Investing In AI-Powered Prosthetics

Investors lessons from Employee #2 at Pinterest & A Deep Dive into AI-Powered Prosthetics.

Last month the EPO community invested over $3M into startups. For a community this young, that is unheard of. 

As this community grows, so does our level of access. 

So i'm excited to share a new regular feature here at EPO - Investor interviews with the world's top startup investors.

Today We’ll explore: 

  1. 🧰  Investor Interviews: Sahil Lavingia - CEO at Gumroad, employee #2 at Pinterest, and angel investor in over 100 startups.
  2. 🦾  Market Watch: Is there are market to watch in AI Powered Prosthetics?

🧰  Investor Interviews: Sahil Lavingia, CEO at Gumroad and Investor at

This Thursday I am meeting with Sahil Lavingia, CEO at Gumroad, and avid angel investor. After leaving Pinterest in 2012, where he was employee #2, he founded Gumroad, a company that has distributed over $500M to internet creators since its founding.

Because of his standing within the Silicon Valley founder community (check image below!) he was able to raise a rolling fund that invests over $12M into 100 startups per year.

Sahil Lavingia & Patrick Collison, CEO at Stripe (Valuation: $95B)

I’m going to learn about his investing playbook, start trends he’s watching, and companies he’s excited about. But more importantly, what questions do you have for Sahil? Reply to this email with questions and I will get them answered.

🦾   Market Watch: AI Powered Prosthetics

The brain-computer interface market (BCI), a microsphere that includes all the prosthetics and cognitive assistance initiatives underway, is expected to reach $3.7 billion in revenues, with an annual compound growth rate of 15.5 percent during 2020-2027.

For investors in the medical devices space, AI’s successful exploitation of data to solve problems, and the combination of capital, risk and innovation an entrepreneur can use to build businesses, is compelling. 

Enter the age of smart prosthetics: 

Atom Limbs

Artificial intelligence has increased the sophistication of prosthetic development, by taking rough approximations of the data input from the brain and generating actionable movements in the world. 

Atom Limbs, a new player in this space, leverages AI that was developed with over $120 million in R&D grants from DARPA and US Dept. of Defense. Unlike traditional prosthetics that are controlled by physically flexing your bicep & tricep, Atom Touch doesn't require any physical movement. You just think, and it moves. The company uses AI to listen to nerve signals on the skin, and real-time moves a robotic arm with full range of motion. 

The key to Atom Limbs is their non-invasive approach to sensing information, no matter how inadequate, then feeding it to AI algorithms that identify the intent and can execute it via the prosthetics.

The Market for orthopedics is large:

Just how big is this market? 

  • There are 2.1 million people living with limb loss in the USA, and that number is expected to double by 2050.
  • 185,000 people have an amputation each year in the US. This means that 300 to 500 amputations are performed every day. 
  • There are more than 1 million annual limb amputations globally.

Despite the size of this market and the high potential for patient impact, new technology-forward players haven't performed well. ReWalk Robotics, an equally non-invasive robotic walking assistance system that uses powered leg attachments to enable paraplegics to stand upright, went public in September 2014. The company has since seen its share price drop to a market cap of $74.5M. 


The technology deployed by Atom Lims is impressive, but despite the potential for enormous advances, investors need to be conscious of risks committing capital.

  • First, while AI can bring executional clarity to often foggy neural firings, there is a lot of scope in that translation for abuse. 
  • Second (and, frankly, unfortunately) investors need to weigh the revenue per patient against costs. If that figure is negative, great – you’ll have my admiration and that of many others.
  • Third, once the technology is proven, investors need to factor in the cost of FDA approval and potential other regulatory costs.
  • Next, insurance companies already kick at being asked to pay for standard prosthetics. Costly, AI-based systems have yet to be tested in this economic crucible.
  • Finally – consider your portfolio. Companies like Atom Limbs are considered deep tech. Overweighing your portfolio with deep tech hardware products is a risky proposition. These companies are typically extremely capital intensive and require years of heavy R&D and regulatory approval before launch. 

As neuro AI tackles these and related problems, the investment opportunities for the best ideas could significantly enhance human flourishing and well being, but in today's payor led healthcare system, i’d have to ask myself, are we too early?