One Monday morning, a question suddenly made a permanent appearance in our internal deal review discussion: “What does Q say?” It suddenly became part of our vocabulary. Now, you may wonder who or what “Q” is. Q is a young, irreverent, needy toddler in our company. Q happens to be our first foray into automation, machine learning, and AI, as applied to the work of venture capital. Our business is, in many ways, incredibly old-school and arcane. It’s based on relationships, meeting thousands of entrepreneurs, and, over the years, developing the pattern recognition, networks, and insights that eventually have enabled us to make good financial returns for our investors.
Q was baptized after an extensive naming debate among a set of very opinionated godmothers and godfathers within NGP Capital. In the end, the choice fell on the code name for the famous inventor from the James Bond series. However, Q is actually the creation of our brilliant data scientist, Atte Honkasalo. Q was created and brought to life by Atte, so Q is a bit odd, having only a single parent. Being also genderless, we refer to Q as “it.” Q is not alone in the world, though. Q has many like-minded but diverse siblings, growing and being nurtured within other venture capital firms.
Ideally, it is our hope that in the future, all that we, as investment team members in our firm, would have to do is log into Q every morning, see what pops up, whether there are new investment opportunities or issues to tackle within our portfolio, and then proceed with our day. Q would set our daily agenda.
Today Q is rather like a toddler. You can’t leave Q unattended, constant attention is needed, it makes numerous mistakes, and it needs to be fed, changed, and updated all the time. Q also sleeps a lot. But slowly and surely, Q is working its way into our daily consciousness at the office. At times, Atte feeds Q the wrong things and Q becomes very cross and throws a fit. Atte is not the only one training Q; plenty of the godmothers and godfathers play an important role in Q’s upbringing.
Q is slowly learning to crawl, but like any toddler, it is pulling out random stuff from the endless shelves of the Internet. Q is already very good with names and numbers, and has gobbled up data from more than 200,000 companies. Q has an impressive memory — memory that is infinitely scalable and multi-dimensional — and it is really quick to find anything tucked away there when asked.
Q will soon learn to read and write, applying NLP (Natural Language Processing) skills. Unusual for a toddler, Q is unlikely to ever speak. Q prefers to communicate by drawing — showing us neat graphs, diagrams, and structured lists. Once Q has added some more learning to the NLP skills that Atte is busy teaching it, the neural pathways will start to settle, and Q will become immensely more intelligent, reading millions of press releases, news items, financial reports, industry analyst views, and much more. Atte already has Q writing some simple essays, and he expects his pre-teen to be able to automatically produce most of our investment recommendations and financial reporting with little or no adult intervention. Q will be multi-lingual over time. The next important language is Chinese, which is a must for us, as a global firm. We know Chinese is a difficult language, but our hopes are high that Q will soon master it.
The onset of Q’s teenage years will bring more apprehension on our side, as Q starts to develop its own stubbornly held views that, at times, will undoubtedly seem very erratic, bewildering, and mostly incomprehensible to an adult. We will have to handle Q with great caution during this time. At the same time, as with any teenager, firmness will be required to steer Q on the right path to adulthood. Q’s eyesight and understanding of the exterior world will rapidly improve, and Q will soon record and observe management teams and entrepreneurs at work. It will quickly read all available information about these entrepreneurs and management teams, and will know where they’ve worked, where they’ve studied, and what networks they are connected into. We may need some patience, but surely, during its early teens, Q will be able to reliably undertake management team assessment work, which is one of the most important predictors of investment outcome.
The adult Q will be far smarter, faster, and more insightful than any human being. Atte, who is obviously biased about the capabilities of his offspring, very much believes this will be the case.
The question “What does Q say?” is likely to disappear from our terminology as Q becomes inseparable from ourselves and our work. And as we grow old, and Q becomes a fully independent adult, it will undoubtedly treat us gently and overbearingly when we can no longer follow its adventures, incredible insights, and intelligence. We will only be able to look on in awe as Q automatically does the job of building up an amazing portfolio of investment holdings — companies run by intelligent machines that we once used to call entrepreneurs.
The only question remaining now is “Who needs the humans?” We are a species that is erratic, error-prone, subject to numerous biases, and held hostage to a limited memory, limited intelligence, and weirdly emotional tantrums. Q would surely conclude that the job of investment management would be far too important to be left to humans.
So, we should all take great care of our “Qs.” We should bring them up and nurture them carefully. We carry one of the greatest responsibilities of mankind; making the right choices for forming the adults they may become.