tldr: rising senior at Amherst College majoring in CS/Stats. Singapore / Jakarta / California. Fascinated by tech x business.
Hi! I'm a rising senior at Amherst College majoring in CS/stats. I'm interested in the intersections between technology and business. When I'm not in school, I split time between Singapore, Palo Alto, and Jakarta.

I took 3 gap years after high school, during which time I completed my National Service in Singapore in the Civil Defence as an ambulance EMT and volunteered in Indonesia's 2014 presidential campaign. When I'm not reading and writing notes / exploring you can find me on /r/ mildly interesting / in the gym / worrying about the rise of automation. I'm currently using this site as a my online (one of the perks of this being that it is un-lose-able and accessible/share-able from anywhere) commonplace book. A little more bio

Some other things


Articles (September)

If Sapiens were a blog post, amazing 'summary' (still 30m) of Sapiens; tl;dr: cooking, language, imagination, industrialization, and collapse of the family

neilkakkar September 2019

must-read

Some scholars believe there is a direct link between the advent of cooking, the shortening of the human intestinal track, and the growth of the human brain. Since long intestines and large brains are both massive energy consumers, it’s hard to have both. By shortening the intestines and decreasing their energy consumption, cooking inadvertently opened the way to the jumbo brains.

Yet the truly unique feature of our language is not its ability to transmit information about men and lions. Rather, it’s the ability to transmit information about things that do not exist at all.

The first millennium BC witnessed the appearance of three potentially universal orders, whose devotees could for the first time imagine the entire world and the entire human race as a single unit governed by a single set of laws. Everyone was ‘us’, at least potentially. There was no longer ‘them’.

The first universal order to appear was economic: the monetary order.

The second universal order was political: the imperial order.

The third universal order was religious: the order of universal religions such as Buddhism, Christianity and Islam.

Since all social orders and hierarchies are imagined, they are all fragile, and the larger the society, the more fragile it is. The crucial historical role of religion has been to give superhuman legitimacy to these fragile structures. Religions assert that our laws are not the result of human caprice, but are ordained by an absolute and supreme authority. This helps place at least some fundamental laws beyond challenge, thereby ensuring social stability.

The Industrial Revolution turned the timetable and the assembly line into a template for almost all human activities. Shortly after factories imposed their time frames on human behaviour, schools too adopted precise timetables, followed by hospitals, government offices and grocery stores. Even in places devoid of assembly lines and machines, the timetable became king. If the shift at the factory ends at 5 p.m., the local pub had better be open for business by 5:02.

This modest beginning spawned a global network of timetables, synchronised down to the tiniest fractions of a second. When the broadcast media – first radio, then television – made their debut, they entered a world of timetables and became its main enforcers and evangelists.

The state and the market approached people with an offer that could not be refused. ‘Become individuals,’ they said. ‘Marry whomever you desire, without asking permission from your parents. Take up whatever job suits you, even if community elders frown. Live wherever you wish, even if you cannot make it every week to the family dinner. You are no longer dependent on your family or your community. We, the state and the market, will take care of you instead. We will provide food, shelter, education, health, welfare and employment. We will provide pensions, insurance and protection.’

Good citizenship depends on basic statistical literacy,

Aeon September 2019

statistics

Statistics are not cold hard facts – as Nate Silver writes in The Signal and the Noise (2012): ‘The numbers have no way of speaking for themselves. We speak for them. We imbue them with meaning.’ Not only has someone used extensive judgment in choosing what to measure, how to define crucial ideas, and to analyse them, but the manner in which they are communicated can utterly change their emotional impact. Let’s assume that £350 million is the actual weekly contribution to the EU. I often ask audiences to suggest what they would put on the side of the bus if they were on the Remain side. A standard option for making an apparently big number look small is to consider it as a proportion of an even bigger number: for example, the UK’s GDP is currently around £2.3 trillion, and so this contribution would comprise less than 1 per cent of GDP, around six months’ typical growth. An alternative device is to break down expenditure into smaller, more easily grasped units: for example, as there are 66 million people in the UK, £350 million a week is equivalent to around 75p a day, less than $1, say about the cost of a small packet of crisps (potato chips). If the bus had said: We each send the EU the price of a packet of crisps each day, the campaign might not have been so successful.

The second problem is that we are carrying out repeated significance tests, as each year’s new data are added and another test performed. Fortunately, it turns out that there is some remarkable but complex theory, delightfully known as ‘the law of the iterated logarithm’. This shows that if we carry out such repeated testing, even if the null hypothesis is true, then we are certain to eventually reject that null at any significance level we choose.

Fortunately, there are statistical methods for dealing with this problem of sequential testing. They were first developed in the Second World War by teams of statisticians working on industrial quality-control of armaments and other war materiel.

Armaments coming off the production line were being monitored by steadily accumulating total deviations from a standard, much in the same way as monitoring excess mortality. Scientists realised that the law of the iterated logarithm meant that repeated significance testing would always lead eventually to an alert that the industrial process had gone out of strict control, even if in truth everything was functioning fine. Essentially, if we keep on checking on a process, in the end something will look odd just by chance alone.

America, the Gerontocracy, well-known but interesting nonetheless anachronisms of the Constitution and the unintended consequences of an aging democracy

Politico September 2019

aging

The U.S. doesn’t have a Politburo, but if you calculate the median age of the president, the speaker of the House, the majority leader of the Senate, and the three Democrats leading in the presidential polls for 2020, the median age is … uh … 77.

None of this means a septuagenarian can’t function effectively as a political leader. House Speaker Nancy Pelosi and Mitch McConnell are 79 and 77, respectively, and by all reports they’re operating at peak mental capacity. But to affirm that not all elderly people are impaired cognitively is very different from affirming that none is.

Wisdom may be more valuable in the digital age than ever before, because the velocity of information and normative judgments on social media, cable news and elsewhere constantly threatens to make glib idiots of us all.

But here’s the rub: The aging of America’s ruling class does not automatically increase its experience level. In presidential politics, notes Brookings Institution senior fellow Jonathan Rauch, political experience, which “used to be a selling point,” has “become a liability. Voters and the public have come to see experience as inauthenticity.”

In a November 2015 Atlantic article, Rauch plotted experience level for presidential candidates from 1960 to 2012. His graph showed a clear increase in experience level among the losers and a corresponding decrease among the winners. Gerald Ford lost to Jimmy Carter. George H.W. Bush won with more political experience than Michael Dukakis, but four years later lost to Bill Clinton, who had less. John McCain lost to Barack Obama, who’d been in national politics a mere four years.

Donald Trump, who is 73, entered the Oval Office with no political experience at all. The single greatest mental compensation that age provides was therefore unavailable to the oldest president in American history.

Old people really like to vote. In 2016, for instance, 71 percent of eligible elderly voters reported to the Census that they voted. For other age cohorts, the turnout percentages were 67 percent (aged 45-64), 59 percent (aged 30-44) and 46 percent (aged 18-29).

You often hear older Americans complain that the younger generation, with its fixation on social media, can’t distinguish between fact and opinion, making it difficult for them to apply the critical thinking necessary to consume news and be responsible citizens. A 2018 Pew survey found that Americans do indeed experience great difficulty telling these two things apart: Given five factual statements and five statements of opinion, a majority of Americans couldn’t identify them properly.

But younger Americans actually scored better on this test than older ones. Thirty-two percent of 18-49 year-olds were able to identify all five factual statements, and 44 percent were able to identify all five statements of opinion. Among the over-50 cohort, only 20 percent identified all five factual statements correctly, and only 26 percent did the same with the statements of opinion.

The list of the Constitution’s anachronisms and ambiguities is long.

Article One says Congress may “regulate Commerce with foreign Nations, and among the several States,” phrasing that strictly limited the regulation of private business at the federal level until the New Deal, when the Supreme Court reversed itself and concluded the federal government’s power to regulate private business was pretty vast. Had the Founders grasped that the modern economy would all but eliminate purely local commerce—and that it could, unchecked, alter the very climate of planet earth—they might have had more to say on the subject. As things stand, the powers of the regulatory state are the subject of endless legal combat.

Article Two says you must be a “natural born Citizen” to be president, which excludes for no apparent reason Arnold Schwarzenegger and Jennifer Granholm, who previously governed two of the nation’s most populous states. The racist “birther” movement that challenged the legality of Barack Obama’s presidency (and that ushered Donald Trump onto the national political stage) wouldn’t have been possible without Article Two.

Article Two also established that presidents be elected through the Electoral College, an antique mechanism borrowed from the Holy Roman Empire that twice during the past two decades delivered the presidency to the popular-vote loser. Some people have a problem with that.

The Second Amendment frames the right to bear arms within the context of “well-regulated” state militias that no longer exist, an ambiguity that the Supreme Court interpreted in 2008 to mean the Constitution protected the right to bear arms, after holding for the preceding seven decades that it did not. Had the Founders known the extent to which the nation would tear itself apart over the regulation of firearms more deadly than they ever imagined, they might have laid down a few broad parameters.

And so on. None of this would matter much if our government were more amenable to reconsidering first principles, but that’s getting harder, too. The Constitution can be amended, and it has been, 27 times. But growing political polarization in recent years has made that difficult. Only two constitutional amendments were ratified during the past half-century (one giving 18-year-olds the right to vote and another, more anodyne amendment that makes it a little harder for Congress to give itself a raise).

Congress could perhaps pick up some of the slack, but it’s slowed down, too. According to the Pew Research Center, Congress passes fewer substantive laws today than it did 30 years ago. Increased use of the filibuster (which is not mentioned in the Constitution, but has been around almost as long) almost certainly played a role, and a fed-up Senate has during the past decade started phasing out its use. In a provocative June 2018 essay in Commentary, the political scientist Yuval Levin posited that 231 years on, Congress had acquired a problem James Madison never anticipated: a reluctance to compete with the other two branches of government in the exercise of power. Partisanship, he concluded, had displaced ambition to legislate. Senators and representatives, he wrote, now “see themselves as players in a larger political ecosystem the point of which is not legislating or governing but rather engaging in a kind of performative outrage for a partisan audience.” Levin didn’t put it this way, but he seemed to be suggesting that Congress had grown decadent, like fin de siècle Vienna, but without the solace of Sacher tortes.

‘This Is Not the Way Everybody Behaves.’ How Adam Neumann’s Over-the-Top Style Built WeWork.,

WSJ September 2019

We Adam Neumann founder syndrome

For startup investors, the 6-foot-5 Mr. Neumann has always had the qualities they crave in Silicon Valley founders, despite being based in New York. He is intensely ambitious and a masterful storyteller with a magnetic personality who can inspire and sell.

The company emphasized its data and how it was using artificial intelligence to glean insights about buildings.

Past funders and employees tell stories of how an animated Mr. Neumann convinced them within minutes to believe in the company’s epic future.

Even former executives who disliked Mr. Neumann give him credit for an extraordinary ability to motivate employees and push the company.

SoftBank first committed $3.1 billion in new funding in 2017. Mr. Neumann has told others that Mr. Son appreciated how he was crazy—but thought that he needed to be crazier. A SoftBank spokeswoman declined to comment.

Alcohol has been a big part of the culture, particularly in We’s first half-decade. Mr. Neumann has told people he likes how it brings people together, and tequila, his favorite, flows freely. Executive retreats sport numerous cases of Don Julio 1942, with a retail price of more than $110 a bottle, and pours sometimes start in the morning.

Have we hit peak “tech” company?, mentions of "tech" in company documents has been declining

Recode September 2019

Overall, company documents referencing “tech” and “technology” peaked around August 2018, according to Sentieo data pulled from US and international public companies’ earnings calls, financial transcripts, and press releases.

For one, it could indicate that the idea of integrating technology into non-tech businesses has gone from novel and noteworthy to commonplace and unremarkable.

It could also mean that people are avoiding the term as it becomes less useful, since its overuse has rendered it basically meaningless. When companies that hawk everything from shoes to salad call themselves “tech,” what does it even mean?

“It does feel like we’re kind of past this, ‘Any internet company is a tech company,’” Paul Condra, lead emerging technology analyst at PitchBook, told Recode.

It’s also possible that the term “tech” has gone from being lucrative to being a liability.

The Faulty Metric at the Center of Private Equity’s Value Proposition, tactics to inflate IRR

Institutional Investor September 2019

Over the past two years, however, the accuracy of IRR has been called into question thanks to the increasing ubiquity of subscription lines of credit. These loans, also known as commitment facilities, have long been used in the private capital industry to finance transactions before investor capital is called in, easing limited partners’ liquidity needs and making it possible for general partners to jump on deals more quickly. But lately, fund managers have been using subscription credit lines differently — and with greater frequency.

According to alternatives data provider Preqin, the use of commitment facilities among private equity funds has more than tripled in the past decade, with 47 percent of funds launched in 2010 and later having utilized the financing tool, compared with 13 percent of funds launched before 2010. And it’s not just that more private fund managers are taking out these loans. Preqin also reported that these once short-term instruments are now being used to delay capital calls longer — which investors, researchers, and other industry experts claim is leading to artificially inflated IRRs.

Two recent papers — one from a pair of Carnegie Mellon business school professors, one co-authored by two German researchers with a BlackRock private equity director — have explored the effects of subscription credit lines on IRRs and arrived at the conclusion that the loans have meaningfully improved IRRs without increasing the actual amount of money that investors take home.

The German researchers, Pierre Schillinger and Reiner Braun of the Technical University of Munich, worked with BlackRock Private Equity Partners director Jeroen Cornel to simulate how real buyout funds would perform if they had hypothetically used subscription credit lines. Simulating commitment facility use for less than six months resulted in only a 47-basis-point increase in IRR on average, or a 20bp improvement at the median. But increasing the time frame of the loan to a year led to an average IRR bump of 120 basis points — a median increase of 57bp.

“If used extensively, [subscription credit lines] can indeed lift fund returns substantially,” they concluded.

Should Seed Investors Follow On?, data-driven simulation shows that NOT follow-up investing has the highest median return, explained mainly by the opportunity cost of doing so

AngelList September 2019

We are interested in seed (and pre-seed) investments, so we filter down AngelList’s extensive database of early-stage investments to only convertible notes or SAFEs that are either uncapped, or that have a conversion price of less than $20 million dollars. Since, especially when uncapped, these instruments can also be used

for “bridge rounds” between priced equity rounds we additionally exclude all the convertible instruments created after a company has had a priced equity round. Finally, for the purpose of simulation we restrict our universe to only those investments made in 2014, 2015, 2016, or 2017. (This means that the youngest investments we consider have had at least 18 months to season.) All told, our filtering leaves us with a universe of 1,218 seed-stage investments to consider.

Our simulation proceeds as follows. We start by fixing the number of investments an investor wants to make each year, say ten. Then in each year (2014, 2015, 2016, and 2017) of simulation, the investor draws a list of ten potential investments.

We ran 10,000 simulations of the performance of the strategies targeting ten investments per year. We experimented with different numbers of target investments; the biggest change from increasing the number of investments is an increase in median strategy performance, which is consistent with what we would expect

from the heavy-tailed power-law distribution of investment returns.

In both simulations, Never has the highest median and wins most of the head-to-head simulations (fixing the same universe of potential investments) against both Always and Double. Always has the highest mean.

The Pegasus Startup: Flying Over VCs on the Wings of Profits, Jason made a $25K angel investment in Uber, made possible through his employer Sequoia Capital, worth some $125M at IPO... talks about Pegasus vs. unicorn, but also Toptal; his accelerator is called LAUNCH

Jason Calacanis September 2019

The unicorn movement in Silicon Valley has been a great run, with startups plowing mountains of willing capital into audacious products that are changing the world.

However, a new species has taken flight in Silicon Valley, and we’ve been lucky enough to have invested in four of the six we’ve encountered. This new species is called a Pegasus.

A Pegasus startup is one that is so profitable that it is able to use its profits to soar so high, that it skips multiple rounds of funding.

We put $378,000 into Calm.com back in April 2014 and we own around 5% of the company (after selling a small percentage of our position in “idiot insurance”). Calm had $10,000 a month in revenue when we invested and have reported revenue of $7m in 2016, $20m in 2017, $80m in 2018 and an estimate of $150m in 2019.

What people don’t know is what happened between 2014 and 2018, which is that the founders put their heads down and didn’t raise any significant capital. They just poured profits into their growth, team and product — and it worked.

Their first round with venture investors was at ~$250m valuation and the second was at $1b.

No Seed Extension, No Series A, No Series B … straight to what most would consider a Series C financing.

Benefits: This is going to be fairly obvious to anyone in the industry, but to state it explicitly for new founders and civilians, there are two major benefits: (1) For every round of financing you skip, you save 10-20% dilution. If you skip 2-3 rounds of financing this could double your ownership at an exit. (2) You never have to stop working on your product to do a fundraising tour.

It’s been reported, in a spectacular scoop by my pal Amir at the awesome The Information (not fake news), that a company called Toptal raised $1.5m on a convertible note and then never converted that note into equity because it was a Pegasus with a whisper number of $200m in revenue.

The founder, Taso Du Val, also allegedly screwed his employees and co-founder out of their equity by never converting his LLC into a C corp and creating stock.

This is a huge violation of the explicit covenant of startups: “we all lose a decade of our lives trying, or we get fabulously rich together.”

Toptal is the Dark Pegasus, the worst of all animals … a creature so evil and filled with greed that not only do they want to preserve their cap table, they want to do so by screwing everyone they can — including their own family (aka, investors, employees).