Today is day 1️⃣2️⃣ of my 30 day writing challenge.
Yesterday, I wrote my thoughts on how I think Twitter is going to start changing after the platform launches their initial $5m creator fund.
tl;dr: creators will start optimizing for longer form formal content, Tucker Carlson type news episodes, viral threads, etc. In my opinion, this is the start of Twitter 2.0 - an end of the OG "what's on your mind" tweets and the start of a centralized citizen journalism platform. You can read the full post here, "Picture the NYT and Twitter in a boxing ring".
If you're new to The Bigger Picture (TBP), welcome! Subscribe below so you don't miss any future TBP posts 🥂
In today's post, I try to understand more about what deep tech actually means. I've heard the term tossed around a ton on tech twitter, but still wasn't sure what it really meant. At first, it just felt like a fancy term that VCs used to amaze their LPs and convince them for more capital. To learn more, I read through this 12 page paper by MIT that deep dives into deep tech (yes, pun intended).
At a high level, there are 2 things to help filter whether any given technology falls under the deep tech category:
The existence of a high level of information asymmetry
The presence of a high level of capital intensity
My key takeaway from writing this post is that all great tech companies are downstream of a few academic papers and R&D labs that changed the way humanity progressed.
Defining Deep vs. Shallow
First, let's define the difference between deep and shallow tech.
The term "deep tech" was first popularized in 2014 by Swati Chaturvedi, the founder of Propel(x). After the internet started gaining traction in the mid-90s, a ton of the Silicon Valley venture focus for the following two decades shifted to SaaS and consumer business models. In other words, most VC firms were trying to figure out the next big application on the internet, not looking for the next internet itself. And that was the obvious and correct path to take. Any investor would have been doing a disservice to their LPs by missing out on the likes of Google, Facebook, Uber, Salesforce, Instacart, etc.
However, by the mid 2010s, as the momentum of early web and mobile apps started to slow down, investors started looking for larger and more consequential investments once again. Some shifted their focus to blockchain. Others to AI. And some to space and biotech. To differentiate these industries from the normal SaaS/consumer businesses, the term "deep tech" was coined and encapsulated the idea of investing in companies that were working on fundamental technological advancements. The term gained popularity in the VC community and has been widely used the last decade to define tech sectors such as:
AI & Robotics
Of course, there is no one hard definition of deep tech. Every tech circle has their own meaning and categorization. But for the sake of this post, I'll use everyone's favorite site (Wikipedia 😂):
Deep tech refers to a category of startup companies that develop new products based “on scientific-discovery or meaningful engineering innovation
On the other hand, "shallow tech", a term that not that widely used, often refers to tech companies we commonly think of when hearing Silicon Valley. These are the companies that convert a non-digital business to a digital model. The focus here is ruthlessly iterating on product, beating out competitors, and scaling to billions of people. When you think of the terms "product-market fit" or "lean startup", there's a good chance the discussion is around a shallow tech company. For example, Amazon made books digital. Uber made the Taxi model on an app. And Salesforce made HR databases into CRM software.
To be clear, in my opinion, anyone who says deep tech is any more impactful than shallow tech needs an ego check. The challenges are difficult either way for a founder. One relies on scientific breakthroughs while the other relies global psychological mind shifts. Both are a necessity to push technology forward in society.
Deep tech companies push the boundaries of what is possible, while "shallow" tech companies work on applying those breakthroughs to create practical, accessible solutions for end users.
The 5 Commandments of Deep Tech
In the past decade, as the amount of capital and investors present in the venture space starts to increase, the need for a strict definition of deep tech has only increased. The worry many scientists and engineers have is that the term will be overused and applied broadly thus. losing its utility as a specific category of ventures.
The MIT report mentioned above provides 5 points to help filter what counts as "deep" and what doesn't. Let's dive deeper into each of them.
Positioned at the scientific frontier - this translates to technologies that have long and uncertain R&D cycles. Not a few months or years, but more so on the scale of decades. A good example here is how early AI researchers that pioneered the field before most even knew what AI meant have already passed away before seeing their fruits of labor in apps we see today like chatGPT.
Building tangible, often regulated, products and processes - deep tech typically starts out with zero/unclear processes and regulation. Since no one has ever even imagined the capabilities of new technology, it's often (if not always) challenging for governments to have policy on how the tech should be defined in the law. Technologists have to work in parallel with the government to determine what the new frameworks are. For example, in the last decade, the crypto industry has had a constant fight with authorities on determining whether cryptocurrencies are securities or commodities. No one imagined a distributed ledger until Satoshi solved the byzantine generals problem in 2008. That single whitepaper has led to billions of dollars in funding for clear regulation around the new tech.
Linked to key ecosystem stakeholders, especially Higher Education Institutions - this relates to my key takeaway mentioned in the above section, "most deep tech is downstream of academic papers and R&D labs". The reality is that the next big technological breakthrough comes from some obscure lab in a university with a few scientists pouring years of hard work to make a single breakthrough. However, once the breakthrough is achieved, it can provide a canvas for thousands of developers, engineers, etc. to build the next generation of technology. We're seeing it with OpenAI researchers and the release of ChatGPT today. A decade ago, we saw it with the mobile revolution and app store. And we even saw it with the release of GPS technology (i.e. Uber, Google Maps). There's also a clear correlation with technological breakthroughs primarily for military stakeholders which are eventually iterated on to meet consumer demand.
Mission Driven - this one is straightforward. Deep tech companies don't primarily rely on quarterly goals or are worrying about a 10k report with revenue and MAU metrics. Rather, deep tech leaders are focused on creating an almost cult-like employee base that will die for a given mission. The best example here is the employee base at SpaceX that rallied around their holy leader, Elon Musk. These engineers, arguably some of the most talented in the world, worked themselves to near exhaustion over the past decade to reduce space exploration costs. Just recently, after 20 years, is the company starting to see the fruits pay off. If they would have measured their success on revenue early on, the morale would have been always at an all time low.
Built through a dynamic de-risking cycle - Most "shallow tech" companies can be sure that their overall idea makes sense and if there's a need for it pretty early on. Not that the first idea launched will always work, but both founders and VCs have a general idea if the problem space being worked on is one that is feasible and makes sense. The path is pretty linear: build MVP, iterate, find product market fit, scale regionally, then internationally, etc. On the other hand, deep tech companies are exposed to several cycles of risk: technological feasibility, regulation, market acceptance, early adopters, etc. They have to realize that at any point, the project can take blows from any end and have to de-risk the capital and time spent. In fact, they have to be proactive about the de-risking process by creating contingency plans and flexible strategies.
Deep-tech startups avoid falling into “product myopia” or “falling in love” with a business-model or a specific solution
If you're interested in deep tech and want to check out some of the coolest projects being worked on, I made a list to check out:
Space: SpaceX, Relativity Space, Varda,
AI: OpenAI & DeepMind (acquired by Google)
VR: Magic Leap
BioTech: Ginkgo Bioworks, Recursion Pharmaceuticals, Tempus, Zymergen,
Robotics: Boston Dynamics, UiPath, Blue River Technology,
Quantum: Rigetti Computing
Materials: Desktop Metal, QuantumScape, Carbon,
To sum up deep tech companies, here are the 3 things you can remember:
Long term impact and R&D
Long horizon to reach market ready maturity
Intensive capital investment
That's all for today's post - if you enjoyed, please share & subscribe!