Natasha Gunther is an Investment Associate at Innovation Endeavors, an early-stage venture firm co-founded by Eric Schmidt. She spends her time investing across supply chain/logistics, healthcare, ag/food, and all things hard tech. Natasha joined Innovation Endeavors from healthcare SaaS platform, PatientPing, where she was Chief of Staff and Product Manager, after her time as a strategy consultant with McKinsey's operations and healthcare practices. Though a mechanical engineer by training, she is a generalist at heart.
EVCA: What’s your favorite thing about working in venture capital and why? How about your least favorite and why?
Natasha: My least favorite aspect of the job is when we come to pass on a founder. Over the course of a diligence, you are their ally; you get to know them and deeply appreciate them as founders and as people; you spend all of your time thinking of how to make it work. But sometimes, our role needs to be that of the key risk-detector (aka the cynic). Now to the good part - being in VC is to have the insane position of getting to ask any question of any person in any field (often a person 10x smarter and more expert) with no ulterior motive outside of 1) learning and 2) getting to the best answer. As a true nerd, this is an absolute dream.
EVCA: What is your most contrarian view on an existing or emerging technology trend?
Natasha: On digital health: It is incredibly difficult to change anything about the core issues of healthcare (access, cost, discontinuity, and privilege) without investing in the direct provision of care (loose carve out for life sciences + diagnostics + med devices). If you try to "revolutionize" from the outside, you'll eventually just end up frustrated with the barriers thrown by the status quo. On an unrelated note: robots can be good for employees too, not just owners.
Morgan Polotan has over 5 years experience in venture capital, and is currently a principal at Comcast Ventures, one of the most active corporate VCs in the world. Morgan also helps to run Genacast Ventures, Comcast's enterprise seed fund. Morgan's investments include Dandelion (residential geothermal), Zapata (quantum computing platform), Neural Magic (AI performance tooling), goTenna (p2p mobile connectivity), Confetti (marketplace for employee events), and SafeBase (automated security programs for B2B startups).
Prior to Comcast Ventures, Morgan invested in enterprise startups at Bloomberg Beta, a $75M seed fund backed by Bloomberg, where he invested in Donut (automated employee on-boarding) and Stedi (messaging platform for B2B trade), Campuswire (messaging platform for universities), and Virgil Security (encryption library for mobile apps). Before venture, Morgan wrote code, with a preference for JavaScript and Scala. On the side, Morgan likes to shoot hoops (reclaim high school glory), pretend to DJ, and obsessively check the price of Bitcoin and Tesla.
EVCA: What’s your favorite thing about working in venture capital and why? How about your least favorite and why?
Morgan: I love to partner with founders from an early stage to help build their businesses. I imagine myself as a coach. I'm not on the field, but I'm exposed to a lot of plays, so I can help guide the incredibly talented group of founders I'm lucky to work with. There is nothing more satisfying than watching founders achieve in the startup world, and knowing that I played a very small part in their journey.
My least favorite part of venture is telling founders no. There are so many reasons that investors pass, and 80% of the time a pass is not a negative judgment of the founders business. Investing is a 10+ year commitment, and founders and investors need to have chemistry. They need to be a good "fit". Ask any VC about their anti-portfolio, and it almost surely outperforms their actual portfolio (even the best ones!). This fact should give comfort to all the rejection founders must endure.
EVCA: What is your most contrarian view on an existing or emerging technology trend?
Morgan: I think we're reaching the limits of deep learning. GPT-3, which is a deep-learning model, is a monster achievement, and it performs very well under certain situations, but it doesn't seem to be closer to truly understanding text. Read Gary Marcus's critiques of deep learning (Rebooting AI is a good book to start with) to better understand deep learning's limitations. Deep learning bulls believe the answer is bigger and more powerful models, but I'm not so sure. I think we need to revisit some ideas around cognitive architectures and combine that with the best-of-breed deep learning approaches.
Pedro Domingos' wrote a book, The Master Algorithm, that outlines this approach of combining different schools of thought to reach artificial general intelligence (AGI). This approach hasn't received much attention from VCs, which is why I believe it's a contrarian view. DARPA, on the other hand, has funded dozens of academic projects in what they call the 3rd Wave of AI, which are AI models that have a contextual understanding of the world around them and can generalize beyond their training data. I believe this is the future, and that we'll see some of those academic projects turn into startups in the next few years. I'm excited to invest in one or two :).
EVCA is a 501(c)(3) organization, EIN# 83-4254999
