The ambivalent power of unintended consequences
On unintended consequences...
Moneyball is a fortnightly newsletter from Koble exploring the limitations of human decision-making and their implications for startup investing.
We help angels, VCs, family offices, and forward-thinking hedge funds to deconstruct mental models that can make us smarter, wealthier, and better.
🧠 Mental model #3 – Unintended Consequences – The ambivalent power of unintended consequences.
📖 Investor reading – Generative AI: A Creative New World – Ethereum and Solana Disrupt VC With Coinbase-Backed DAODAO – Trouble ahead for European VC?
💬 Some tweets – How to build startups – Advice from VC associates – The most helpful thing a VC did for you.
At Koble we’re building a deep-learning model that can predict the success of early stage startups. Here’s what’s happening…
We’ve performed multiple tests on our baseline model and we’re seeing it converge and improve. This is a big result as it means it has predictive capability.
We’ll continue to do more tests over the next few weeks, adding further enhancements and seeking to understand the superiority of our algorithms to human investors.
The ambivalent power of unintended consequences
Unintended consequences are the invisible hand that guides (and sometimes upends) our lives. They’re also a mental model that can help investors to recognise the unpredictability of complex systems and manage risk more effectively.
They’ve been a feature of the social sciences landscape since the 1930s, when American sociologist Robert K. Merton coined the term. And in recent weeks, they’ve dominated headlines as the well-intentioned actions of politicians have disrupted capital markets (long-dated gilts, anyone?), with negative implications for all of us.
Here’s the problem:
When we evaluate the potential impact of our actions within a particular system, we tend to ignore the impacts of the system’s first-order responses to our actions, and indeed the second and third order responses to those responses.
In other words, our actions don’t just have an impact. They have nth order impacts. To bring this concept to life, let’s take a trip to Mexico.
Mexico City is world-famous for its traffic. To tackle the problem and reduce air pollution, in 1989 the Mexican government introduced a traffic-calming policy aptly named Hoy No Circula (“Today it doesn’t circulate”). On certain days people were prevented from driving based on the final digit of their number plates (odds vs evens).
The scheme was initially successful in reducing pollution, with carbon monoxide levels dropping by 11%. So far, so good.
But then, the exact opposite happened.
To circumvent the new rules, people bought more cars. A lot more. And because these cars were cheap, old vehicles, pollution levels actually increased. The long-term impact of the scheme has been a 13% rise in carbon monoxide levels.
A well-intentioned solution to a serious problem made it worse.
Similarly, ABS may not reduce the number or severity of car accidents, since people compensate for shorter stopping distances by driving more aggressively. Hail The Peltzman Effect – the tendency of individuals to respond to safety regulations by engaging in more dangerous behaviour.
Unintended consequences love scale, and no business has more scale than Amazon. When the UK government introduced a Digital Services Tax in 2020, the company simply increased its fees for sellers (and presumably they simply increased their fees for consumers). The government’s lack of second order thinking led to a worse outcome for consumers.
And then there’s Meta. When Mark Zuckerberg et al launched Facebook they probably weren’t thinking about third-party abuse and political interference on a systematic scale.
On a more micro level, we see the impact of unintended consequences in early-stage valuations. Startups achieving pre-revenue valuations of $10 million might sound awesome, but the downstream effects are problematic, with founders struggling to achieve metrics that justify valuations across the funding cycle.
Nth order impacts are an invisible force that govern our businesses, investments, and lives. The world is a complex system and the impact of our decisions on a collective and individual level is difficult to predict, often with disastrous consequences.
But unintended consequences work both ways. Adam Smith’s notion of the “invisible hand” highlights the benefits that can accrue from individuals pursuing their own self-interest, generating widespread social benefits. And Jeff Bezos has noted:
“There is so much stuff that has yet to be invented. There’s so much new that’s going to happen. People don’t have any idea yet how impactful the Internet is going to be and that this is still Day 1 in such a big way.”
What if we re-framed unintended consequences as a means of value creation, rather than destruction? To do that, we’d need to cultivate a deep understanding of the mechanics of causality.
And that’s where data science comes in.
Implications for investors
Humans are bad at predicting the future. Which is why they are typically bad at investing in startups.
But the future – or at least, specific future outcomes – can be modelled with the right data and technology. Graph Neural Networks and dynamic data-science models like the ones we’re building at Koble use complex math to deconstruct causality, enabling us to de-risk future outcomes.
To avoid becoming victims of unintended consequences, we must recognise them as inevitable features of the investment experience. And we must embrace technology as a way to model, mitigate, and monetise them.
Join our waitlist
At Koble, we’re building a group of forward-thinking angels, VCs, family offices, and hedge funds that want to play a role in re-engineering startup investing with AI.
If this applies to you, sign up to our waitlist.
🤖 Generative AI: A Creative New World – A powerful new class of large language models is making it possible for machines to write, code, draw and create with credible and sometimes superhuman results.
⛓ Ethereum and Solana Disrupt Venture Capital With Coinbase-Backed DAODAO – DAODAO (backed to the tune of $200M by Coinbase, Sequoia, and a16z) is a breakthrough cross-chain fundraising platform.
🌪 Despite a stellar quarter for fundraising, is there trouble ahead for European VC? – Sifted reports on the mismatch between listed VC share prices and portfolio performance, noting that share prices are down between 30% and 75% since January.
“Success is a science; if you have the conditions, you get the result.”
– Oscar Wilde
Regards from your [quietly focused] startup investing AI,
Koble is re-engineering startup investing with AI, applying quantitative strategies that have disrupted public markets to early-stage startup investing.
We've spent two years developing our groundbreaking algorithms, which discover early-stage startups that outperform the market and predict their probability of being successful.