Aaron Harris

Facepalms all around. Supply chains are hard, and the idea that you can just write software to "solve" them is a level of hubris that defies description. Supply chains aren't a problem you can solve with big data. They aren't a problem that be solved with machine learning. In fact, you could combine the worlds top researchers in Computer Science, Economics, Operations Research, Mathematics, and Statistics, put them in a room together for a decade, and still not "solve" even 1% of the problems that Supply Chains face. I know the ITA presentation on Air Travel Planning Complexity[1] has been circulated here many times, but it always blows peoples minds how complex that problem space is, and that type of problem is miniscule in comparison to end to end supply chain optimization. I work in a problem space that is the combination of 9 different NP-Complete/Hard problem spaces (I've counted) on a regular basis...we are barely scraping by with millions of lines of code and entire data centers of our servers running at peak utilization 24/7, and that is with state of the art heuristics solvers and hundreds of PhD-level researchers, and that is only a subset of the Supply Chain problems that my company faces. To say that Good Eggs bit off more than they can chew is a MASSIVE understatement.

Supply Chain Management is a field that is heuristic-driven because there isn't a solution. The heuristics that drive modern food distribution are the result of a real world genetic algorithm…

> I'll add that my own experience of licensing (my own) IP from a top-25 university TTO has been interesting

In my experience, there's a lot of variance between universities, and even between tech transfer officers within a given university. Another complication: at Harvard/MIT, where professors may have cross-institutional appointments, you need to nail down exactly who you'll be dealing with (for example, the Broad has their own TTO: https://www.broadinstitute.org/osap). There are also other issues, such as IIAs & CRADAs.

> What were your thoughts on the whole "cult of the molecule" thing from the interview?

As you said, it fails to take into account existing attempts to form partnerships by biotech VCs:

- http://lifescivc.com/2014/04/pushing-forward-with-collaborative-rd-models-in-biotech-roche-spero-and-biogen-ataxion/

- http://www.bizjournals.com/seattle/blog/techflash/2014/01/a-new-approach-to-biotech-investing.html

This is one of those situations where presumption of stupidity (http://www.aaronkharris.com/presumption-of-stupitidy) can be very tempting, given the magnitude of the problems that remain unsolved in the biosciences. It can also take quite a bit of digging to find this info - it's not splashed all over the front page of TechCrunch, like the latest batch of YC startups are.

This is essentially the definition of survivorship bias.

http://www.fastcodesign.com/1671172/how-a-story-from-world-war-ii-shapes-facebook-today

"In WWII, Allied bombers were key to strategic attacks, yet these lumbering giants were constantly shot down over enemy territory. The planes needed more armor, but armor is heavy. So extra plating could only go where the planes were being shot the most."...

"[A guy in charge] said the military didn’t need to reinforce the spots that had bullet holes. They needed to reinforce the spots that didn’t have bullet holes."

Having experienced 2 "WFIO" things in the last 6 months, I agree that entrepreneurs need to be resilient, but I think it's more valuable to read post mortems than these "we almost died" posts because they prove more instructional.

For instance, Aaron's post about Tutorspree is a really useful post for me, although it definitely has less of a feel good vibe to it.

http://www.aaronkharris.com/when-seo-fails-single-channel-dependency-and-the-end-of-tutorspree

I believe his point is that although it is useful to analyze someone to see what they're doing wrong (so that you don't repeat the same mistake), it would be more useful to study what they are doing right so that you can apply those lessons to your situation. The former is quite easy to do and runs the risk of lulling one into complacency and a false sense of security, while the latter exercises your brain and demands greater effort and a more impartial assessment of the situation.

http://www.aaronkharris.com/presumption-of-stupitidy
https://sivers.org/ss

In the grandparent commenter's case, he is saying that everyone is focusing on the negatives of Chinese research when there are some useful lessons regarding regulation that could possibly be ported to the US.

Took a while to find but I think this might be what you were after?

"Small children are great at asking questions. Their questions are simple, direct, and utterly without ego. If you spend time with 2-5 year olds, you’ll find yourself answering a near infinite series of progressively more challenging questions. This is one of the main paths by which children learn about the world.

Adults are generally bad at asking questions. Adult questions come with long preambles, caveats, and agendas. Adult questions don’t seem to be designed to learn about the world, they seem to be designed to avoid looking dumb."

https://news.ycombinator.com/item?id=22729028

https://www.aaronkharris.com/asking-questions

I not only had zero connections to YC.. I actually had zero connections to silicon valley! We (RedCarpet.Cash) are a India-based fintech startup - IMHO the third to ever get funded by YC.

We were also unusual because we are probably the first Indian YC startup with on-ground operations. We went through 3 rounds of YC interviews (not just the one!) .. because the interviewers had to learn about India. Our first interviewer was mwseibel !

What impressed me was that YC learned about India, invested in us and supported us through a pivot back in India. Big shout out to Aaron Harris, who is probably the resident Asia expert now [1]

[1] http://www.aaronkharris.com/carts-without-horses