Two years in the making, thousands of hours of research, hundreds of conversations, three startups, and two investing jobs later, here is my Unified Theory of Deep Tech 🧵
- How to invest in deep tech?
- Why now?
- Interesting markets?
- Common pitfalls?
I've now researched all
@Lux_Capital
portcos, here's part 2 of top 5 highest "Wait What Scores":
1.
@relativityspace
: 3D printed rockets
Relativity's Terran R is a fully reusable LEO rocket 3D printed in its Factory of the Future with the largest 3D metal printer in the world.
Is no one else freaking out about this 🤯
DoD’s new Office of Strategic Capital that has a Small Business Administration which is matching 2:1 private LP commitments towards deep tech funds
I've now researched 100
@Lux_Capital
portcos, here are my top 5 highest "Wait What Scores":
1. CTRL-Labs: non-invasive BMI for mind control
With a wrist-wore device, neural signals are picked up and processed allowing users to control devices with just thoughts.
My personal investing thesis: Preemptive Infra.
10 areas in AI I'll be watching in 2023:
1. QA/fact checking services for LLMs.
- Tech trend: all-interfaces-conversational
- Emerging problem: misinformation.
@rabois
should you delegate or overrule when:
1. High consequence decision + low conviction in own judgement + high task-specific maturity of employee in question
2. High consequence decision + low conviction in own judgement + low task-specific maturity of employee in question
A key skill I learned in the past month has been building the muscle to recognize that some 100x technologies are really marginal improvements.
Many breakthrough sciences outperform benchmarks by orders of magnitude, but the advantage gets squeezed when they become startups.
The most under appreciated shocking trend in venture rn is the % of investors who only do SaaS. It’s like being an investor in Q4 1991 and still only doing semi/bio deals and refusing to touch internet infra
In the last months I started internalizing TAM on decks as competition adjusted. Ie $10B market in obscure manufacturing niche w/ 2 existing players where tech breakthrough req’d to enter is bigger TAM than $100B devtools market w/ 250 existing solutions where any SWE can enter
I'm still surprised by how often I get sent decks from companies going into tiny but highly competitive markets.
Oftentimes they're seduced by a sexy idea or problem, such as monitoring space debris or building a niche developer tool.
Subletting my nyc apartment 7/22-8/11, flatiron area (29th & 6th), fully furnished bedroom and master bath with one roommate and one cat. Lots of books
Plz DM!
Some timeless categories of investments that I've seen a lot from studying venture history:
1. Do things faster --> use software/automation to replace or improve traditionally manual tasks
(Stripe, Figma, Zapier, Formic, RunwayML, etc)
Learned today that li-on batteries weren't developed until the 1990s, by Sony! <15 years later Eberhart introduced the Tesla Roadster. Really puts into perspective that we're still in the early innings of battery tech.
This morning learned about:
Intelligent robotic arms
Quantum AI
Diffusion for protein design
From experts in each field. Pretty great to be working at
@Lux_Capital
I've now researched all
@Lux_Capital
portcos, here's part 2 of top 5 highest "Wait What Scores":
1.
@relativityspace
: 3D printed rockets
Relativity's Terran R is a fully reusable LEO rocket 3D printed in its Factory of the Future with the largest 3D metal printer in the world.
I've now researched 100
@Lux_Capital
portcos, here are my top 5 highest "Wait What Scores":
1. CTRL-Labs: non-invasive BMI for mind control
With a wrist-wore device, neural signals are picked up and processed allowing users to control devices with just thoughts.
We’re tackling noninvasive mind reading
@The_Mind_Co
. Please dm me if you’re:
- a swe/ml eng/ee/meche
- in broader bay area
- native english speaker
Looking for summer engineers and participants for data collection trials!
Sharing some WIP market research that
@BrianMikeWu
and I have been working on:
🤖 Robotics is benefitting from decreasing hardware costs + increasing software capabilities --> many application opportunities in agriculture, consumer, industrial etc
Since I'm investing in deep tech & critical national sectors via our fund
@humbavc
, I wanted to better understand if deep tech is VC-fundable.
So, I did what any engineer would: I analyzed data!
TLDR: deep tech is probably the best place to invest today.
Can attest that
@DStrachman
is one of the best resources for young founders and a great human!
When young founders ask me for inception/preseed investor recs Danielle and
@cory
are the two I bring up most often
Random shoutout to
@1517fund
! (🫡
@DStrachman
)
1517 is the most consistent name brought up amongst young, technical, super ambitious founders with just ideas to make a dent on the universe!
And thus we keep finding ourselves co-investing 🤝
Great talk from our
@wolfejosh
at
#StanfordETL
today about why most entrepreneurs are risk killers, why we like to invest in things where there’s time arbitrage & why an entrepreneur should be comfortable being a contrarian
Looking to crowdsource ideas.
What can we do if we have embeddings of all scientific literature? Maybe ask gpt to cross-pollinate ideas from disjoint fields? Resurface old research that's relevant again? Algorithmically discover promising drug candidates in biopharma literature?
Two years ago I built my first BCI app with Muse. It was a $250 4-channel headset.
Today,
@rayhanfahmed
and
@The_Mind_Co
team are excited to launch Hans, a $499 16-channel EEG headset with comparable performance to OpenBCI's $3699 16-channel headset.
Yesterday was the 100th anniversary of the first human EEG recording by Hans Berger.
Today we’re excited to introduce Hans🧠, the world’s first affordable, research-grade, non-invasive mind reading EEG headset.
Hans can be used for a variety of EEG tasks, including
Historic deficit + QT = more bonds for private investors
Inverted yield curve = less bond demand from fund managers
+ more market volatility => more bond demand from risk-averse LPs => need to sell liquid assets (equities) => denominator effect => less VC allocations?
@wolfejosh
4/ a REALIST view of the FUTURE is that scientists, engineers and founders
will shape it into a MORE desirable place we all wish to live in––
—which will also be far LESS rosy than techno-optimist utopianist hallucinations
Focus areas:
—> Biotech, quantum, telecom, materials
—> AI, network systems, microelectronics, space, renewables, computing, human machine interfaces
—> Directed energy, hypersonics, sensors
As a young investor it’s never been clearer why DEEP TECH is the next gen fund returner
I don’t think there’s one optimized way to spend time as a VC, just leave enough room for serendipity and see what happens
Prepared mind + curated serendipity is the job
Trying something out with
@fattah_zaid
We want to highlight some AI deals that happened in the past week, so founders and investors know the market price and what others are building/investing in.
Let us know if you're interested in reading this weekly
👇🧵
Releasing "Paradise Portfolio" 🪽
Finally got a chance to break down my definition of "Deep Tech". The most common one is "taking majority tech risk", but I argue it's much more.
Deep Tech = Tech-Centric Frontier Markets. How they perform financially👇
Starting to realize that as a young person learning is always the top priority.
Building is great but only if it helps facilitate learning, and learning breaks the barriers on what can be built.
What are good example of GREAT venture-backed companies since 2000 that aren't significantly software-enabled? I know there are many that aren't venture-backed, but just looking for VC-backed examples.
1/ Experienced founders have a competitive edge in building companies in their industries, but what people don’t realize is that many of the greatest companies were founded by OUTSIDERS – no experience required.
Here’s 5 examples and the method they use to succeed: 🧵
Really enjoy truly unique and deeply thought-through fund theses. Like an intellectual breath of fresh air
@Lux_Capital
@bedrock
@a16z
@ThriveCapital
all have very different but extremely clever approaches
Join MindCo!
We're a team of 8 working out of a 3-acre ranch in the mountains and provide food and living. Come work with us if you want to GSD across ML, SWE, EE, ME, data collection, and growth/ops.
Please don't apply if you want work-life balance.
Ambient deal flow:
Steve Jang met Brian Armstrong on a plane
Ben Sun met Coupang founder playing basketball
Delian sourced Ramp cofounder playing video games
Big reason Kleiner and Perkins succeeded was they pioneered the modern fund structure, which they were able to do because they were LPs themselves! After the Fairchild stint they became LPs in Rock & Davis and felt the “dead cows” pain points. Aka know your customers
Main takeaways from reading financial/investing history:
1. No strategy is evergreen
2. All great investors became great by outperforming in down markets
- 2a. corr: relative > absolute performance
- 2b. corr: consistency = longevity
3. The whole industry is spanking new
Saw Oppenheimer again last night with the
@Mach_Industries
team
Takeaway is the power of romanticizing your craft. Nolan made physicists more than just scientists and I think every ambitious person has to romanticize the titans of their industry in the same way.