marketing bullshit. This measurement in chips used to be half the distance between the centers of two features. Once marketing took over, they started measuring half the distance between the edges of two features. Instant improvement! Then they started measuring other random stuff. Other kinds of improvements in chip design helped to gloss over the fact that we are no longer able to shrink the size of transistors. Today, there are extraordinary geopolitical machinations to control chip production. The U.S. has tariffs and export controls akin to those for fighter jets and ICBMs. Access to chip production is as critical to superpowers as oil. Lace Lithography had been in stealth since we invested in them a few years ago. They’ve invented the technology that can go beyond Extreme UV and put Moore’s Law back on track. By using helium atoms instead of light, they can make transistors 10x smaller than the physical limit of ultraviolet light can. https://lnkd.in/dq8ruxHV

98

6 Comments

I’ve been working on the theory of why small, shippable factories are the perfect inversion to the centralized prefab factory model. It all collapsed into one line of math (credit to Gilles Retsin): Factory Efficiency = Usage × (Throughput × Product Value ) / ( CapEx + Deployment OpEx) The idea is that the closer you can drive Usage to 100 % while keeping the denominator tiny, the faster the flywheel spins.  A shippable microfactory tackles this by primarily focusing on a high Usage factor. Because the microfactory can be packed up afterwards and reused, its CapEx gets amortized over multiple projects, effectively raising its overall Usage across the year. If one project alone doesn’t fully occupy the factory, it can simply roll over to another job to maintain high Usage. Its CapEx is often an order of magnitude lower than a centralized factory (< $1M vs. $10-50M min), so the breakeven throughput is more achievable on a small pipeline of work. The real breakthrough here is portability. There is near-zero stranded capital and almost no idle time. To flesh this theory out, I wrote 3000 words on the topic for Brad Hargreaves and Thesis Driven. The article breaks down: - why “fixed factories, shipped goods” is being inverted to shipped factories, fixed goods - what a sub‑$1 M robotic cell does to CapEx per home versus a £45 M off‑site plant - early production/cost data on a shippable microfactory from Mollie Claypool, Gilles Retsin, Sam Baker and the Automated Architecture (AUAR) team - the hybrid reality of robots tackling the heavy structural components and human crews assembling faster and cheaper - technical constraints to pull this model off and future considerations on the optimal business model (own vs. rent) Major shoutout to Gilles Retsin for helping me pull this together. https://lnkd.in/eRTnbZPa

54

2 Comments

While investing in #AI, many corporations fail to achieve strong ROI as a result. The root of this failure is not technological. It is macroeconomic and organizational. US enterprises are attempting to execute a paradigm-shifting technological transition within a market environment that offers them no structural shock absorber, in the way a formal industrial policy can. These enterprises must navigate the AI transition while addressing a rapidly changing market environment and the AI Stakeholder Squeeze. The post builds expands on some of the ideas presented in the recent podcast with Ben Lorica 罗瑞卡 about remaining employable in the AI era (https://lnkd.in/gAGRYxQN) #AIstakeholdersqueeze #AItransition #industrialpolicy

15

2 Comments

100 founders, builders, researchers, investors and thinkers. One big question: how to solve some of the largest challenges building at the frontiers of AI. From physical compute scaling constraints to applying AI to unlock civilizational progress from material science to drug discovery. Camps, hikes and heated debates on human x AI co-existence: this was Climbing Hard AI Peaks. Thanks to everyone for coming out and making it so special. We learned a lot from all of you. Big shout-out to our founders working on some of the biggest AI unlocks, the many researchers and builders from Center for Digital Technology and Management (CDTM), ETH Zürich, Stanford University and special guests from Anthropic, OpenAI, Google DeepMind and more. David Byrd Michael Wax Jason D. Whitmire Chad Fowler Jonah Anders Kaplan Christine Fuchs Sam Harrison Mehdi Ghissassi Antoine Moyroud Wulfie Bain Eva Spannagl Nathan Gruber Judith Dada Juan Benet Zoe Weinberg Haya Hanna Sophia Kalanovska, PhD Clayton Mellina Ricardo Sequerra Amram Evan Phoenix Hendrik Dietz Katie Hewitt Molly Mackinlay Omer Shlomovits Zied Bahrouni

129

16 Comments

You can have groundbreaking science, elegant engineering, and even early customer interest. But without this one thing, you don't have an investable business. After two decades running IP strategy projects and now leading UC Berkeley's Deep Tech Innovation Lab while building the Berkeley Gateway Accelerator, I've watched countless brilliant technologies die in the valley between breakthrough and business. The problem isn't what most technical founders think it is. It's not about having better tech. It's not about getting more funding. It's not even about finding product-market fit. The fatal flaw shows up much earlier—and it's almost always the same mistake. In my latest piece, I answer questions from a Taiwanese entrepreneur about what really separates deep tech ventures that scale from those that stall. Including one provocative suggestion for Asian ecosystems that has nothing to do with technology. Read the full conversation, linked below. #DeepTech #Innovation #Startups #IntellectualProperty #VentureCapital

60

6 Comments