A16Z Crypto Research
By Joe Sweeny
In my last newsletter, I asked “If A16Z is determined to become the full-stack player in venture, what will they do next?” Then they launched…
A16Z Crypto Research
Introducing A16Z Crypto Research:
- A new R&D division pursuing foundational research into the hardest problems in crypto
- Led by Stanford computer science professors Tim Roughgarden and Dan Boneh
- Shipping production-grade code for blockchain developers and the crypto community
Why It Matters
Crypto runs on hype cycles. The 18 months from mid-2020 to 2022 saw an explosion in crypto markets. Ethereum prices soared from $142 in March 2020 to $4,627 in November 2021. Hundreds of billions of dollars were invested. NFTs happened.
But during that period crypto vacuumed something even more important than capital: Talent. Software engineers, product managers, and operators of all stripes left their big tech jobs to migrate en mass to the next tech wave: Web3.
Today, ETH is hovering around $3,000 and it feels like we’re in a bit of a lull. But even as the hype subsides, the talent is still there, hammering away on new startups, infrastructure, and products that will fuel the next crypto boom – if it arrives.
A16Z Crypto Research is a way to cement Andreessen Horowitz’s position at the center of the industry, and bet on the next crypto wave. It leverages their deep bench of expertise and massive capital reserves in a mission to push the entire industry forward — and differentiate them from other crypto VCs.
It’s easy to forget but when Andreessen Horowitz first raised the first $350M A16Z Crypto fund back in 2018 it was a non-obvious bet.
Today A16Z Crypto looks like this:
- $2.2B Fund III and currently raising a $4.5B Fund 4
- 11 investors and a 50+ people supporting startups they invest in
- Regulatory, Engineering, Marketing, Recruiting, and Editorial teams
But back in 2018 the market had just crashed and a multi-year crypto winter was beginning.
Raising that initial A16Z crypto fund enabled them to invest in companies like Uniswap, Solana, dydx, Dapper Labs, and Compound. A16Z Crypto Research is a similar bet today.
We have an “all weather” fund*.** We plan to invest consistently over time, regardless of market conditions. If there is another “crypto winter,” we’ll keep investing aggressively.*
– Chris Dixon, Head of A16Z Crypto announcing A16Z Crypto Fund I in 2018
Open AI’s DALL·E 2
- Image generating AI software from Open AI
- Creates photo-realistic images from a written caption
- Second generation, named after the artist Salvador Dalí and Pixar’s WALL-E
DALL·E 2 launched earlier this month but we haven’t talked about it yet on this newsletter so I’m not going to miss the opportunity.
It’s an AI that can create incredibly realistic images from a written prompt. It spans styles, from photorealistic to ‘vaporwave’ to the work of specific artists. Like Open AI’s text generative algo, GPT-3, it’s been trained on virtually the entire internet and has near limitless potential.
It can also add new images to existing photos. Like these flamingos:
And like all great creative tools some of the best work has been created by the community. Just search DALL·E 2 and you’ll find thousands of pieces of content:
Not Fun Fact: It’s also highly biased. When you ask it for a nurse or a personal assistant it generates only women. When you ask it for a lawyer or a CEO only men. Open AI is acutely aware of this fact and gives a full rundown in their Risk and Limitations section. I’d argue that identifying these biases in the research is more instructive than hiding them but the docs state:
DALL·E 2 tends to serve completions that suggest stereotypes, including race and gender stereotypes.
Those are just the most obvious ones. When you consider the long tail impact these technologies could have… More than a bit scary.
Today, we’re flooded with content but all of it is created by other people. In the near future, a lot of the content you consume will be generated by a computer.
Want to read a new book? Just ask your AI and it’ll generate an entire novel. In the style of your favorite author. Based on a plot you give it.
A new Netflix series? A new video game? AI can generate that. With photo-realistic graphics and sound to match.
How do we prepare for that?
For more on this tech, I interviewed Cristobal Valenzuela, the founder of Runway, an AI company making it easy for anyone to create movie-grade CGI and special effects. Listen here.
- Brex acquired Pry, a financial planning software tool for startups for $90M
- Pry had raised $4M to date for financial planning software for startups
- Pry was growing 30% MoM since launch in March 2021
- The Playbook for up and coming fintech startups
Normally I wouldn’t cover a single acquisition, but Brex’s evolution is a case study for fintech. If you understand the Brex story you can predict where dozens of other neobanks will go next.
Brex started with a simple message: Credit Cards for Startups. High limits, lots of points. $5,000 in AWS credits for new customers. 6X Points on lab equipment. 4x points on software purchases.
How Credit Cards Work
For those of you who aren’t credit card obsessed, ‘points’ are just a cash back mechanism. When you swipe your card, the credit card issuer charges the business you’re buying from 1.5% to 3.5%. The interchange fee. Visa or Mastercard take a percentage of that fee and then the credit card provider, Brex in this case, but it could be Chase or Apple or Macys, take the rest. Then to attract customers the provider, Brex, gives back a percentage of every transaction.
Fun Fact: Points don’t exist in Europe because regulation caps interchange fees at around 1% so card issuers don’t have the margin to give rewards to European customers. Also, I kind of can’t believe credit cards were ever adopted. Businesses went from accepting cash to accepting credit, having to wait 30 days for payment, and getting charged a 3% fee. How did that happen?
Cards → Bank → Software
Compared to the rest of finance, it is incredibly easy to start a credit card issuing business. Today you can even get a white label version from Stripe. But as you may have noticed from the interchange breakdown, there isn’t a lot of margin to build a durable business on top of.
So Brex did what you can expect from every new startup that launches with a credit card product. They built bank accounts. Now they’re working to expand into the highest margin business of all: SaaS software.
Evolution of a Fintech
- Business Credit Cards: Your Wedge into the Market
- Next is Bank Accounts, Expense Management, Bill Pay, Payroll
- Last is Software, usually financial modeling/planning software
Today Brex aims to be the financial operating system for startups. A diversified set of products with a diversified set of revenue streams. Banking, financial services, and software. Watch out for this same playbook from up and coming credit card startups like Ramp and Divvy.
Twitter Sells to Elon Musk
As I said Friday, the only question for the board was whether they’d be able to drive the price over $52.40 over the next 6 to 18 months.
Looks like they decided it was better to cave than risk lawsuits from shareholders if they turned Elon down and the stock plunged after he sold his 9% stake. Will be interesting to see what changes, if any, he makes to the product.
Despite the speculation, no one but Elon knows what he plans to do with Twitter:
FROM THE PODCAST
Every week I interview up and coming founders to break down their business, how it fits into the market, and get a glimpse of what’s coming next in tech. Here’s a quick take from the Just Raised pod this week:
Blue Origin with Rob Meyerson
A new economy is being built in space. Bootstrapped by entrepreneurs and startups far more concerned with operating costs and profits than government space missions ever were, it’s not about exploration anymore its about business.
Former President of Blue Origin and current VC at C5 Capital, Rob Meyerson joins the show to map out the next 10 to 50 years in space.
Products: Private space stations, transit networks to the moon, 3D printing, and lunar colonies
Traction: VC Investment into space startups hit ~$14.5B in 2021, up 50% over 2020
I think the next big leverage point in lowering the cost of space missions is utilization of space resources.
If it costs X to launch a satellite, what if you didn’t have to launch that satellite at all? What if you 3D printed it in space and assembled it and deployed it from the space station? What if you 3D printed it with resources that you sourced from the moon or an asteroid?
– Rob Meyerson, Former President of Blue Origin and VC at C5 Capital
- Raise: $38M Series A from Andreessen Horowitz
- Including: A killer set of angels including
- Plus: $5M from their user community crowdfunded on WeFunder
- One Liner: Glucose monitoring to track your metabolic health
I started my career covering wearable startups for Lux Research and I am very bearish on wearables. Step tracking and heart rate is just not valuable data for the average consumer. The collapse of Fitbit’s share price is proof.
Levels is different. Minimally invasive blood glucose monitoring has long been the holy grail of wearables. Their monitor allows you to tap the blood stream directly to access data foundational to diseases form diabetes to obesity and heart disease. I’m a Level’s user and a fan.
We dive deep into the tech and the business strategy on the podcast. An interview with Levels Founder and CEO Sam Corcos. Listen Here.
Also: Oura Health on the other hand just announced they’re bringing on new CEO to bring the company public. At some point every company needs to exit but I am not optimistic…
- Raise: $23M from Index
- Including: GV (formerly known as Google Ventures), Figma founder Dylan Fields, and David Nothacker founder of Sennder
- One Liner: Seafood Trading Software + Marketplace
Fishing is a massive industry with hundreds of thousands of unique players. Boats and fisheries, processing facilities, wholesalers, distributors, restaurants, and fishmongers all play a role before seafood reaches consumers.
Add to that hundreds of different fish all in different sizes you end up with thousands of skews of a very perishable product. It’s a very messy industry with massive amounts of waste.
Keith Rabois has a recipe for building a killer business:
Find large highly fragmented industry with low NPS; vertically integrate a solution to simplify value product.
Keith Rabois, Founders Fund
Rooser is a clear cut marketplace business and an obvious example of a fantastic niche, vertical software businesses.
- Raise: $10M from NVIDIA
- One Liner: Self-driving sidewalk delivery robots
- Why It Matters: NVIDIA needs autonomous driving to materialize
Nvidia is an incredible cash flowing silicon chip business. It’s the 8th largest market cap company in the world at a monster **$474B and it’s growing by ~50% a year.
But it has only $12.5B in revenue.
To justify that valuation that need massive future markets to materialize, including autonomous vehicles – self-driving cars. So a $10M investment and ‘strategic tech partnership’ with Serve Robotics makes sense if it can improve their self-driving AI to even the smallest degree.
That said I do not expect sidewalk delivery robots rolling around major cities any time soon.
Deep Dive: Acquired just did a fantastic 2 part breakdown of NVIDIA which I highly recommend.
Also: NVIDIA isn’t the only one in the hunt for autonomous driving. In addition to American companies like Tesla and Waymo, Chinese firm Pony AI just announced they received a license to operate self-driving taxis in Guangzhou. Given China’s centralized leadership and push into EVs I wouldn’t be surprised if self-driving cars and taxis rolled out across China in the next few years, well before the US.