Leave it to ME to check out "this show people have been talking about" and the people were 15 accounts on Tumblr, 2 deactivated, 5 of them made a post about it three years ago and now I have a new hyperfixation nobody wants to listen to

Meet The Genetic Sleuths Solving Decades-Old Trans Murders

On the blustery morning of November 26, 1983, a beachgoer spotted a still-warm body in Half Moon Bay, California. The victim, who looked about 20 years old, had been stabbed more than 20 times and left in the sands near Pillar Point Bluff. Their wrists were slashed, their face bruised and swollen. (Out of respect for the victim’s unknown gender identity, Motherboard is using they/them as a pronoun.)

At their time of death, the person was presenting as a stylish, slender woman. They were 5-foot-10 inches tall, wearing an auburn pixie cut and casual clothes: yellow capri pants and a turtleneck over a foam-form bra, fishnet hose, and two pairs of feminine underwear. A Madonna-style white metal crucifix hung around their neck. 

When taken to a medical examiner, the victim's body was misidentified as male, and nobody ever came to claim it. In an effort to identify the individual, cops dubbed them “John Doe #83-26” and released a crime sketch depicting a man. It failed to convey their gender identity or expression, including that they were likely wearing makeup and going by a woman’s name.

The case of Pillar Point Doe soon went cold and their identity remained a mystery for 35 years—until two genealogy sleuths recently cracked the case. The trans couple, who specialize in cold cases involving trans and gender non-conforming people, found the forgotten victim’s birth name through an online DNA database, reviving the hunt for their killer.

“I would work until I passed out”

Lee and Anthony Redgrave traced Pillar Point Doe’s relatives from Wales to Utah using the family history site GEDmatch, known for its role in finding the notorious Golden State Killer. The search was close to home for the Redgraves, who toiled obsessively for months without pay. 

“I would work until I passed out. I’d cry myself to sleep at night, and have dreams where I was woken up thinking that [the victim] was telling me their name,” said Anthony, who along with Lee, ran a small team for the DNA Doe Project, a non-profit that identifies deceased people through forensic genealogy.

Lee added: “There are a lot of factors—and homicide detectives have absolutely no idea how to do this.”

The Redgraves were inspired by personal tragedy to help solve the case. In January 2018, a transgender friend of theirs, Christa Steele-Knudslien—a beauty pageant organizer and trans activist—was beaten and stabbed to death. The attack, which came after another friend’s suicide, sent Lee spiraling into a depression.

“It really tore my brain up,” said Lee, 41, a non-binary night owl with arms full of tattoos. “I got depressed, and when that happens I usually throw myself into a project.”

Anthony and Lee Redgrave

When a true crime-loving friend recommended they volunteer for the DNA Doe Project, it seemed like a good distraction.“Partially, I’m sure, it was her being like, ‘You have to stop being in a funk,’” Lee said. “We both felt really helpless about Christa—and this was something we actually could help with.”

They had plenty of experience with genetic genealogy, but they knew the limits of DNA and family tree matches for transgender victims. Database searches often lead to “dead names”—birth names victims no longer use, and aren’t known by in their communities. And most law enforcement systems don’t allow searches across sex marker categories, blinding them to some gender non-conforming folks.

A “trans-informed” perspective could shed some light, considering trans people are more likely to be the target of violent, unresolved crime. “Being a trans person, I know I’ve been incredibly fortunate not to have had a bunch of horrible things happen to me,” said Anthony, 38, a soft-spoken Civil War buff with a long ginger beard. “That was a driving force.”

When the Redgraves first heard about the Pillar Point Doe case in July 2018, they knew almost immediately it was ripe for a genetic gumshoeing.

An Unlikely Partnership

The victim, who was carrying no identification, had been found only two hours after they were stabbed in the neck and chest, allowing cops to collect a piece of blood-soaked blotter paper known as a “blood card.” This meant Pillar Point Doe’s DNA didn’t have to be extracted from bone, a longer and more expensive  process. And yet it would likely show a complete picture of the victim’s entire genetic makeup, one that could be extracted in a lab and uploaded to GEDmatch, they said.

But the couple still had to convince the San Mateo County Sheriff’s Office to hand over the blood sample—and to team up with them on the investigation.  

The Redgraves had a hunch cops would be willing. Their request came on the heels of the Golden State Killer’s arrest a few counties away, and in a section of northern California that tends to be queer-friendly.

“The thought was it was a good case because it was the Bay Area. We expected  there would be more friendly law enforcement and a LBGTQ liaison in the Bay Area,” said Anthony. “We had to give an elevator pitch to the department, like, ‘This is why we want this specific case, and this is how it will benefit you.’”

The cops, it turned out, were game. The Redgraves signed non-disclosure agreements and— in a rare move—police released Pillar Point Doe’s private case files,  including the blood card along with crime scene and coroner photos. 

The Redgraves agreed to do the genetic sleuthing, then pass off next of kin matches to police, who would talk to relatives, collect DNA samples and handle the investigation from there.

It was an unlikely partnership. Many trans people refuse to work with cops since law enforcement has routinely targeted the community, trans activists and experts said. According to a 2015 survey, at least 57 percent of trans respondents said they would be afraid or uncomfortable going to police for help.

“It comes from being abused or not taken seriously by officers who historically have been disrespectful or dismissive of trans people,” said Rodrigo Heng-Lehtinen, a policy expert for the National Center for Transgender Equality. “There’s a real stigma. Sometimes just being visibly trans in public is enough to get stopped or harassed by police on suspicion of being a sex worker—whether or not you actually are.” 

Cases with trans victims have long been de-prioritized by cops, who assume victims are sex workers, living a “high risk” lifestyle or are “disowned” by their families, Heng-Lehtinen said. 

But a lack of trans awareness is bad for police, too. Detectives who are ignorant about the community are more likely to use a transgender person’s dead name  because it was printed on a government-issued ID, or to seek outdated information from estranged family members who knew them pre-transition, he said. It keeps those investigators from understanding the whole picture. 

“If you’re an officer who’s asking around for Mark Smith and everybody in the neighborhood knows her as Marcia, that’s not helping anybody,” Heng-Lehtinen said.

Lee chalks it up to lack of education and training. “If you look at popular media over the past 20 years, the characters that are dressing opposite of what they’re ‘supposed to be’ are usually trying to trick somebody or get away with something—the end of ‘Ace Ventura’ is a classic example, or ‘The Crying Game,’” Lee said. “You get a lot of that mentality still in law enforcement, just because they haven’t had an alternate education.”

Identifying Doe

The Redgraves quickly got to work on creating a more gender-accurate forensic sketch of Pillar Point Doe. In the 80s and 90s, at least three drawings had been made of them, all wildly different. 

One showed a “partially-Asian goth” guy with boxy slicked-back black hair, Lee said. Another depicted a shaggy-haired Val Kilmer look-alike with almond eyes. All were of men, and none were quite right.

“It seemed like [police artists] were trying to make this person look male,” Lee said. “Considering they had natural hair, not a wig, and were wearing pants with multiple layers of hose and underwear, it’s likely that they were tucking to have a more female appearance,” Lee said, citing details about the victim's outfit.

“They were probably attempting to pass as female as opposed to someone who was [a] drag performer or engaging in prostitution while cross-dressing.”

Based on those clues, the male sketch on fliers would have likely been lost on Pillar Point Doe’s queer “chosen family”—or anyone who saw them the night of the murder, the couple said. So using crime scene and coroner photos, the Redgraves and an artist came up with a new sketch that depicts the victim with a more feminine look, a yellow outfit and natural-style makeup.

In March 2019, Pillar Point’s blood card came back from the lab. It showed Pillar Point Doe’s entire genome sequence on a huge hard drive. 

From their cozy home office in central Massachusetts, the Redgraves and a small team plugged those chunks of genetic code into GEDmatch, which compares DNA from testing sites like 23andMe and ancestry.com to find possible relatives with similar genetic makeups. Unlike law enforcement’s Combined DNA Index System (CODIS), the site can pinpoint distant ancestors, not just immediate family members. 

It works like this: Say you find a painting in a park with no signature and you want to learn the name of the artist. If you could somehow scan the piece’s complex colors and brush strokes into a massive database of art, you might be able to match it to the person who made it. Other paintings by the artist with similar patterns—a distant cousin, in this analogy—may also pop up.

In general, DNA evidence is only as accurate as the people who collect and analyze it. Technicians have been known to misinterpret samples, and police have submitted tainted or mixed genetic material. But Pillar Point’s blood card appeared to be a solid sample, the couple said.  

The search led the Redgraves to a small town in Wales, where Pillar Point Doe’s distant relatives once worked at a glove factory. “We kept finding people who descended from this really specific family, but then finding the right branch turned out to be really hard,” Lee said.

Scores of unwed mothers hailed from the town for unknown reasons, leading to frustrating genealogy dead-ends. “It happened over and over again in this one little town,” Lee said.

The couple built a massive family tree and cross-referenced names with public records. They traced that to a group of relatives to a Utah pioneer community with roots in the Mormon Church of Jesus Christ of Latter-day Saints. “When you work on someone’s genealogy, you get to know them in a really intimate way through their ancestors,” Lee said. “You know you’re getting close when they start looking like who you’re looking for.”

The couple soon discovered Pillar Point’s cousins belonged to an intermarried clan of families. Some men had multiple wives and children, amounting to a genealogy headache. “It’s a problem that’s common in isolated religious communities. The fancy word is endogamy,” Lee said. “You end up with a  whole lot of half-relations and unreliable predictions.”

Setbacks

Then a fiasco unfolded. The arrest of the Golden State Killer in 2018 had sparked privacy fears from critics who claimed GEDMatch could be used for nefarious reasons. When a criminal case centering on a minor stirred up more controversy on the site the next year, the owners abruptly purged the “law enforcement matching” section of it in May 2019, according to the Redgraves.

With no warning, it left the couple with only about 20 percent of the genetic clues they’d had before. “Think of it as the number of letters turned around on your ‘Wheel of Fortune’ puzzle. [Afterwards] there were 80 percent less letters, and we still had to guess the phrase,” Lee said.

The setback forced them to get creative. They sought uploads from people who descended from early settlers in Utah, along with the Mormon church, and mapped out “clusters” of potential relatives.

Anthony spent hours tinkering with DNA Painter, a tool that helps genealogists make sense of matches. Eventually, it led to Pillar Point’s possible great grandfather. 

During an all-nighter in October 2019, they had a breakthrough. When they got to one of the possible great grandfather’s relatives, they checked records for proof of the person’s life after 1983, and found none. Lee pulled out Pillar Point’s crime scene photo and checked it against a high school yearbook photo of the grandchild.  

It all added up:  Here was the long-forgotten face of Pillar Point Doe.

They both burst into tears. “There were periods of crying and shaking for a few days afterwards. It was really intense,” Lee said.

The team then sent Pillar Point’s birth name to cops, who collected DNA from a relative to confirm the match, reinvigorating the investigation.

San Mateo County police have since declined to release Pillar Point Doe’s birth name—or to allow the couple to—saying it could hurt their hunt for the killer. “This homicide is actively being investigated. Unfortunately, disclosing information about the details may hinder our investigation,” Sergeant William Young, from the San Mateo County Sheriff’s Office, told Motherboard.

Cold Case, Close To Home

Now, the Redgraves want more answers. “Ideally, police will find the perpetrator,” Lee said. “[Cops] definitely want to tell us something but they can’t. It makes us feel hopeful.”

Not long ago, Lee got a tattoo of poppies in Pillar Point Doe’s honor. It was inspired by the California flower bloom that could be seen from space in March 2019, the week the couple began searching for the victim's identity. “I am absolutely forever changed from working on this case,” Lee said.

Ultimately, the Redgraves hope Pillar Point Doe will be remembered for who they were—a complex and loved person, not a forgotten John Doe. “Hopefully someone who loved them will carry on their memory,” Lee said.

The couple now runs the Trans Doe Task Force, a research group that helps police and medical examiners with transgender and gender-expansive cold cases. Recently, they launched a database that allows for DNA comparisons across sex marker categories. They also founded their own firm, Redgrave Research Forensic Services, and Anthony has helped train law enforcement departments on five continents.

These days, the couple has a small framed high school photo of Pillar Point Doe in their home, near portraits of other people from cases close to their hearts. 

“Pillar Point has become part of our family. I feel like we are basically like their  foster parents,” Anthony said. “I’m going to feel that way until I know exactly how this case ends.”

Meet The Genetic Sleuths Solving Decades-Old Trans Murders syndicated from https://triviaqaweb.wordpress.com/feed/

(AngelList Radio)

In the second episode of AngelList Radio, we interview Sumon Sadhu and learn how to improve as an angel investor and why he thinks “national empires” are good investment opportunities.

Sumon explains why he’s excited about investing in international startups like Mobius Motors (an African car company), Chaldal (a Bangladeshi delivery company), and ClearTax (a Indian electronic tax filing company).

Here are a few of my favorite lessons from this interview:

Now is the time to build big international startups. Billions of people are entering the middle class or coming online for the first time. New startups that serve their local communities can build important companies, especially in large countries like India and China. Sumon discusses examples like Alibaba, and talks in depth about his investments in ClearTax (India) and Chaldal (Bangladesh).

Academic research can help you predict technology trends. Academic publications show us what is becoming technically possible. If you combine that with new behaviours you see in the world, you can use research that is five years old to predict new trends.

Be a signal. As a contrarian, Sumon often has to convince other investors to consider companies that they normally wouldn’t. Fast forward to 31:00 to hear how Sumon did this with Chaldal.

Share your favorite part by leaving a comment on Soundcloud.

Learn more about Sumon’s investments on AngelList, follow him on Twitter and read his blog.

Read on for the full transcript of the interview.

Tyler Willis: Hi, everybody. I’m Tyler Willis. I am an entrepreneur and an angel investor. This year I’m your host for season one of “AngelList Radio.”

[0:17] In our first season we’re interviewing different investors. We’re learning how they invest and what’s made them successful.

[0:23] Today I’m joined by Sumon Sadhu, the co-founder of Snaptalent and the first employee of Quid, which had raised $49 million from top investors. Most recently he co-founded Muse.

[0:32] Sumon is also an angel investor. In talking with him, I’ve really enjoyed how defined, well reasoned, and unique his thought process is. In a business where it pays to be contrarian and right, this is usually a signal worth paying attention to. I’m excited to hear what he has to say today.

[0:47] Sumon, thanks for joining us.

Sumon Sadhu: [0:48] Thank you, Tyler, for having me.

Tyler: [0:50] Before we get started and dive into it, give us an overview of what your career has been like today. How do you find yourself here?

Sumon: [0:58] I’m originally from the UK, originally from Northwest London. I studied biochemistry and infectious disease, both at Oxford and in Imperial College, London, so by training, I’m a scientist.

[1:12] My original sort of motivation for going into that was, I was really fascinated with how nature works. Back in 2000, when the human genome emerged, was really a time where there was exponential knowledge being created in biology relative to other sciences, so I made a bet that that would be a good place to be.

[1:30] As any sort of good student does, I was originally really focused on research, and managed to sort of publish a paper during undergrad and sort of get myself over to an internship in France.

[1:41] It was there in France where a professor that I was working with said to me, “For a scientist, you talk about business in a really sort of interesting way, in that you seem to understand,” we’re talking about stocks. I was analyzing a company. The way that I was describing it to him was certainly different to how he, any previous student has described to him.

[2:01] I went back to Oxford, sort of final year of university, and ended up joining up a society called Oxford Entrepreneurs, which was the student entrepreneurship society at the time, and attempted to start a company.

[2:16] My first ever sort of venture was trying to build software to bring together…there’s a lot of latent knowledge that exists inside of a student population, so how do you turn that knowledge into something, so conceived of something that was sort of Yahoo Answers back in 2005, in the UK.

[2:34] That was really my first experience in trying to put together a project. Obviously, didn’t work, programmers walked out on me, ended up finding a bunch of people to come together, but what it left me was with this desire to sort of turn ideas into projects.

[2:52] That got me started practically in entrepreneurship, and then the summer between Oxford and going to grad school at Imperial College London

[3:00] At that point I didn’t know what I wanted to become but finally I got this glimpse of entrepreneurship because of the Student Entrepreneurship Society. I ended up…again, this was part of the telephone campaign, raising money and did pretty well. I was one of the top callers.

[3:17] Ended up swinging another internship over the summer where I was part of a team that was responsible for performing the first census study of venture capital spending in the UK. My job as an aspiring entrepreneur was to call up every single entrepreneur in the UK, and try and figure out how much they raised and what their companies did.

[3:38] I learned a lot about how to analyze venture back in that summer between university and grad school. The two strands of analysis, as well as practical entrepreneurship, are two things which have combined over time to lead to investing.

[3:54] I went to London. I went to Imperial College London. My goal there was not to be a PhD student. It was actually, really, to spin out the society that was in Oxford to Imperial College London.

[4:09] I attempted that, then I had a really massive setback. I paralyzed my right arm. That was a pretty critical injury. For six months I had to recover from that. What happened was that I was either going to start the society or I was going to be considered a failure.

[4:26] My friend Bob, who earlier had gone to Silicon Valley. He was headhunted by Max Levchin. He was the first employee of Yelp. He wrote me an email, actually. That email was pretty interesting. It said, “If you don’t make this thing real then you’re going to be considered someone who talks, who doesn’t act. You’re going to leave yourself all these splendid starts.”

[4:46] After this injury, I was determined to make this happen. I ended up starting Imperial Entrepreneurs, which was a spinout of Oxford Entrepreneurs at Imperial College London. Imperial College London at the time, 2005-2006, didn’t really have a culture of entrepreneurship. The goal was this is the MIT of Europe. How can we build really great companies out of here?

[5:09] Through that initiative was really great to meet some entrepreneurs who had gone to Imperial. I had to raise money for the student society. I ended up pitching Index for money, for £1,000. Can you give me a £1,000 to start the society?

[5:26] Got some money from Michael Birch, who was an Imperial alum at the time who started Bebo, is in Silicon Valley. Then also at the time I met Saul Klein who was just leaving Video Island and starting up as a partner at Index.

[5:42] When meeting Saul and pitching him this vision for a student society and sparking entrepreneurship in Europe, he said to me, “Hey, there’s an initiative that I’m starting up which your counsel might be very useful for.”

[5:55] That was Seedcamp. Here I was, 22 years old in London, organizing these events, bringing together entrepreneurs from outside and the student population, and I ended up getting to know a lot of the European venture capital community.

[6:12] At that time, Imperial Entrepreneurs was my startup, and then I got to sit on the advisory board of Seedcamp, help to start that. When Seedcamp needed space to start up, we used our student connections to get some space at Imperial, and managed to help build their first website with some of my later co-founders in the startup, and that was a really great experience.

[6:37] The connection that’s interesting is being on the Seedcamp website actually alerted another major incubator at the time, which was Y Combinator. Obviously my aspiration was not really to be a PhD student, so I was getting by, doing enough work, but at the same time, leaving the lab to go do Seedcamp stuff or do an interview at the BBC about entrepreneurship in Europe.

[7:02] Came to San Francisco over a summer with a couple of friends, and managed to get an email in my inbox from Paul Graham, of all people. He’d funded a few friends of mine from the UK, and he said to me, “Hey, we were looking for founders that were promising, and your name appeared familiar. I recognized you from the Seedcamp website. Would you consider pitching me in California?”

[7:28] That email was the opportunity of a lifetime. With a couple of friends, I ended up putting together a prototype over the summer, and we were accepted early, during the summer of ‘07. I went straight back to the UK, quit the PhD, and that started the entrepreneurial journey into Y Combinator.

[7:52] Until that point, there had been a bunch of key inflection points, and then I came to the Valley in 2007. We were staying in the Y-scraper, which obviously was the cheapest place to rent at the time in North Beach. It was great.

[8:11] Some of the first people I met in San Francisco were the founders of Scribd, the founders of Weebly, the founders of Dropbox. They were all in that building.

Tyler: [8:20] I’d not heard that before, “The Y-scraper.” Is this where all the early Y Combinator founders were renting?

Sumon: [8:24] Yeah. The reason why it was quite popular is because it was one of the places where if you didn’t have a credit history, you didn’t have any kind of financial backing, it was somewhere that you could rent that was fully furnished, month to month, and also at the time, it happened to be that a few other people were there, and so the secret was out.

[8:45] It wasn’t so popular that everyone was going to do it, and really for us, we had no option. I was a PhD student. I wasn’t very good at working for other people. I’d always had this aspiration of turning projects that I thought of into companies, and so this one opportunity to come out to California and be funded was a great one.

[9:10] We managed to bridge ourselves between getting accepted into Y Combinator. Back then, funding was you got around $17,000. It was me, Tim, Jamie. Jamie was from the states, so he didn’t have to go back and forth, but me and Tim had to fly back and forth. Just before YC, he brought on a fourth guy called Brad, from MIT. We met on “Hacker News.”

[9:37] We were four guys in Y Combinator in the winter of '08, and on $17K, so we had to make it last, and we had to make money. What happened was that the idea we applied with was not the idea that we executed on. We’d been working a lot on understanding student employment, and we also reasoned that the way that students behaved online were very different to how ordinary consumers were behaving online.

[10:06] The base idea behind Snap Talent, which is the company I started in winter '08, was, “How do you distribute recruitment advertising to the sites that people already spend their time, not on job sites?”

[10:19] We conceived a distributed recruitment ad network. we changed the formats of those ads from text to rich media. We built novel targeting mechanisms, so if you were coming from a Google IP address, how could I show you an ad that said, “Leave Google. Join Microsoft”?

[10:36] That was the basis of what Snap Talent was. During YC, we signed up 18 companies. We made about $30,000 inside of Y Combinator, and ended up raising $1.5 million after YC, and we really needed to raise that money, because we were running off of a cliff, flights back and forth from the UK.

[10:57] That was my first venture, and within…it truly did accelerate our progress, because four months before that, we were just students in the UK, and five months later, we’d come out with a company that had money and actually had brand-name investors. It was fortunate to work with a bunch of really great investors back then.

[11:19] We were in Zachary’s first fund. We raised money from early Googlers like Andrea Zurek and Aiden Sankkt, Paul Buchheit. We raised money from Index, at the time, and Jeff Clavier was also an investor in the company, and Saul, who probably said Matt through the course of this innovation work, back in the UK, and came to back Snap Talent.

[11:45] We put together a really all-star set of investors around the company, and so Snap Talent was an interesting experience, because it didn’t go right. I managed to learn a bunch of lessons from Snap Talent. One is of market timing. First of all, what we were conceiving was incredibly future-looking relative to the state of the market. It also happened to be prior to the recession.

[12:14] Also, I think another thing is we probably made every entrepreneurial mistake in the book. We did very well in raising money. We then had to go straight back to the UK, because of visas, and so keeping momentum of the company up post funding was challenging.

[12:30] The second thing is, internally, while doing YC, we were pretty together. There were certainly moments in the company where we weren’t, and I think learnt a lot about who you work with, why you work with them, and really, the way that you know your cofounders is in crisis, not in actual, in glory.

[12:51] The short lesson was, we built four or five different products inside of Snap Talent. It wasn’t a very long period, but in a year and a half, superstar backers get impatient. We ended up with a board ultimatum, which was, “Shut down the company or get viral,” which is a very incredible request, if you consider it.

[13:13] We had over a million dollars, $1.5 million in the bank, and what we decided, instead of pivoting again, was to cut our losses and return the money back to investors. That was a moment of crisis, a very difficult decision as a CEO to go do that to your first company. That led to the next chapter, looking for the next thing.

Tyler: [13:41] Was it right after that that you joined Quid, or was there some space in between?

Sumon: [13:46] I pretty much jumped in as soon as I really to put my head down, and get back into a project, whether it’s my project or a friend’s project, and it happened to be that Bob and I were really close. We’d always advised each other on each other’s projects, and the two things that they were doing at the time were fascinating to me.

[14:04] One, because they were analyzing networks of information and trying to pull out insights, and two, they were doing that for the understanding of the venture-backed ecosystem, and I’d done both of those things. I’d both looked at networks as a bio-informatician, and I was able to analyze. I had learned the fundamentals of analyzing venture capital.

[14:25] A network of venture-backed companies was probably foreign to someone who didn’t understand venture capital and some of the trends that you could pull out from that data, and so we built a platform that effectively reduced the number of hours that humans had to both read as well as connect ideas, and ultimately that led to a system that could answer many, many valuable questions, leading to the automation of knowledge and knowledge industries.

[14:55] A lot of the problems that Quid software solves is where the question that you are asking does not have a concrete end answer, so, “What is the future of the electric car? What is the state of technology in China? What are the ways in which cloud computing will completely impact the enterprise?”

[15:17] It was through both building that system to look at data and understand connections from natural language as well as apply that in industry, and then eventually go and scale the application of that across many, many different clients. We both built an impressive system that actually, today, we do many secret projects that power many important decisions, as well as I built a massive knowledge base of understanding technology from first principles.

Tyler: [15:45] How do you, Sumon…? Walk me through one of those areas, maybe the electric car, or cloud computing, or something like that. When you’re setting your mind towards analyzing a trend, how do you discover the trend, and then what’s the next step from there?

Sumon: [15:57] You can look at technology trends from a set of fundamentals. One thing is first, what are the enabling waves that come together in order to make new things possible? Those technology waves could be in the case of the electric car, could be the fact that batteries are becoming smaller and more powerful.

[16:18] In the case, from a macro perspective, you’re looking at things like reducing our dependence on fossil fuels, because there’s immense pressure on natural reserves. For all technology possibilities, it comes from, initially, technology trends that are four to five years out of research, which can come together to enable new possibilities.

[16:45] Then you’re looking for needs which exist in the world, which intersect with that. so in the case of transportation, it could be that the miniaturization of batteries and the increasing power density of those batteries enable other forms of transportation beyond large vehicles, and it could be that smaller vehicles like the boosted Ford are new forms of transportation that fit into novel needs. Urbanization creates a need for transportation that lives alongside you.

[17:20] You see all sorts…it’s the intersection between societal behavioral change, because ultimately, if you want to build a technology, it’s got to be used by a lot of people, and technology possibility. Those two things coming together is what creates the opportunity for timing as well as possibility.

[17:36] You can anticipate a lot of those things by watching those things very closely.

Tyler: [17:41] Give me some specifics there, of when you’re watching very closely. Where are you looking for sources of data that you’re starting to pull out. Are you following specific blogs or information sources?

Sumon: [17:51] If you’re trying to understand technology from first principles, you really do have to understand research, and so what’s interesting is a lot of great ideas are already described in literature. There’s no surprises in trying to understand the next big thing to be commercialized, because it’s already been described probably 10, 20 years ago.

[18:14] Naturally, there’s a big treasure trove of information in academic papers in understanding any fundamental technology trend. For example, if you, as a biochemist, have been following synthetic biology since it’s inception – it’s something I wanted to do as an undergraduate, and obviously, you see the iGEM competition and how it arises.

[18:37] It wasn’t around when I finished my studies, but 10 years later, in 2010, you start looking into this stuff, and you see the understanding of the space increase.

[18:49] We have funds-submit enabling technologies. We can read DNA. We can start to write DNA. We can start to understand that we have a bunch more experiments as to how you can recompose those elements to bring novel changes to cells. You could start to see some of it bubble up in 2010, 2011. I actually crashed a synthetic biology conference, Synberc, in Berkley, pretending I was an on-leave PhD student.

[19:16] You can start to see a lot of these big ideas bubble up in academia, being very well described, where the application’s very well described.

Tyler: [19:25] Your thesis is that that often takes several years before the commercial application. You’ll start seeing it in the research and in the academic pursuits?

Sumon: [19:34] Yeah, you’ll see the really crazy ideas be…anything which was the vogue of research, probably four or five years ago, is something which should be commercialized today, and so you should quite heavily looking at those ideas, those algorithms, those opportunities, as something that is commercializable today, because there will be follow-up worked to make them robust.

[20:00] Then there’s the difference between server…the way that I think about new ideas is several fold, and this is inspired by John Maeda, who used to be a professor of design at RISD, and later on joined Kleiner Perkins. He says innovation comes from the thrust of four different disciplines – art, which questions, and questions without needing answers. You just question.

[20:27] The second thing is science, which obviously provides a foundation for possibility, engineering, which turns possibility into reality, and then sales and marketing, which ultimately drives things which become possible to things which become usable, and then mass market.

[20:46] Whenever you look at new ideas, there’s the artistic phase, which is, “This is crazy, and it has no purpose, but it’s worth asking the question why it doesn’t exist.” Then, as you go through thinking about new ideas, whether it’s from the perspective of inventing new ideas or from the perspective of investing in new ideas and assessing where we are along those things, you can put on a different hat and create a different lens for what you’re looking for.

Tyler: [21:16] Let’s step back really quickly. When you think about yourself as an investor, what kind of investor are you? It sounds like you’re obviously very data-driven and very analytical. You’re looking at particular spaces.

[21:28] How do you think of yourself in the universe of other angel investors? What makes you unique, what makes you different, and where are you aligned with a stereotype if there is one?

Sumon: [21:37] A lot of my influences in investing come from very classic things. I’m very influenced by the way that the Medici family in the Renaissance were able to cultivate a set of talent in order to push forward. One is, it’s very people-centric, in terms of I believe that if you can align with people’s fundamental motivations to do something, and you can enable them, then you empower them to create new things.

[22:14] Capital drives the permission to do new things, and so that’s one place, is in influencing people who typically have strong aspirations to invest, to be psychologically confident to make it to the next level.

[22:29] On the other end, it’s combining that with a very rational understanding of market timing, and so from that, the influences are really sort of the classic investing mode to Don Valentine at Sequoia. Market timing is everything, the size of the market, the confluence of factors that drive the market is absolutely everything.

[22:52] When you can combine both, when you can align the psychology of people’s desires to be great, and you can cultivate that together with aligning with market timing, you get great investments.

[23:04] In terms of where I fit into the spectrum of angel investors, I’m investing in things that are timeless, from a thematic perspective, because when you eliminate time, you end up in thinking about things in a very different way. You don’t follow trends.

[23:21] I’m very passionate about investing in for societal needs, and this drives both the projects I am involved with as well as the ones that I invest in, and I don’t look at the two things as being separate. There are some things that I can do myself in terms of companies that I could start, and there’s some things where I should be supporting other individuals, because I believe passionately in those ideals.

[23:42] Those four things that I believe in is – and these are fundamental societal needs that should exist for a very long time, are one, we should make economies more efficient, whether they’re physical or digital economies. We should create economies that create new jobs, so changing the efficiency of economies is one big thing.

[24:04] The second big thing that I’m investing in is the extension of human life span, whether it’s direct or indirect, sort of enablers. I can talk about examples of companies in these themes as well.

[24:15] The third thing is amplification of human intelligence. We live in a complex world. We seek to understand that world in order to make decisions for our survival, for our survival on this planet, for our survival on other planets. We need to extend the capabilities of the human mind to address that complexity.

[24:35] Then the fourth thing that I believe in is this throwback to empires. We’ve sort of forgotten about empiric thinking, and investing in entrepreneurs that want to build empires, particularly national empires, is something I’m pretty passionate about, because those companies can attract and monopolize talent and their local markets, and really build something empiric. The ambition that those entrepreneurs have is fascinating.

Tyler: [25:00] I want to talk about each of those separately, but the kind of building an empire, especially building a national empire piece is particularly interesting because it’s very unique. I don’t think a lot of people think about that today.

[25:11] You see some of these empires being built. The ones in China have gotten some attention recently. What are the other areas where you think national empires can be built? What are you evaluating when you’re looking for a potential national empire?

Sumon: [25:23] Obviously, the first thing is there could be a massive national market, or it could be a massive market where a national focus gives you real emphasis. For example, one of my investments is a company called ClearTax. ClearTax is essentially building the Intuit of India, and every single year, people do their e-filings through ClearTax.

[25:54] This year should be around three percent of e-filings in India go through ClearTax. Next year, 10 percent, and this is an example of an opportunity where an economy is going through a transition from having nothing to going digital. That is an opportunity in the history of that country for a company to enter and take advantage of that shift, but the fact that it’s starting to build a transactional relationship with so many Indians, over a very long period of time, leads to many other opportunities to digitize, to change the behavior of those companies.

[26:28] That’s an example of this historical shift where, for the first time, you can do x, and it seems very obvious to everyone sitting in the US or in the UK, but you get a chance to take that thing and have it become something completely different over time. That’s very exciting.

Tyler: [26:48] There’s a saying that I’ve heard from another investor that I’ve stolen for myself, which is ideally you find somebody where their initial market is very obviously massive, and then there’s optionality towards much larger markets after that.

[27:04] There’s a clear path to a billion dollars, but then if you’re really looking to find the company that’s a $10, $50, $100, multi-$100 billion company, they’re going to have to venture out of their core business at some point, and so you’re looking for somebody who has both a clear initial business as well as then great future optionality.

[27:23] Is that a fair way to say that a lot of these companies that you’re looking at, because they have a monopoly focus internal to one country, they have a lot of optionality to go elsewhere in that country?

Sumon: [27:34] That’s fair to say that, but what’s also true about these entrepreneurs is, because they’re looking at the problem for the first time, they inherently don’t have any limitations to their thinking. They’re not checking themselves because there’s competition left and right, and they’re veering in one direction. It just means that they’re very free about how they’re thinking about things, and that’s a very different way of thinking about an opportunity.

[27:59] When there’s nothing, and you’re the only guy that can get capital to address that opportunity, you can do anything. Then the choice becomes what do you do, and why do you do that? Those are the choices that I help those entrepreneurs make, and I remind them. It’s like we often check ourselves in the US, because there’s someone doing something on the left, and someone doing something on the right.

[28:19] It’s very strange to find yourself in a place where no one’s doing anything, where if you do it, everyone will follow you, and everyone will work for you. That’s kind of an interesting opportunity that many of these emerging markets entrepreneurs have. From a fundamentals perspective, they’re very large opportunities, so we see that as a chance to make a company worth $1 billion, but there’s also a chance to make a company worth $100 billion.They’re not focused on the financial outcome. They’re focused on making something for their country. They’re focused on, for example, creating a type of work that their country doesn’t have, and also, they have no limitations. Often, the conversations with these guys are very crazy.

[29:02] For example, my investment in Chaldal, which is the Amazon of Bangladesh, their motivation for starting that company was a very cheeky remark, which is like, “There’s hundreds of millions of people with money. How do we take it?” You wouldn’t make that statement if you were starting an e-commerce company based out of New York.

[29:24] It would be much more reserved, and so, I think there’s just a different mindset, which if you can cultivate and tune that ambition in those entrepreneurs, they really end up going for something incredible.

Tyler: [29:35] One of the first things we, basically the way we met, was around this Chaldal investment. Is it OK if we tell the story there? Is that…?

Sumon: [29:42] Yeah.

Tyler: [29:45] You organized an angel list fund to invest your XY Combinator founder to invest in Y Combinator startups for that particular batch, and I thought this was a couple of things I wanted to dive into. Chaldal, but I also want to talk about the process used for evaluating companies.

[30:02] One of the things that actually made my ears perk up was like, “Oh, I want to know that guy, and I want to learn more about how he sees the world and what he’s going,” is the fact that you didn’t take the safe route. You didn’t raise a bunch of money to deploy into a batch of 80 or 90 companies, and then really diversify. You didn’t go for 15 investments or 20 investments.

[30:22] You put the bulk of it into one company, into Chaldal. Tell me a little bit about how you made that decision, because obviously diversification is widely regarded as, “Oh, it’s the safe strategy, and it’s what everyone should do.” Why did you buck that trend?

Sumon: [30:37] I do believe in an all winners strategy, and in evaluating the Y Combinator batch, I did say to myself…I’m interested in returns, and I’m interested in disproportionate returns, and ultimately, that’s how you get remembered as an investor. That’s how you make money for your LPs.

[31:00] My focus there was simply the thinking around, “This company has a national economy to dominate, and if we’re able to appropriately capitalize this company, I give enough of a signal to other investors, then they will disproportionately capitalize in this funding round relative to their competitors, therefore giving them a significant advantage in doing this.”

[31:26] That doesn’t necessitate a $25K investment. That involves being the lead signal for the round. Obviously, it wasn’t a decision that was made in a day, but it was fairly quick, from meeting the entrepreneur to doing several days of diligence, really trying to understand the national opportunity, the risks, the psychology of the entrepreneur.

[31:53] It felt like supporting this entrepreneur…and the other thing about the company is, the company was making significant money on a monthly, recurring basis, so objectively, you’re investing in something that returns your capital in cash.

[32:17] Unobjectively, you’re investing with the upside of if you owned a piece of a company that had…Bangladesh is the eighth largest country in the world, and was the number one transactional commerce provider. What else gets built off of that platform? That was the upside, and so I felt that it was a better way of turning $400 into $40 million, $50 million, than other bets.

Tyler: [32:47] You talk about three things there, in those days of diligence. You’re analyzing the opportunity, the risks, and then the entrepreneurs’ psychology. Let’s walk through to those. How do you evaluate an opportunity? What are the things you look for in upside potential? What are some of the questions you ask, or things, your criteria you look for?

Sumon: [33:05] Proxy dollars don’t really go away. Even in the most obscure opportunities, there’s always some proxy spend, that there’s always dollars for displacement. In the case of Chaldal, there is a share of wallet of the Bangladeshi middle class, and that money is going to be spent in some way or another, so there’s proxy dollars at stake.

[33:31] The other thing is that a lot of the risks have been taken out as to whether this company was one that could take money from those individuals. They had already built a transactional relationship which was growing at a significant rate per week.

[33:47] Just to get some of these things off the ground is often difficult. You’ve got to master the logistics, you’ve got to master marketing. You’ve got to actually brand the thing. You’ve got to get it out there. They were on a path where they could be profitable by the end of the year, and so a lot of that risk had been taken out.

[34:02] The third thing is, is the entrepreneur crazy, because he could just stop at being an e-commerce business, and it happened to be that we’re seeing from Chaldal. There’s somewhat an interesting aspect of the psychology of the entrepreneur in that he was a former colleague of Parker and Laks at Zenefits. They’d all come out of SigFig, which is a really interesting company in that it’s created founders that have gone on to do really great things.

[34:28] In his peer group of friends, his immediate influences are to build something massive, because they feel…there was a psychological aspect where one, this was a guy that wasn’t normal. He was slightly crazy. Two, he was in a peer group of founders that were building big things, and he clearly wanted to do that, and he didn’t want to stop.

[34:53] Their original motivation for building that company is to build the Internet company of Bangladesh so that they could create the types of opportunities that they experienced in Silicon Valley in their home country, and that was a better way to do economic development than just throw money at things. That felt like someone who could go a long way.

Tyler: [35:12] You said there a thing that I think was really critical, which is slightly crazy, and you said that as a good thing, where I think a lot of people would say, “Well, is the entrepreneur crazy? If so, I don’t want to back that person.” You actually look for that as a signal. What are you looking for in the right type of crazy?

Sumon: [35:29] It’s really a scale of ambition. Scale of ambition means talking about something that takes many steps to execute, but talking about it in a very confident way. If you’re willing to entertain the possibilities that will take 20, 30, 40 steps of execution, then you’re, as far as I’m concerned, the right level of ambition, because we can figure out how to descope it, how to make it more tractable, and ultimately most entrepreneurs do that.

[36:01] One of the differences is the scale at which people talk about their ideas. There’s some absolute thing that’s really driving them, and most people will check their ambition. Checking your ambition is a sign of not being as determined or as crazy as you’d like. I like those entrepreneurs, because I can spur them to think at that level.

[36:23] That’s another reason why people are interested in working with me, is because that’s the level that I like to participate in in companies.

Tyler: [36:32] I want to take you back to the early days, let’s say your first investment that you made. When was that, and were you at Snap Talent at the time? Were you at Quid? Were you already past that?

Sumon: [36:43] I was at Quid. It was a year or two before I left. It was 2013, 2014.

Tyler: [36:55] What was that company?

Sumon: [36:56] Zesty.

Tyler: [36:57] Tell me a little bit about Zesty. They do food delivery, right?

Sumon: [37:00] They do food delivery. The original reason for investing…I was actually very nervous in making that investment. I was very nervous in making that investment, because it involved one factor, which is how determined was David Langer, who was a friend that I know from Oxford University, in building that business?

[37:21] He had previously started a company in the UK, which he was not so happy with the success of, and so this was his chance. He’d finally made it to Silicon Valley. This was his chance to do something big, but at the time, the configuration of the business was not the one that you see today.

[37:40] At the time, the idea was very simple. I was a user of their app, and as well as a passionate, enthusiastic supporter of David, and he wanted some help in sequencing his round. This was the first round of the company, very close to the friends and family round, and they needed more capital.

[38:00] One of the things that I was obviously very good at is in understanding how do you position a pitch, and then going and making upstream introductions.

[38:09] I had been doing that for a little while. I was walking out of the room, having helped him put together a spreadsheet of upstream investors, and it just occurred to me that maybe I should really just do something here myself instead of talking about it. It was that very nervous first step in making that investment.

Tyler: [38:28] Did you have money set aside at that point? Did you have a thesis set aside, or was this really just kind of a, “Hey, there feels like something here, and I can afford to take a risk”?

Sumon: [38:37] I had actually seen the connection between nutrition and life extension, and so the fact is that he was really focused on delivering healthy food, and taking a swath of all of the possible meals that one could order from restaurant delivery services. They weren’t optimizing for what is healthy, and my fascination with that was that if you could deliver that at scale, you’d actually be impacting the nutrition of a lot of people, and that would have some meaningful impact on lifespan.

[39:07] That was really the thesis, like, “You’re the anti-McDonald’s, and you’re impacting.” It was an irrational connection of things which I was passionate about that led me to say, “No, this is how you should pitch it. This is the big idea, and this is what I believe in.” I psyched myself up to invest. I’d put aside some money for investing, whether it was in public equities, but didn’t necessarily know that that would be how I would get into angel investing.

Tyler: [39:37] You had set aside some money for investing, but it wasn’t clear to you yet that that would be private tech companies?

Sumon: [39:42] Once you make that first investment, and you start building that inductive feel for how to do it, it’s very scary, and I was certainly scared. I was like, “Well, I’m probably going to lose all of this money.” Fortunately, he pulled off a $17 million Series B, Series A with Index and Founders Fund recently, again building on that thesis of extending human lifespan through food.

[40:11] You could tell he would get that. Food is a big business, so a lot of space.

Tyler: [40:17] Do you think about, once you had made that first investment, were you committed to building a portfolio? Do you think about building a portfolio as a way of mitigating risk? How do you think of constructing where you invest?

Sumon: [40:31] I actually made two, three other investments in quick succession after I had made that investment, and the reason was, again, it gave me conviction to back some of the things I believe. When I look at investing – we talked about the four things before – it’s like, “Can we extend human lifespan? Can we expand the efficiency of economies? Can we amplify human intelligence? Can we support these national monopolies?”

[41:03] I believe them regardless of whether I’m investing or founding companies, and so it was more that I had belief in those areas, and as a consequence, I was able to make investments that fit those areas. Another investment was Benchling. Benchling at the time no one wanted to back, because they didn’t believe that cloud software for life sciences was a market.

[41:25] Even people in the industry had said, “It’s not a market,” but as a scientist, as someone who had said, “There’s a lot of people creating a lot of GDP in life sciences. If you can change the efficiency of that, that’s a really valuable thing that you could to do the world. What would happen if you changed the efficiency by one percent, two percent, three percent?” That’s the big idea. I believed in that idea, and so I backed them to pursue that.

[41:48] These are all about the founders. The founders are incredibly determined.

Tyler: [41:52] Do you think about downstream effects? Do you look at something and say, “Well, shoot, I’m investing early, and I don’t know that there’s going to be a lot of people behind me,” especially in the case of a Benchling, where it’s not clear that a Series A investor would look and say, “Oh, there’s a market there. I should bet.”

[42:09] Are you considering that before you make an investment, or are you trying to prove that by your investment?

Sumon: [42:14] In the case of Benchling, there’s proxy dollars. A significant portion of GDP comes from life science industry, so there’s actual proxy dollars. Those companies do make money, and so have money, but they don’t buy software, because no one had built the relevant software for them to buy.

[42:31] By proxy, I am investing in fundamentals – markets, market timing, the psychology of the entrepreneur, as well as my belief in when those things will intersect correctly. I’m very independent in making those decisions, and quite often, my investment is a signal for others to come in and validate that. Often the thing that I do with entrepreneurs, I change their deck, and I work with them to re-architect how they tell their story.

Tyler: [43:01] Is it fair to say that part of your value-add is then helping them tell that story and show that signal, show the path for the Series A or Series B investor who may be lagging behind?

Sumon: [43:10] Yeah, it’s giving them conviction or early conviction where maybe others don’t give them conviction, and I think once the entrepreneur is sufficiently determined, then that’s wind in their sails. That initial set of belief is really valuable.

Tyler: [43:26] You did three or four investments in fairly rapid succession. Was that too fast, too slow? If you had to go back and do it again, would you have changed anything?

Sumon: [43:35] You sort of invest in waves. You first have to put out a set of bets, like, “What do you believe in?”, because then you don’t have enough inductive experience. You don’t know what a good first meeting feels like. You don’t know what a good set of co-founders feels like. You don’t know what a mistake feels like. You have no range of experience to draw from. You have to pay for that.

[43:57] It’s like college tuition. You’re paying for that. You’re going to pay for that regardless. Maybe if you make money from those companies, that’s fantastic, and that shows some strong conviction, but really at that point, it’s like you’re putting down your college tuition and it’s going to cost you.

[44:14] For me it was really important to make a set of investments, and then to sit back and evaluate before making another set of investments.

Tyler: [44:22] You did three or four, and then how long was the break that you gave after that?

Sumon: [44:26] Finally, what happens is that as soon as some of those companies do well, and your a value add to them, obviously more people approach you. I did my first six investments reasonably quickly, within a year of going, and then I’ve slowed down since then, and really focused on…my pace is actually two to three a year. That’s my actual pace. I’m on a slowdown with that.

Tyler: [45:03] I want to put you in the hot seat of giving advice to somebody listening here. We’ll end up with, obviously, lots of different types of people listening to the podcast, but I have in mind two archetypes of people that are probably listening to this, and they’re right on the cusp of making a decision of whether they should become active angel investors or not.

[45:24] One example of that might be somebody who’s an insider, who’s steeped in technology already, but is figuring out whether they want to risk capital and become an investor. The second archetype is somebody who’s maybe flush with capital, but is unfamiliar with technology. I want to run through each of those with you and get a sense for what your advice would be.

[45:45] For the person who’s an insider, let’s call them an x founder. Maybe they have sold their first company, maybe they have gotten some secondary liquidity, and so they’ve got some capital that they can put to work here.

[45:56] If one of your friends who’s in that scenario called you up and said, “Hey, I’m wondering whether I should or should not be an angel investor,” what are the first things you ask them or direct them towards looking at to help make that decision?

Sumon: [46:08] A lot of good experiences with investing in my life have come because of benevolence. You’re not investing because of your capital, because there’s no amount of capital that could provide value. There are people with way more capital than you, and there are people with less capital but maybe more value add.

[46:28] The first thing is, can you practically make a difference in this entrepreneur’s life, whether it’s through giving them perspective, whether it’s giving them experience, whether it’s protecting them from thinking about things in the wrong way, whether it’s protecting them from distraction.

[46:43] I use “Protection” as a really important word, because a lot of the best investors don’t distract. They’re available in the right place, and they’re incredibly benevolent, and eventually karma pays you back. All of the experiences I’ve had that are positive from Y Combinator onwards have been about benevolence.

[47:02] When you’re making a decision about whether you’re investing in something, if it’s a good investment, typically it will follow a very ambitious person who, often, that’s the only constant that you’re investing in. The idea will change, how they execute it will change, their mind will change about how they do it.

[47:19] But if they’re intention and projection of that intention is to do something big – it’s the right time in their life to really commit to something, and you can be benevolent to help them in some way, then do it, because that is your only constant.

[47:33] If you are evaluating on external circumstances like whether the market’s there or not, whether you believe in this team, whether there’s competition, then you’re evaluating it in the wrong way. Obviously, those factors matter, but very much so in 5 years’ time, in 10 years’ time, as companies get going, the competitors won’t be around, because the determination of that person, the magnitude of their ambition, is going to be the thing that drives them to do something.

[48:01] If you have a connection to that person, if you can help them in a benevolent way, if you can bring them some value, whether it’s protecting them from distraction, or giving them advice at the right stage, then invest, but otherwise, stay out of that.

Tyler: [48:14] I want to, maybe even a little bit more tactically than that, assume that there’s plenty of ways to add value, as a friend, as an advisor, as an investor. If you’re on that fence of deciding, “Should I start writing angel investment checks?”, is there a certain number of companies you have to commit to?

[48:30] Let’s take the tuition example, earlier. How much of a tuition should you expect to pay before you actually have learned the business or understand whether it’s for you or not?

Sumon: [48:39] I think you’re going to spend $50,000 to $100,000 in tuition.

Tyler: [48:45] Those initial investments are $10K, $25K? Is that a rough assessment?

Sumon: [48:50] You should start small and slowly build up, as you build conviction behind your investments. I started with $5K investments to build, and then later on, I went back to those companies and put in more money as I had more conviction. I think that you want to feel like you can lose it. You’re slightly scared of losing this money, but you want to feel like it’s meaningful to you.

[49:13] If you want to make money long term – and I haven’t had any exits, so I obviously can’t talk to that, but a lot of up rounds in the portfolio. I think you should always start small, and you should deploy widely. Like I said, you’re going to pay tuition on your first 5 to 10 investments.

Tyler: [49:36] To the point of tuition, that you were saying earlier, the person who’s coming in who maybe kind of learning tech may also have to pay a larger tuition cost. They’re starting to learn what makes a good investment, what makes a good CEO, what makes a good leader, what makes a good founding team, for example, especially around tech, where they may not be used to evaluating technical risk.

[49:59] Do you see people learning that fairly quickly, or should somebody who’s coming in totally green to set aside a little bit more capital?

Sumon: [50:07] Setting aside capital, and there’s also learning from history. A lot of things that I’ve looked at have been historical investments. The thesis around investing internationally came from looking at Celtel, for example, was a deal that Bessemer did. It was a infrastructure company focused on telecomms in Africa, and if you would invested in that, you would have generated 200x return on your capital.

[50:35] Alibaba was a national monopoly. I think that you can learn a lot from history, and learn a lot by historically understanding the decisions that people have made, but yeah, also I think inductive experience like having that first meeting, doing that first check, getting in a round with other people, feeling what that’s like, you only get that from doing it.

Tyler: [50:57] So you think really the first few are important for building reputation, for building experience, network? What are those things that are interesting to you?

Sumon: [51:04] It’s building a track record of service to entrepreneurs, because those entrepreneurs will talk. When you want to do your third deal or your fourth deal, the first guy that you backed that’s really happy with you will give you a reference to get into that deal, and that’s the stuff that matters, that later on compounds.

[51:21] Focusing on having a really good experience for the entrepreneurs that you back initially, and then using that to compound your involvement, rather than having a very light touch involvement. Your name on the CAP table, it doesn’t matter. You have no relationship that you can draw on to gain leverage from later on.

Tyler: [51:40] I want to switch gears a little bit for the other type of investor, the other archetype. The person who is…perhaps they’re coming from real estate, or they’re coming from financial services, or something separately, where they have a lot of capital. They may be users of technology. They may deeply understand technology, but they’re not steeped in the business. They haven’t been operating a company, or investing in tech companies.

[52:02] What would be your advice to that person who’s sitting on the sidelines and looking at, “Hey, maybe I do want to invest in technology, but I don’t have a clear, deep, competitive advantage across folks that have been operating?”

Sumon: [52:16] That’s two things to think about. That one is maybe there is an angle that you have that is valuable. If you’ve been in real estate, maybe there’s some advantage you can provide beyond capital, or if you’ve been in finance, there’s some rigor that you can add, but for the most part, if you’re completely new to the business, then I think find ways of being involved that are helpful, whether it’s empowering others to invest on your behalf, whether it’s…

[52:48] Again, it comes down to being helpful. If there’s a situation where you have an opportunity to invest, think about how you can be helpful. Quite often, writing a big check and getting an entrepreneur to a certain funding milestone is helpful.

[53:01] In fact, it’s like, beating around the bush, waiting for a check to arrive or asking too many questions often causes disappointment. Maybe there’s a way in which you can get a great reputation by being a guy that’s reliable that ask good questions but was quick and did things.

[53:18] There are ways of differentiating, that’s if you’re doing direct investing and there are opportunities for this indirect investing through value added individuals. Get to know them, reach out to them, talk to them, understand how they invest. Then pick a few of those that you back indirectly, it could be through syndicates. It could be by being an LP in a fund.

[53:38] There are lots of opportunities to participate in the ecosystem beyond that vantage point.

Tyler: [53:46] For those folks, let’s say, was there ever a point where you would say, “Hey, don’t do direct investing”? If somebody came to you, what are the times in which you would tell them as a potential, “Hey, you should really go do LP or you should really go back syndicates instead of trying to do direct investing”?

Sumon: [54:02] Direct investing is incredibly stressful. Quite often, if your relationship with the entrepreneur, you have to be incredibly patient capital, and it requires a sort of understanding of the human situation an entrepreneur’s going through.

[54:17] If you’re not used to those ups and downs, then don’t invest directly. It’s not sort of a good thing. If you’re willing to be patient capital and sort of see things out over a long period of time, and smile through thick or thin, then I would say do it, but in general, it’s not a good thing if you’re not patient in that way.

Tyler: [54:41] Yeah, I think your point earlier about reputation compounding is a double-edged sword. If you have a negative in reputation early on, you get bad selection bias after it.

Sumon: [54:51] Yeah, and quite often, there are some angels that I know that I just known as being the guy that you go to for a few hundred K. Those are the guys, they invest passively, they often interact over email, but they’re known to be reliable. If they want to do a deal, they do a deal. They’re getting into a lot of stuff.

[55:09] Being reliable, being virtuous to an entrepreneur, that’s how you build a stellar reputation for yourself.

Tyler: [55:16] When you’re talking about a reputation, the things that are really critical for an LP, being able to provide some sort of value add, and then being consistent. Is that, those are the kind of the two things that you see as being important?

Sumon: [55:29] You, as a GP or as an LP?

Tyler: [55:31] As a direct investor.

Sumon: [55:34] As a direct investor. Again, I think it’s about, can you impact the levers that drive a company’s sort of positive value? One thing is, in raising financing, do you know other investors that you can bring in? Quite often the conversations that are happening, like, can you help focus the entrepreneur in on a specific way of describing the opportunity that is valuable to them?

[56:05] It might be the central thesis that drives the company or something like that, but can you help focus them on that? When the entrepreneur’s distracted can you bring them back to focus?

[56:15] Also, the other thing is that there’s lots of ups and downs in starting companies. Can you be a source of positive psychology for the entrepreneur? You often see a lot of entrepreneurs – that’s what they’re looking for in terms of service. Can you be the guy that you go to?

[56:34] Quite often in Snap Talent Chris Sacca was the guy that I would call up who was the founder’s friend. Being the founder’s friend is a really valuable position.

Tyler: [56:45] Talk a little bit more about that. I know it’s valuable from the entrepreneurial side, where there are people that I’ve called and gotten great boosts to morale. Something’s gone wrong. I really need somebody to tell me it’s not the end of the world.

[56:58] As an angel investor, though, I think it’s also similarly valuable to be the investor that’s closer to the company, the investor that’s let into the wall a little bit more deeply. Do you see that as well playing out in your role given that you’re really early and really connected to the company?

Sumon: [57:15] If you can accept imperfect circumstances then you’re really valuable. Quite often investors meet companies at inflection points. Those companies are perfect for those inflection points. They have got the right messaging. They’ve got the right team, they’ve got the right funding. But the point is that you know that three weeks before they made up all of that stuff.

[57:34] If you can understand the relationship between the stuff you make up and the stuff that’s perfect, then you’re going to be the guy that sees the stuff that’s imperfect. Quite often that means you’re meeting the companies at a point where they are not refined.

[57:48] If you can assess something that is not refined and you can make it become more refined that’s a very valuable place to be in the ecosystem. The guy that gets the call about the board deck before the board deck goes up is valuable. The guy that sees the pitch before you even have the third cofounder, that guy’s valuable. The guy that can persuade the fifth guy to join, that guy’s valuable.

[58:11] It’s about understanding imperfection. I think that’s a role to play that a lot of investors don’t lurk in.

Tyler: [58:22] There was something that triggered in my head from the very start of this conversation. You described getting the email that changes your life. That’s Paul Graham reaching out and saying, “Hey, would you consider coming and pitching us?”

[58:35] One thing that blows me away, especially seeing how big Y Combinator has gotten today, is the concept that they would actually hunt – they would actually go out and look for talented entrepreneurs or people that possessed the potential that they wanted to tap into.

[58:50] It’s my understanding that even today they still do some of this. Partners are very active in evangelizing YC and trying to find raw talent that might not be convinced to join yet.

[59:01] You’ve been close to YC for a long time. Do you think that’s a core part of their success?

Sumon: [59:08] The two core parts of the success is…I think that there’s an adage that there’s 15 great companies that get built in Silicon Valley every year. Andy Ratcliff talks about the study that he did, comes out of benchmark capital and Marc Andreessen references it.

[59:24] The truth is, how do you make 15 companies that didn’t exist that become those 15 great companies? There are certainly more ideas and more entrepreneurs of unfulfilled potential.

[59:35] That’s the gap that Y Combinator fills. The matching the total idea space for technology together with the total opportunity space for individuals to fulfill their destiny as being great in that.That’s the marketplace opportunity that Y Combinator is looking to fulfill.

[59:59] Being benevolent to those individuals, believing in those individuals before they became big because it was their only opportunity. It was their first opportunity to be believed in. That’s the other place where that is consistent with the values of Y Combinator.

[60:12] Ultimately what that means is an army of angels, an army of founders that are schooled in that way. That ultimately leads to an ecosystem that has those values.

Tyler: [60:28] When you were raising money for your own companies, what were you looking for in your angel investors? When you were talking to the angel investors for the company you just founded what were the criteria that you were looking for?

Sumon: [60:40] Again, I think that when you’re building a company with a long time horizon and you’re building a company with a long technology time horizon, at least at the beginning, you’re looking for investors, again, that are willing to believe in the founders and the capability of the founders to lead you through that uncertain jungle to commercialization.

[61:06] In my experience, that patience was something that very few people had. That was something that I’m looking for because the journey in your first year, in your first two years is going to be very different from when you’ve taken the money, too.

[61:28] I really wanted to find investors that were willing to support that, that believed in me and my cofounder. That’s important.

Tyler: [61:36] When you think about when you started investing and perhaps today, have there been large scale changes? This is only a few years ago that you started. Is it still the same as when you started or are you perceiving changes in people that are just starting to angel invest today?

Sumon: [61:56] What’s true is, like any sort of movement, just line in entrepreneurship, there’s certainly more acceptance of how to do angel investing and how to do it better. There’s also, certainly, more opportunities for angels to fill in the gaps as venture and institutional investing moves upstream.

[62:17] There are certainly more ways in which angel investors can become noticed and build brand. I was fortunate to be hacking with AngelList Syndicates back in 2014. That was because I wanted to scale up my ability to be benevolent to founders, but also as a way in which…getting real inductive experience in investing other people’s money.

[62:41] There are definitely new opportunities to…I think any sort of revolution in product or technology comes when things become more consumer-centric. There’s a consumer need and then there’s a market need. Now there’s more opportunities to bundle capital advice and branding in a way that you can serve the entrepreneur more specifically. That movement has taken off.

Tyler: [63:10] You think there’s an increased ability to bundle those things? I know some of the other people in the industry have said actually this process is about unbundling. You’ll get advice separate from capital. You’ll get angels well before you get venture investing.

[63:26] Do you see this actually as a rebundling across smaller vehicles? Or is this an unbundling that’s happening across the industry?

Sumon: [63:36] What it is is that it’s the relationship between service and those bundles. It’s the opportunity to…

[63:45] For example, if you have those three ingredients, you have X amount of capital, Y amount of brand, and Z amount of service typically you can have more choices for how those things come together. Often they will reflect the optimal make up for an entrepreneur to take them.

[64:05] If the entrepreneur’s the customer there’s going to be more options for them with those three components. You can say, “They’re 10 percent capital and they’re 5 percent brand but they’re 50 percent service and maybe I’ll take more of that than this.”

Tyler: [64:18] How do you see yourself as…what’s the mix-up of your bundle? Maybe how that’s changed over time, 'cause I would assume with a Five K check you’re obviously small amounts of capital.

Sumon: [64:28] That was at the beginning I think. I’m writing much bigger…

[64:31] [crosstalk]

Sumon: [64:31] …checks now.

Tyler: [64:32] Now you’re probably changed the mix of it.

Sumon: [64:35] The difference is obviously I think with institutional support, with some of the work they’ve done with Maiden Lane, some of the work they have done with the YC Funds Project out of AngelList as well as the syndicate means that I see myself competing in the super-angel slot and often using the ability to drive much more capital to create new types of outcomes, whether it’s sequencing rounds, whether it’s pricing rounds.

[65:05] It’s moving upstream. The relationship between service and capital – eventually you’ll have the best service providers indistinguishable from existing capital providers, but because they provide the best service…my goal is to go upstream. I see that emerging with the different tools that are coming out.

Tyler: [65:28] Do you think, long time scale – use whatever time period you want – do you think you will work as a venture capitalist, or as a traditional venture capitalist, at a firm?

Sumon: [65:38] My goal is to pursue my mission, which is again, to further humanity on those four fronts. One way of doing that is obviously driving capital towards the ideas that further those things. Obviously by some reason that will involve controlling the flow of capital. That may involve traditional investing in that sense, whether it’s my own fund or whether it’s part of something else. At some point those things will come together.

[66:10] But my secret goal has been for a long time is to be one of the best investors in the world. I’ve been studying venture for a long time, since 2005. I’ve done my 20,000 hours of really understanding the future of markets and all that stuff. My angel investing activities is part of keeping a lot of that experience going.

Tyler: [66:36] Do you feel competitive with that person who is maybe on the sideline maybe thinking about investing, or is it cooperative? Let’s say I’m an entrepreneur. I’m considering writing my first 10K or 25K check or 5K check. Am I competing with you for access into companies, or is this a cooperative, helpful “We’re all in it together-type economy”?

Sumon: [66:54] It’s cooperative because every single company has different needs and different constellations of stars will come together around those companies. I think that it’s not a thing about competition.

[67:12] I mainly focus on, one, service to entrepreneurs, and two, my mission as an individual on this planet. If those two things are interesting to people, if those two things are complementary then certainly, very open to talking. But I think it’s all about service to the entrepreneur and ultimately that will dictate things.

[67:31] If people are genuinely benevolent, if people are genuinely value-added, if people are genuinely complementary then they’ll find themselves in constellations with other people.

Tyler: [67:41] This is…it’s an industry where the median return is negative – at least venture, and I would assume the same thing extends to angel and some of the studies I’ve seen have shown that maybe it’s less negative. But it’s still a game where being in the middle of the pack means you’ve lost.

Sumon: [67:57] It’s an elitist sport.

Tyler: [67:59] Yeah. Is your thesis that by being truly benevolent, by being truly value-added and focusing on long-term help for the companies that you do rise to the elite level? Or is this that being particularly good at analysis or particularly good at evaluating people is a critical start skill that you need to have and then all the rest of the stuff allows you to play the long term?

Sumon: [68:26] To be a great investor you have to be great at analysis. You have to be able to do a deal. You have to be able to assess people, and you have to have some differentiated view on the world at the time of the investment, because then that’s where you create an advantage. If you have an advantage in all four of those things they compound. As an athlete my focus is obviously to be really good at those four things.

[68:50] Then in terms of regarding how other people invest, if you want to invest, don’t invest casually. Invest seriously. Invest as a sport. Invest in a way where you want to get into the best deals, where you want to provide the best service for entrepreneurs. Then ultimately there will be 10 percent remaining in 10 years’ time that are still doing it.

[69:14] If you look at the cohorts of venture not everyone makes it. Not everyone generates returns. Ultimately I am very benevolent for those entrepreneurs that have massive ambitions, but mainly because by aligning around that you hopefully will generate big returns, which then allows you to get the next cohort of incredibly ambitious people to back again. It is about returns.

Tyler: [69:44] You said you do two or three in a given year? How many companies do you look at in a given year that you actually do analysis on?

Sumon: [69:49] Because of the selective nature of my thesis, I have a lot of inbound pitches, but I’m not actually focused on inbound. I’m focused on outbound. If there’s something that I like that I’m passionate about, I’ll go after it.

Tyler: [70:04] How are you identifying those?

Sumon: [70:06] Again, they fit within the four investment theses that I have. If I talk or write widely about those things, those…

[70:18] There are certain ideas that I believe in and they resonate with certain people. If I’m resonating with people then those entrepreneurs will come through. If there’s something I particularly like I move quite fast and try and be first into that company.

[70:35] In that respect I am competitive. The pattern that I look for is if lots of people are in it I won’t invest in it. For me being first or being misunderstood is important and for it to be confusing to me and confusing to others is also important.

Tyler: [70:53] Why is that?

Sumon: [70:54] Again, if we look at the 15 companies example, if there’s 15 great companies built in Silicon Valley, but another 15 can be created, I think that there’s less competition for the ones that get created because it requires a different skill set to find them, to nurture them, to build those up. That’s what’s driving a lot of angel investments.

Tyler: [71:18] I want to be respectful of your time. You’ve given us a ton. Two quick last questions. The first one is, what is the biggest obvious mistake you made in the first year? What’s an obvious pitfall you can help somebody avoid?

Sumon: [71:33] One thing is to make an investment decision really emotionally. One is to fall in love with the theory of the idea, and then to invest on your interpretation of that. You have to be listening for objective signals and I think it’s really interesting in angel pitches…

[71:54] There’s a company I was involved in as an advisor. I was in love with the theory of what they could build, and ultimately the fundamentals of the company were not correct. That company ended up being acquired, but mainly because acquisition was the only way to get out of that company.

[72:13] I fell in love with what I thought they could build, and that was the thing that made me invest. It wasn’t any of the fundamentals. You do have to remain very objective and unemotional, to some extent.

[72:26] That’s a big mistake in angel investing, is to lead with your dream of what this company becomes and use that to drive your decision, not something that’s much more objective that comes from the company, from the fundamentals of the company.

Tyler: [72:42] You publish thoughts online. You tweet fairly significantly. You blog. If people want to learn more about your thesis and how you view the world where do they find that?

Sumon: [72:52] My website, sumonsadhu.com, has a bunch of my thoughts on there. My twitter handle, twitter.com/sharpshoot has a bunch of my nascent thoughts.

[73:03] [background music]

Tyler: [73:04] Thanks so much for joining us and the time.

Sumon: [73:07] Thank you, Tyler, for having me.

Tyler: [73:08] Cheers.

11 What will break me? This is the question that consumes me over the next three days as we wait to be released from our prison of safety. What will break me into a million pieces so that I am beyond repair, beyond usefulness? I mention it to no one, but it devours my waking hours and weaves itself throughout my nightmares. Four more bunker missiles fall over this period, all massive, all very damaging, but there's no urgency to the attack. The bombs are spread out over the long hours so that just when you think the raid is over, another blast sends shock waves through your guts. It feels more designed to keep us in lockdown than to decimate 13. Cripple the district, yes. Give the people plenty to do to get the place running again. But destroy it? No. Coin was right on that point. You don't destroy what you want to acquire in the future. I assume what they really want, in the short term, is to stop the Airtime Assaults and keep me off the televisions of Panem. We receive next to no information about what is happening. Our screens never come on, and we get only brief audio updates from Coin about the nature of the bombs. Certainly, the war is still being waged, but as to its status, we're in the dark. Inside the bunker, cooperation is the order of the day. We adhere to a strict schedule for meals and bathing, exercise and sleep. Small periods of socialization are granted to alleviate the tedium. Our space becomes very popular because both children and adults have a fascination with Buttercup. He attains celebrity status with his evening game of Crazy Cat. I created this by accident a few years ago, during a winter blackout. You simply wiggle a flashlight beam around on the floor, and Buttercup tries to catch it. I'm petty enough to enjoy it because I think it makes him look stupid. Inexplicably, everyone here thinks he's clever and delightful. I'm even issued a special set of batteries - an enormous waste - to be used for this purpose. The citizens of 13 are truly starved for entertainment. It's on the third night, during our game, that I answer the question eating away at me. Crazy Cat becomes a metaphor for my situation. I am Buttercup. Peeta, the thing I want so badly to secure, is the light. As long as Buttercup feels he has the chance of catching the elusive light under his paws, he's bristling with aggression. (That's how I've been since I left the arena, with Peeta alive.) When the light goes out completely, Buttercup's temporarily distraught and confused, but he recovers and moves on to other things. (That's what would happen if Peeta died.) But the one thing that sends Buttercup into a tailspin is when I leave the light on but put it hopelessly out of his reach, high on the wall, beyond even his jumping skills. He paces below the wall, wails, and can't be comforted or distracted. He's useless until I shut the light off. (That's what Snow is trying to do to me now, only I don't know what form his game takes.) Maybe this realization on my part is all Snow needs. Thinking that Peeta was in his possession and being tortured for rebel information was bad. But thinking that he's being tortured specifically to incapacitate me is unendurable. And it's under the weight of this revelation that I truly begin to break. After Crazy Cat, we're directed to bed. The power's been coming and going; sometimes the lamps burn at full brightness, other times we squint at one another in the brownouts. At bedtime they turn the lamps to near darkness and activate safety lights in each space. Prim, who's decided the walls will hold up, snuggles with Buttercup on the lower bunk. My mother's on the upper. I offer to take a bunk, but they make me keep to the floor mattress since I flail around so much when I'm sleeping. I'm not flailing now, as my muscles are rigid with the tension of holding myself together. The pain over my heart returns, and from it I imagine tiny fissures spreading out into my body. Through my torso, down my arms and legs, over my face, leaving it crisscrossed with cracks. One good jolt of a bunker missile and I could shatter into strange, razor-sharp shards. When the restless, wiggling majority has settled into sleep, I carefully extricate myself from my blanket and tiptoe through the cavern until I find Finnick, feeling for some unspecified reason that he will understand. He sits under the safety light in his space, knotting his rope, not even pretending to rest. As I whisper my discovery of Snow's plan to break me, it dawns on me. This strategy is very old news to Finnick. It's what broke him. "This is what they're doing to you with Annie, isn't it?" I ask. "Well, they didn't arrest her because they thought she'd be a wealth of rebel information," he says. "They know I'd never have risked telling her anything like that. For her own protection." "Oh, Finnick. I'm so sorry," I say. "No, I'm sorry. That I didn't warn you somehow," he tells me. Suddenly, a memory surfaces. I'm strapped to my bed, mad with rage and grief after the rescue. Finnick is trying to console me about Peeta. "They'll figure out he doesn't know anything pretty fast. And they won't kill him if they think they can use him against you." "You did warn me, though. On the hovercraft. Only when you said they'd use Peeta against me, I thought you meant like bait. To lure me into the Capitol somehow," I say. "I shouldn't have said even that. It was too late for it to be of any help to you. Since I hadn't warned you before the Quarter Quell, I should've shut up about how Snow operates." Finnick yanks on the end of his rope, and an intricate knot becomes a straight line again. "It's just that I didn't understand when I met you. After your first Games, I thought the whole romance was an act on your part. We all expected you'd continue that strategy. But it wasn't until Peeta hit the force field and nearly died that I - " Finnick hesitates. I think back to the arena. How I sobbed when Finnick revived Peeta. The quizzical look on Finnick's face. The way he excused my behavior, blaming it on my pretend pregnancy. "That you what?" "That I knew I'd misjudged you. That you do love him. I'm not saying in what way. Maybe you don't know yourself. But anyone paying attention could see how much you care about him," he says gently. Anyone? On Snow's visit before the Victory Tour, he challenged me to erase any doubts of my love for Peeta. "Convince me," Snow said. It seems, under that hot pink sky with Peeta's life in limbo, I finally did. And in doing so, I gave him the weapon he needed to break me. Finnick and I sit for a long time in silence, watching the knots bloom and vanish, before I can ask, "How do you bear it?" Finnick looks at me in disbelief. "I don't, Katniss! Obviously, I don't. I drag myself out of nightmares each morning and find there's no relief in waking." Something in my expression stops him. "Better not to give in to it. It takes ten times as long to put yourself back together as it does to fall apart." Well, he must know. I take a deep breath, forcing myself back into one piece. "The more you can distract yourself, the better," he says. "First thing tomorrow, we'll get you your own rope. Until then, take mine." I spend the rest of the night on my mattress obsessively making knots, holding them up for Buttercup's inspection. If one looks suspicious, he swipes it out of the air and bites it a few times to make sure it's dead. By morning, my fingers are sore, but I'm still holding on. With twenty-four hours of quiet behind us, Coin finally announces we can leave the bunker. Our old quarters have been destroyed by the bombings. Everyone must follow exact directions to their new compartments. We clean our spaces, as directed, and file obediently toward the door. Before I'm halfway there, Boggs appears and pulls me from the line. He signals for Gale and Finnick to join us. People move aside to let us by. Some even smile at me since the Crazy Cat game seems to have made me more lovable. Out the door, up the stairs, down the hall to one of those multidirectional elevators, and finally we arrive at Special Defense. Nothing along our route has been damaged, but we are still very deep. Boggs ushers us into a room virtually identical to Command. Coin, Plutarch, Haymitch, Cressida, and everybody else around the table looks exhausted. Someone has finally broken out the coffee - although I'm sure it's viewed only as an emergency stimulant - and Plutarch has both hands wrapped tightly around his cup as if at any moment it might be taken away. There's no small talk. "We need all four of you suited up and aboveground," says the president. "You have two hours to get footage showing the damage from the bombing, establish that Thirteen's military unit remains not only functional but dominant, and, most important, that the Mockingjay is still alive. Any questions?" "Can we have a coffee?" asks Finnick. Steaming cups are handed out. I stare distastefully at the shiny black liquid, never having been much of a fan of the stuff, but thinking it might help me stay on my feet. Finnick sloshes some cream in my cup and reaches into the sugar bowl. "Want a sugar cube?" he asks in his old seductive voice. That's how we met, with Finnick offering me sugar. Surrounded by horses and chariots, costumed and painted for the crowds, before we were allies. Before I had any idea what made him tick. The memory actually coaxes a smile out of me. "Here, it improves the taste," he says in his real voice, plunking three cubes in my cup. As I turn to go suit up as the Mockingjay, I catch Gale watching me and Finnick unhappily. What now? Does he actually think something's going on between us? Maybe he saw me go to Finnick's last night. I would've passed the Hawthornes' space to get there. I guess that probably rubbed him the wrong way. Me seeking out Finnick's company instead of his. Well, fine. I've got rope burn on my fingers, I can barely hold my eyes open, and a camera crew's waiting for me to do something brilliant. And Snow's got Peeta. Gale can think whatever he wants. In my new Remake Room in Special Defense, my prep team slaps me into my Mockingjay suit, arranges my hair, and applies minimal makeup before my coffee's even cooled. In ten minutes, the cast and crew of the next propos are making the circuitous trek to the outside. I slurp my coffee as we travel, finding that the cream and sugar greatly enhance its flavor. As I knock back the dregs that have settled to the bottom of the cup, I feel a slight buzz start to run through my veins. After climbing a final ladder, Boggs hits a lever that opens a trapdoor. Fresh air rushes in. I take big gulps and for the first time allow myself to feel how much I hated the bunker. We emerge into the woods, and my hands run through the leaves overhead. Some are just starting to turn. "What day is it?" I ask no one in particular. Boggs tells me September begins next week. September. That means Snow has had Peeta in his clutches for five, maybe six weeks. I examine a leaf on my palm and see I'm shaking. I can't will myself to stop. I blame the coffee and try to focus on slowing my breathing, which is far too rapid for my pace. Debris begins to litter the forest floor. We come to our first crater, thirty yards wide and I can't tell how deep. Very. Boggs says anyone on the first ten levels would likely have been killed. We skirt the pit and continue on. "Can you rebuild it?" Gale asks. "Not anytime soon. That one didn't get much. A few backup generators and a poultry farm," says Boggs. "We'll just seal it off." The trees disappear as we enter the area inside the fence. The craters are ringed with a mixture of old and new rubble. Before the bombing, very little of the current 13 was aboveground. A few guard stations. The training area. About a foot of the top floor of our building - where Buttercup's window jutted out - with several feet of steel on top of it. Even that was never meant to withstand more than a superficial attack. "How much of an edge did the boy's warning give you?" asks Haymitch. "About ten minutes before our own systems would've detected the missiles," says Boggs. "But it did help, right?" I ask. I can't bear it if he says no. "Absolutely," Boggs replies. "Civilian evacuation was completed. Seconds count when you're under attack. Ten minutes meant lives saved." Prim, I think. And Gale. They were in the bunker only a couple of minutes before the first missile hit. Peeta might have saved them. Add their names to the list of things I can never stop owing him for. Cressida has the idea to film me in front of the ruins of the old Justice Building, which is something of a joke since the Capitol's been using it as a backdrop for fake news broadcasts for years, to show that the district no longer existed. Now, with the recent attack, the Justice Building sits about ten yards away from the edge of a new crater. As we approach what used to be the grand entrance, Gale points out something and the whole party slows down. I don't know what the problem is at first and then I see the ground strewn with fresh pink and red roses. "Don't touch them!" I yell. "They're for me!" The sickeningly sweet smell hits my nose, and my heart begins to hammer against my chest. So I didn't imagine it. The rose on my dresser. Before me lies Snow's second delivery. Long-stemmed pink and red beauties, the very flowers that decorated the set where Peeta and I performed our post-victory interview. Flowers not meant for one, but for a pair of lovers. I explain to the others as best I can. Upon inspection, they appear to be harmless, if genetically enhanced, flowers. Two dozen roses. Slightly wilted. Most likely dropped after the last bombing. A crew in special suits collects them and carts them away. I feel certain they will find nothing extraordinary in them, though. Snow knows exactly what he's doing to me. It's like having Cinna beaten to a pulp while I watch from my tribute tube. Designed to unhinge me. Like then, I try to rally and fight back. But as Cressida gets Castor and Pollux in place, I feel my anxiety building. I'm so tired, so wired, and so unable to keep my mind on anything but Peeta since I've seen the roses. The coffee was a huge mistake. What I didn't need was a stimulant. My body visibly shakes and I can't seem to catch my breath. After days in the bunker, I'm squinting no matter what direction I turn, and the light hurts. Even in the cool breeze, sweat trickles down my face. "So, what exactly do you need from me again?" I ask. "Just a few quick lines that show you're alive and still fighting," says Cressida. "Okay." I take my position and then I'm staring into the red light. Staring. Staring. "I'm sorry, I've got nothing." Cressida walks up to me. "You feeling okay?" I nod. She pulls a small cloth from her pocket and blots my face. "How about we do the old Q-and-A thing?" "Yeah. That would help, I think." I cross my arms to hide the shaking. Glance at Finnick, who gives me a thumbs-up. But he's looking pretty shaky himself. Cressida's back in position now. "So, Katniss. You've survived the Capitol bombing of Thirteen. How did it compare with what you experienced on the ground in Eight?" "We were so far underground this time, there was no real danger. Thirteen's alive and well and so am - " My voice cuts off in a dry, squeaking sound. "Try the line again," says Cressida. "'Thirteen's alive and well and so am I.'" I take a breath, trying to force air down into my diaphragm. "Thirteen's alive and so - " No, that's wrong. I swear I can still smell those roses. "Katniss, just this one line and you're done today. I promise," says Cressida. "'Thirteen's alive and well and so am I.'" I swing my arms to loosen myself up. Place my fists on my hips. Then drop them to my sides. Saliva's filling my mouth at a ridiculous rate and I feel vomit at the back of my throat. I swallow hard and open my lips so I can get the stupid line out and go hide in the woods and - that's when I start crying. It's impossible to be the Mockingjay. Impossible to complete even this one sentence. Because now I know that everything I say will be directly taken out on Peeta. Result in his torture. But not his death, no, nothing so merciful as that. Snow will ensure that his life is much worse than death. "Cut," I hear Cressida say quietly. "What's wrong with her?" Plutarch says under his breath. "She's figured out how Snow's using Peeta," says Finnick. There's something like a collective sigh of regret from the semicircle of people spread out before me. Because I know this now. Because there will never be a way for me to not know this again. Because, beyond the military disadvantage losing a Mockingjay entails, I am broken. Several sets of arms would embrace me. But in the end, the only person I truly want to comfort me is Haymitch, because he loves Peeta, too. I reach out for him and say something like his name and he's there, holding me and patting my back. "It's okay. It'll be okay, sweetheart." He sits me on a length of broken marble pillar and keeps an arm around me while I sob. "I can't do this anymore," I say. "I know," he says. "All I can think of is - what he's going to do to Peeta - because I'm the Mockingjay!" I get out. "I know." Haymitch's arm tightens around me. "Did you see? How weird he acted? What are they - doing to him?" I'm gasping for air between sobs, but I manage one last phrase. "It's my fault!" And then I cross some line into hysteria and there's a needle in my arm and the world slips away. It must be strong, whatever they shot into me, because it's a full day before I come to. My sleep wasn't peaceful, though. I have the sense of emerging from a world of dark, haunted places where I traveled alone. Haymitch sits in the chair by my bed, his skin waxen, his eyes bloodshot. I remember about Peeta and start to tremble again. Haymitch reaches out and squeezes my shoulder. "It's all right. We're going to try to get Peeta out." "What?" That makes no sense. "Plutarch's sending in a rescue team. He has people on the inside. He thinks we can get Peeta back alive," he says. "Why didn't we before?" I say. "Because it's costly. But everyone agrees this is the thing to do. It's the same choice we made in the arena. To do whatever it takes to keep you going. We can't lose the Mockingjay now. And you can't perform unless you know Snow can't take it out on Peeta." Haymitch offers me a cup. "Here, drink something." I slowly sit up and take a sip of water. "What do you mean, costly?" He shrugs. "Covers will be blown. People may die. But keep in mind that they're dying every day. And it's not just Peeta; we're getting Annie out for Finnick, too." "Where is he?" I ask. "Behind that screen, sleeping his sedative off. He lost it right after we knocked you out," says Haymitch. I smile a little, feel a bit less weak. "Yeah, it was a really excellent shoot. You two cracked up and Boggs left to arrange the mission to get Peeta. We're officially in reruns." "Well, if Boggs is leading it, that's a plus," I say. "Oh, he's on top of it. It was volunteer only, but he pretended not to notice me waving my hand in the air," says Haymitch. "See? He's already demonstrated good judgment." Something's wrong. Haymitch's trying a little too hard to cheer me up. It's not really his style. "So who else volunteered?" "I think there were seven altogether," he says evasively. I get a bad feeling in the pit of my stomach. "Who else, Haymitch?" I insist. Haymitch finally drops the good-natured act. "You know who else, Katniss. You know who stepped up first." Of course I do. Gale.