Anti-fragility in the startup world

by Adrian Lloyd

If you haven’t read Nasssim Nicholas Taleb’s book, Antifragile, I suggest you do.  So did Vinod Khosla in a talk I attended at Stanford Business School, my alma mater, last week.


I have almost finished the book and it is stuffed with thought provoking ideas that I want to write about, but I want to choose the controversial subject of the fallacy of entrepreneurship that is endemic in the Bay Area and creeping into London.

First, a brief description of what antifragility is.  Taleb, who also wrote Black Swan, noticed that in this World there are the:

  • Fragile – those systems or objects that will be damaged by an energetic attack or volatlity.  Your china vase and highly leveraged financial systems
  • Robust – those systems or objects that will withstand an energetic attack or volatility.  A plastic cup (to a degree) and smaller, unleveraged bank
  • Antifragile – those systems or objects that will benefit from an energetic attack or volatility.  The Hydra of Greek mythology that grew an additional head each time one was removed or, simply, a financial instrument that increases in value as volatility increases

Back to the fallacy of entrepreneurship.

Taleb writes that the longer a system or object has existed, the longer it will survive into the future.  He is a statistics obsessed person, and so this is purely a concept based on statistics and it stands up to rigourorous analysis.  It’s such a simple concept, but one that made this venture capitalist shudder at first.  He has great respect for the ancient (he is a classical scholar) and great disdain for the new until it has proved the test of time.  He LOVES time as a measure of something’s worth.  In fact it is just about the only thing he measures something’s value on.

He uses numerous simple examples: the utensils we use to cook and how we cook (pots over fire) has not changed in a few thousand years.  We wear similar kinds of shoes as the oldest examples found to date, and in fact latest releases from Nike and the like are more like ancient, thin-soled shoes than the well padded versions we grew up with.  The same goes for ideas – he cites many examples of ancient philosophers who’s ideas are alive and well today (all the major religions for e.g.), usually emulated by moden philosophers who claim to be coming up with something new.  A book that has been in print for 40 years is more likely to last 40 more years than a book that has been in print for 1 or 5 or 10 years.  Aeroplanes and cars have hardly changed in basic principal since they were invented.    He believes that these ancient and old technologies will be around for a lot longer than the brand new given how well they have stood the test of time – they are antifragile at best and robust at least.  What he gets excited about is not what will be new in the future, but which fragile systems and objects will have disappeared in the future.  It’s a reductionist view of innovation.  Futurists may imagine a future full of flying cars and hover-craft, robots and printed hamburgers, but they will probably (statistically speaking) be wrong according to Taleb.  The basics which have been around forever will likely be around forever (statistically speaking – of course there are examples that will buck this rule).

Does this mean Google X should give up on creating driverless cars and other extraordinary innovations?  Of course not – it’s human nature, and particularly Google’s nature, to push at the boundaries.  But they should perhaps consider innovating in a radical way within the realms of this concept.

Does this mean Episode 1 should give up trying to invest in the cutting edge technology it finds in the market?  Of course not.  But we should be aware that we can learn from what has survived the test of time, investigate why it has survived the test of time and perhaps look for old or even ancient analogues to what we find in the market today.  It took me a few seconds on Google to find ancient Greek and Middle Age examples of peer-to-peer lending systems a la Rate Setter or Funding Circle for example.  Perhaps we should also look to invest more in those startups that are attacking and replacing the fragile – like peer-to-peer systems that are disintermediating the large, too-big-to-fail, financial institutions.


  • aelloyd

    My colleague Paul McNabb criticised my blog for not discussing those new innovations that have succeeded such as telegraph, or electricity, or the microchip or perhaps the smart phone and why they have succeeded and how hard it is to displace things that are “good enough”. It was a fair challenge and one I will address now.
    The answer is simple: those innovations mentioned above are amongst the few that were genuninely new to the human race and added so much obvious, tangible and immediate value that they were able to survive. This was my point in the blog about Google X and Episode 1 persevering in their quests to respectivley invent and invest in the best new technologies out there. The telegraph allowed people to do better what they have always done – communicate. Electricity allowed people to do better what they have always done – use light to brighten the night to allow for continued socialising and later work. The microchip allowed people to do better what they have always done – calculate and analyse data. The smart phone allowed poeple to do better what they have always done – communicate with each other and calculate and analyse data etc etc.
    The good enough point is critical and something we think about every time we see a pitch. My rule of thumb for an enterprise business is that if it’s not 10 times better than the current incumbent or way of doing things, it won’t get the attention of time-pressed executives who need to decide. Obviously the number 10 is arbitrary in a way, but when execs assess a new technical offering they will make an RoI calculation and work out how much better than what they already have the new startup is. When you factor in political, emotional and other switching costs, it has to be a hell of a lot better to make the grade. Inertia and “good enough” are dangerous forces pitched against startups every day. As a Founder you need to convince your customers and us that your solution is 10x better than what exists.

  • krdavis

    Great post Adrian. I think the nature of VC investing is itself antifragile. Another aspect of antifragile systems is that the downside is limited (ie the worst-case-scenario is the norm) and the upside is near-unlimited. Startups are a perfect example of this: the baseline is crashing and burning, whereas $100bn-IPOs are entirely possible. VC portfolios have ‘baked in’ a high probability of failure, so that every startup that exits is a plus, and you simply cross fingers that the exit is significant enough to carry the others.

    Within a startup, they can make themselves more antifragile by lowering their fixed cost base, staying lean, not hiring too early, and taking the minimum necessary funding. Besides maximising runway, it limits the downside while stile exposing them to enormous potential upside.

    • aelloyd

      Well said and thank you for contributing to the conversation


Website by

We Are Ignition