Nadia Eghbal

I often scribble half-baked ideas, reactions to things I’ve read, or something useful I’ve heard. Sometimes they turn into longer blog posts or projects, but most of them sit in my notes app, unused.

I’ve decided to start publishing some of these as a faster way to get ideas out there. They’re updated monthly and loosely organized into sections below.

To get monthly updates, subscribe here. And if you ever want to discuss something in here, drop me a line on Twitter or email

(Please note: you are crawling my brain. These are rough notes, which means ideas are experimental and conviction level is highly variable!)




Digital goods with open production models:

Both factor into open source software (and other forms of knowledge curation) not being perfectly scalable vs. other digital goods


How does our behavior change when we presume an environment of abundance instead of scarcity? (SV tech is a good example of this vs. other professional cities/industries)


Theory: you only need 1 or 2 competitors to keep an otherwise-monopoly in check. (Ex. Gitlab for GitHub, Lyft for Uber, Google Cloud and Azure for AWS). “Minimum viable competition”? Is that a thing?


Infinite use but finite production explains cognitive overload among creators in general (not just OSS devs but even people on a daily basis). Kinda parallels the dynamics of “there are only so many apps you want to own/use regularly”. Information overload


Thought experiment: what if monopolies aren’t bad? And the actual question is more “how does everyone get fairly heard” and other adaptations, but not actually breaking up monopolies entirely? Much like conservationism and the belief that we’re supposed to return back to some former state of nature that doesn’t actually exist. What if we’ve assumed this is all bad under classic economic conditions, but under post-capitalism it’s actually the new natural order of things?



Problem with UBI: there seems to be this utopian assumption that UBI will give people the power to negotiate, walk away from bad opportunities, etc. UBI might prevent you from starving, but you’re still poor. People still want to save money. They still want to have extra cash. People will still want extra paid work over top of UBI bc most ppl want more $$, status and power. They will still want purpose in life. UBI provides basic services that a society should afford to all its citizens, and it gives opportunity to some ppl who will want to use it to take risks on new ideas. But lots of underestimated cultural externalities, including that it will highlight and exacerbate differences in aptitude. (Much like how the internet connected us a little too much by making us aware of the existence of ppl we don’t like)


Scarcity and how it plays into funding behavior for creators. Patreon hasn’t really changed fundamental dynamics of how creators ask for money or get paid. There are an infinite number of people that you could donate to, that’s why it’s impossible to pick. Patreon, GoFundMe all have this problem.

What would a “reverse Patreon” look like, where patrons, rather than creators, to advertise that they’re funding XYZ opportunities? Patron-centric platform (more like AngelList) Companies don’t usually post on their websites that they’re looking for funding, it’s investors who make themselves known, so why do we expect creators to?

It’s on the source of limited capital - patrons - to hang a sign on the door and advertise that they’re spending. Bc they’re the source of scarcity. There are endless funding opportunities and that makes it stressful/impossible for a funder to distinguish (this is like Nick Szabo’s micropayments and mental transaction costs)



Protecting against malevolent miners by forming strategic partnerships with big players is like building up a project’s defensibility via military base. You’re basically starting a new country and you need a defense system to go with it


Theory: in open source projects, governance matters most when financial interests are involved (true for big corporate foundations, also true for crypto stuff, but not true for ex. tiny npm modules)

Without external influence of money, project communities can be trusted to self-organize (is that true, though? scale is prob another factor here, bc communities also need governance for efficiency/coordination purposes). Maybe not about money, but presence of divergent interests? I guess that was basically Ostrom’s theory


Rewrite democratic principles according to software principles. NOT: using software to enforce existing principles. But using observations of software principles to rethink how we self-organize in government.


Idea for structuring foundation board elections: board seats are forever, but anyone can propose an election for the upcoming year. (Basically, if you’re doing a good job as a leader, and you’re well-respected, why not stay on for a very long time? Reduces switching/onboarding costs. But, there should be an easy process for turning you over if/when people start to feel differently, and you don’t step down on your own)


Without a verifiable “proof of personhood” (meaning both identity and reputation), online voting is worthless. Without knowing what the total population looks like, we don’t know who matters more/less. I guess the point of democratic voting is that everyone’s vote matters equally, IMO not as true for decentralized communities. As we try to rethink voting from the ground up today, it seems like we’d want to weight some people’s votes more heavily than others, especially depending on the topic


Reputation as enforcement mechanism for enabling recursive governance? If you can somehow make reputation immutable, then people wouldn’t want to risk their reputation on stupid things. Laws enforce themselves bc you don’t want to permanently damage your reputation. The system doesn’t need to monitor you; you monitor yourself and your own behavior based on social norms. The law theoretically should reflect social norms today, but even better to remove one step and make the law == social norms (Goal is to trigger self-regulation instead of relying on external mechanisms)


If ancient Greece and Rome serve as the foundation of modern democracy, then software is the foundation of modern technocracy (government by experts, which is more feasible/democratic today than how the term has historically been used)

Software design principles also describe optimal paths for how people self-organize (which one is natural order: gov’t or software? Neither?)


Strive for smaller populations of governance. Not reducing government itself, but the size of populations that can be governed

Membership between populations should be more fluid. Would market dynamics incentivize everyone to do better? Imagine if governments were on a marketplace and had to compete for citizens via their offerings, how would they perform? This is already theoretically somewhat true today, except it’s way/easier harder to leave/enter certain countries. Also, what would it look like to have a fully digital gov’t that competed in an open market along with physical ones? Could it do better?

I assume common critique would be that price goes up for the best governments. If we were able to keep cost of services low, could they compete on price? Would it converge into one monopolistic uber-government?

Curators create “spikes” of new governed populations by virtue of their leadership. They tell us what to pay attention to


Thinking about how many government services could be made digital. There are obvious tasks like registering a business, but other interesting services like emergency response or protection.



(from a friend) “We are different and we might disagree on stuff, but you’re still within my Goldilocks zone”


With Uber out of the picture, Tesla has taken its place as the new media whipping boy. We always need a tech villain in the media. (If Tesla were knocked out somehow, who would take its place? Not biggest tech cos which are on a different level of media scrutiny, but among companies that are oversized startups)


The act of dating is like reincarnation via multiple relationships until you reach enlightenment?


There’s an unspoken rule that you can’t be a self-aware contrarian, or else it invites mocking and derision for being “better” than everyone else. Ex. nobody else WANTS to call themselves a hipster, they’ll only do so if mocking themselves or others, even if they actually are. I dunno exactly how this applies to neoliberalism but I suspect it’s why the “brand” is so intertwined with being a harmless, optimistic, somewhat awkward but cheerful nerd


Dunno how to phrase this, but: I wonder whether the historical focus on “make great ideas happen by building startups” was influenced by a masculine-dominant industry (“making things is the way you express your good ideas”).

Either way, I feel like the longer historical narrative/arc is that startups were this one crude tool we had for making good ideas happen, and now slowly we’re seeing a much richer ecosystem form, where you can capitalize on having good ideas, and reap the reputational rewards, in faster and more agile ways. Over time, we’re increasing the speed at which good ideas can come to life, and which we can quickly learn/discern who does/doesn’t have good ideas.

When you look at it that way, building a company is this extremely risky, slow, sluggish, difficult way of doing it, that requires you to be locked up in a long timeline. (Although the rewards can be correspondingly higher than smaller, more iterative ways)


App that just scrolls endless boring Lorem ipsum with occasional puppy images. Like prayer beads, fidget spinner, or nicotine patch to replace your social media addiction


Naming things is hard, not just figuring out the name itself, but bc of the timing. Sometimes it’s the wrong time to propose a new name, it won’t stick


Existential fiction featuring a trophy wife protagonist as Gregor Samsa


Are we always looking for music no one’s ever heard of bc music is so personal and we don’t want to feel that emotional connection towards something that lots of other people also feel? “My emotions are unique”


Manic pixie dream girl is the equivalent brand counterpart to fortune cookie pseudo-intellectual on Twitter


Why do wealth and interest in the environment seem to be correlated? John Muir, Teddy Roosevelt, etc. (Part of this is just NIMBYism, though, i.e. protecting environment for its aesthetic value) And also developing countries that don’t prioritize the environment at all. Does that make the issue frivolous, or is it just a Maslow hierarchy thing?


Why has interest in sensational media stories declined? (ex. cultural touchpoint stories about serial killers, kidnappings, etc) Maybe violence itself declined? But also, I guess sensationalized social media replaced it?


Diets as correlated to politics. Vegetarianism (far left) is all about sacrificing bc you’re thinking of others. Paleo (libertarian) is about doing what’s right for you. Individual-centric


(from a convo with a friend) “Culture as liturgy”. Why get rid of liturgy in religion? We embrace it everywhere else. It’s not outdated or old-fashioned, it what keeps us closer to God. The mall as an example of cultural liturgy. We’re all chasing novelty, but novelty doesn’t bring us closer to each other. We’re creatures of habit, we’re always remembering and trying to remember.

(Side note: maybe this explains the wave of 90s nostalgia and remakes. It’s not that we ran out of ideas, it that we’re looking for a cultural touchpoint for remembering together)


Mechanics of synesthesia: thinking about how it’s 2018, and remarking that the color of 2018 isn’t going to change much from 2017, because 7 and 8 are such similar colors (8 is just more purplish, but both are very dark).

However. In the number 2000, 2’s color is most prominent, so that the overall number takes on that color. But in 2017, the color of 7 is most prominent, and the overall number takes on that color. That suggests that perhaps the color of a number is intertwined with which aspect of that number I find most significant

So in 2000, the 2 is most important bc it’s telling you how many thousands you have. But in 2017, the 7 is most important bc it tells you which year it is.

To use another example, in the word Nadia, it’s purple overall, but that’s bc N is purple. But A is red (like 2), D is cyan (like 3), I is pale yellow (like 1). None of those colors are visible when we string them together into Nadia, except for N, which is purple. That makes sense bc your first initial is often used to signify your name.

Online social interactions


Communities can have heterogeneity and diversity, as long as they’re still high context and high trust (that often comes from ppl being similar to each other, but similar values and superficial differences works just fine too)


So, in open source, “users” are useful to an open source project bc they can also serve as evangelists, even if they never directly interact with the project. Any one user might not mean much, but in aggregate, they’re the engine of a project (bc they boost popularity and adoption)

Similar case for social amplification of the news. It’s easy to deride people who post about the news all the time bc they’re not engaging in meaningful civic participation, but in aggregate, they are actually the news “engine” that ensures headlines get paid attention to

But there’s a trade-off in incentives between my time as an individual vs. the cost of participating in this cycle. This makes it a sort of commons-type problem to resolve: it’s in the collective best interest that I participate in the news cycle, but it’s in my personal worst interest to do so, so I don’t

There’s no short-term danger in avoiding the news cycle bc there are plenty of ppl who will still do it for you, but in the long term, does this create a stratification/caste system where “low cognition” ppl mindlessly contribute their time to news cycle engine, and “high cognition” ppl opt out to focus their time elsewhere? Although, also, you don’t have to fully opt in or out to the news cycle. Maybe you lend your voice when it’s convenient/relevant and ignore the topics that aren’t, which ensures equitable distribution of the collective signal boosting burden


Part of the problem with measuring financial value based on contributions themselves is that whether a contribution gets accepted is often tied to social merit, not the intrinsic value of the contribution.

So basically, there might be an even “worthier”/more valuable contribution that doesn’t get accepted bc of preexisting social norms. If you only give $$ to the contributions that get accepted, isn’t this some version of “wealth begets more wealth”?

(Although, I don’t really know how you’d overcome that bias, either. And in some sense, maybe “usefulness to project” is the right way to evaluate. Maybe figuring out how to follow the social and technical norms of a project is part of the process of attaining value)


Research suggests high context matters in open source:

So using that as a basis, how do we design online products and platforms to retain high-context situations?

Ex. in a social media context, emphasize small group interactions and filtering out unwanted interactions. In fundraising context, raise $$ from people/companies who already use and know you


One of the best things about Twitter is that your social circles can evolve with your interests. You can follow/unfollow people whose feeds you might not resonate with anymore. Whereas FB is like a mausoleum of people you used to know that are constantly being shoved in your face like “remember this? look at it”

Conversely, it’s also really nice to see little social peer “clusters” pop up among the people you follow on Twitter. You might follow all of them, but you can see diff peer groups naturally cluster around different tweets.


Cities and their neighborhoods are a good physical model for scalable online communities. Each neighborhood is its own “pod” with its own shops, services, etc but you can easily get to another neighborhood if needed, and everyone has a broader interest in city-level needs. Also, preserving mobility between neighborhoods is important!


Context collapse is just another term for high discount rate. When you think your actions don’t matter long term, you’re more willing to act in ways that are reckless and damaging


Distributed programming languages as an analogy for running societies at scale?


I wonder if college campuses are a metaphor for what happened on the internet. Too many people smooshed into the same room, becoming aware of each other, then provokes outrage, but maybe we’ll actually just all settle down and be fine as we “grow up” together and learn to accept differences (or filter out into our own tribes). Maybe this current stage is just our collective adolescence, the internet’s “Culture Wars”. Maybe we don’t actually need to worry about outrage culture in the long run.


Social media decentralized the concept of friendships (I guess this is true for dating, too) You used to only be able to make friends based on physical location/connections. And in the early days, internet friends were people you liked but didn’t really meet up with, they were their own separate world. But now you can actually make new, real friends that are location-agnostic. You don’t need updates from your purely physical friend world now. You want updates from your virtualized internet friend world that you hand-picked based on shared interests.


Closing of public spaces = this is something that government and digital platforms are both capable of

Open source


Maintainers should treat user-evangelists like they do developer-contributors. Give them the tools to make their contributions additive/free to you. Ex. creating a dedicated support channel and letting users run it themselves


Onion model of an open source project isn’t quite right…more like a “spider” model, with maintainers as central hub, and each spoke following its own onion model


Maintainers continue to maintain bc of obligation to community which is also another way of saying high-context incentives


Theory: JavaScript might have more staying power than other languages, despite its complexity, bc of the community, willingness to talk and share knowledge and educate. Proliferation of workshops, blog posts, videos, etc. Thesis is that human skills increasingly win over technical advantages in choosing what you want to learn.


Pet peeve: when people cite Linux as a sustainable model for open source. Linux is to open source what Facebook is to startups: fascinating stories, but extreme black swan events. Yet I’m surprised by how often Linux is cited in mainstream conversation. Maybe bc it’s the last big cultural signpost that open source has.

Who is today’s version of Linux/Linus Torvalds? What characteristics will they have? (Or has the game changed entirely, so that this question is irrelevant?)

Organizational culture


What are examples of companies who have gone through organizational trauma? Examples:

Who else?

“Organizational trauma” = a huge event that the company went through, which marked the end/beginning of a cultural period, and the memory (and fear) of which still impacts how they make decisions today


From Call to Commitment, on why the “new guard” eventually becomes the “old guard” (p.53):

“The young have not proved more pioneering than those who have gone before them until they have been tested. Strangely enough, we often find that those who protest change the most are those who were most attracted by the adventurousness of the group.”

“It is easy for one generation to overthrow the structures of another and to think itself bold and adventurous. But the test comes in whether we can part with the structure we ourselves have created, for new forms, like the old, can come to represent safeness and security.”


How crypto companies might look different from traditional startups:


(from a convo with an HKS student) “Zuckerberg is the George Washington of Facebook”

Tech problems are the same as policy problems. It’s about governance, community, people. False dichotomy between “bringing tech to policy” or “bringing policy to tech”, though they have different norms/cultures. Belief for awhile within tech that they could do better than gov’t and avoid bureaucracy, but it turns out at a certain scale, all institutions have problems. There was nothing special about tech. (This is really scary and also true, I think! Tech isn’t actually immune to bureaucracy, it just managed to avoid these problems for 10-20 years bc it was so new). Zuck, as the “George Washington of Facebook”, has to figure this stuff out now just like anyone else.


Two kinds of leaders: those who work to become heroes, and those who work to become obsolete


You don’t want to be a “values-driven” company to the point where it distracts from your company’s purpose. You should just be a company with basic standards of decency, excellence, professionalism.

Also, theory: you can only afford to be a mission-driven company in a space that doesn’t have monopolistic competitors. Mission-driven companies with network effects can also last forever…until they get out-innovated. So a friendly organic cosmetics company can theoretically exist in a world with P&G, bc consumers have choices, but couldn’t in a world where there’s only one main choice (ex. Facebook is an imminent threat). Also bc there’s nothing to really acquire in this case, vs. ex. Coca-Cola acquiring Honest Tea for its brand. But Uber wouldn’t acquire Lyft for its brand.

Personal growth


“He’s very good at resting” - I can’t stop thinking about this. Resting as something you’re actively good at (that means not playing on your phone or even meditating, but literally just sitting there and resting). Been trying to incorporate moments of resting throughout the day. I want to be good at resting!


How do you know when your brain’s dependencies are outdated? We operate based on the values/tastes/preferences we assume we have, maybe they worked for years and years, and they inform the actions you take daily without thinking. You’ll even feel like you’re truly “being yourself” but they produce actions that don’t feel quite right. At some point they become outdated and you have to update your underlying system. By definition, can you only recognize your dependencies are outdated once it’s already happened, or can you anticipate beforehand? (Maybe this is an equivalent analogy to hitting a “local maximum”)


(from a conversation with fellow researchers)

Essential q: “Who are you trying to impress?”

I like this question better than the “average of the 5 ppl you spend the most time with” as a means of measuring your quality barometer. A person, or the idea of a person, can motivate you for years and years, even if you don’t talk to them anymore, even if you’ve never met them, even if they don’t know they’re influencing you.

Example of this: Stewart Brand asked three friends for their photos to keep by his desk while writing “How Buildings Learn”. Less about checking in with them, but thinking “what would impress them” was enough to motivate


Ze Frank has this video about how when people have ideas, they’re afraid it’ll be their last really good idea, so they tend to hold on to them in their head, and not share them (it’s like “brain crack” - “I know I’m smart bc I thought of this cool thing, I can hold on to this idea that I’m smarter/awesome bc it’s still in my head, I haven’t put it out there to have anyone argue with me, or customers react to it, etc”)


Creative process is mostly miserable bc you have ideas that are insanely important to you but no one else cares about. But over time you come to realize this is a blessing and a sign that you’re into something good. (Or you don’t and you get depressed)


It seems that with any extremely close relationship, we’re constantly navigating this struggle, and perhaps experiencing grief, each time as we learn that no person can possibly live up to the standard we’ve created. Sometimes people deal with this by only playing in the shallows: casual acquaintances, superficial connections. Other people dive too deep: expecting the impossible from those around them, demanding that they change. Ultimately, we have to learn to respect the other person’s agency, allow them to breathe and be the person they truly are, not the identity we insist on imposing onto them. We are all actually alone on this journey, after all.


What’s the right balance between giving ppl opportunity, vs. handouts? Serious question. If you focus on removing barriers to entry so, e.g. a developer can “just code” or a founder can focus on “just building their product”, the counterargument seems to use the same logic, which is “a really good developer/founder/etc will find ways around those obstacles on their own”. How much do we expect people to specialize vs. generalize?

For the record, I gravitate towards the former, but I can’t decide why it feels intuitively different from handouts or preferential treatment or quotas or whatever else. I guess bc you’re not giving them the reward, just reducing their obstacles? Is that the dividing line?


(from a friend) Being a server is fun for him, bc your job is just to care. It’s so easy to go slightly beyond the average and make other people ridiculously happy. “Your product is your personality”


Theory: Work is for doubling down on your very best skills. Hobbies are for doing things you’re naturally bad at (but enjoy the process of getting better at), it’s like cross-training for your brain, so you get better at your weakest skills


I think the busiest people make a point not to overschedule their lives bc it’s the only control they have left. Being constantly overscheduled makes you feel out of control, being spontaneous and refusing to overly schedule social stuff gives you control again.


Common wisdom is to list the things you’re grateful for (i.e. things others have done for you), but I suspect it also feels good to reflect upon the things you’ve done to cause gratitude in others (i.e. things you’ve done for others)


We pick different types of media to be obsessed with, ex. movies, TV, music, or books (and relatedly: knowledge). But people are rarely good at keeping up with all of these together, and people who are obsessed with one form vs. another tend to have their own set of behaviors and subculture. Maybe an obsession with ideas/reading is simply yet another form of media, not better or worse than TV (except it does feel more generative?)


“Free speech is not an value in itself, but a critical enabler of critical inquiry” (World After Capital)



For a long time, we assumed FB would just get out-innovated by the next platform, like Friendster or Myspace before it. (Or with GitHub, it’s the threat of SourceForge) But really, the platforms that were created in the last wave are here to stay. And if they’re here to stay, then instead we’re putting down roots. We were ok with it before when we felt like we could “exit” (switch to another platform). Once we realized we couldn’t anymore, now everyone’s freaking out. And in that sense, maybe regulation is the wrong answer to this question bc it legitimizes their power. Technically, we have the “right to fork” and build new platforms (much like revolting and starting a new country), but the energy required to change the system is much higher than it was previously. But still possible.


FB doesn’t want to be seen as a utility anymore, even though they are, bc it invites scrutiny. They’d rather be a media company etc. We’re treating lions like mice


Belief that “as long as you have basic infrastructure allowed, you can always set up a website and it won’t be that hard to get your ideas out” seems naive/short-sighted. If a website is up and nobody will index or link do it, does it really exist? Distribution is obviously what matters most. If that site were banned by all social media companies they wouldn’t be able to share their ideas, it’s just as bad as if you could only whisper about it at home.


Although it’s not being framed as such, this is maybe the first time we’re witnessing “libertarianism at scale” as an experiment (via private tech companies as platforms). Bc they’re private companies, but also, essential to the fabric of everyday lives, serving a purpose that gov’t normally would’ve, and also resisting regulation. Parallels with other large-scale historical attempts at new forms of gov’t? Also, for the first time, “the people” being governed is borderless, not tied to any one geography. FB’s constituency is not the United States, even if US regulators come after it, it’s the world.



Common approach to measuring social influence is based on who’s following you (a la PageRank/measuring links), but it’s actually your regular interactions that say more. Who follows you is a static measure, who you interact with is a dynamic measure.

Meaning: it’s not enough if someone influential on Twitter simply follows you. It’s more interesting to look at whether you two regularly interact. And similarly, it’s not enough if someone is, say, listed a contributor to an open source project. Maybe they got a patch merged once, but they’re not known/influential in the project. You want to see who’s regularly talking to each other.

IMO this is another data point in favor of an “open production” model being different from digital goods in closed production settings. You can’t really look at these things in terms of fixed-state activity


Ongoing list of open questions about how reputation is measured. Some of these are notes from conversations, others are just musings…

Does reputation degrade over time? What would an inflation schedule for reputation look like?

In open source, some ppl perceived as “high reputation” aren’t the most active maintainers today, bc they did a bunch of work in the past and then kinda cash in and sit on their reputations. So if “work done” is the input for measuring reputation, you’re not gonna see the ppl on top that you expect. And maybe that’s a good thing! Ex. a project author who still weighs in on the project but doesn’t actively work on it anymore, should they be able to do that, or should they cede control?

Is reputation transferrable, or tied inextricably to your identity?

Why do we value reputation? As a promise of how we expect you to act on the future. Reputation is speculation. Another way of asking this is, what is the difference between identity and reputation? Identity is a snapshot of who you are today. Reputation is who we expect you to be in the future. But both are based on past behavior? Reputation isn’t inherently interesting if we don’t think that past behavior transfers to your future behavior. Maybe it’s: “history”=past (training data), “identity”=today (present state), “reputation”=tomorrow (future state)

Quality of your reputation depends not just on the size of your following, but its quality. “Inbound links” from a few high-signal people is worth more than many “links” from low-signal people

Is reputation (and by extension, power) a zero sum game? Is there a finite number of high-reputation people who can occupy the same problem space? Seems like it is zero-sum within a specific niche, but not generally? (This is why academic researchers get more and more niche, to remain a big fish in a small pond, and thus attract funding. Also ditto nonprofits, targeting specialized funders)

Reputation begets reputation, much like wealth: if you already have a huge following, anything you share out will have higher “returns” on your reputation than if you don’t have a following

Do you “spend” your reputation to vouch for others, and if so, does this also work like wealth? (If you have a lot of reputation amassed, you can spend a bit to gain more, if it goes poorly, you haven’t lost much. If you don’t have a lot of reputation, you vouching for someone is a riskier stake)

What are the best universal reputation systems we have today? Most reputation systems don’t transfer (ex. GitHub or Stack Overflow). Twitter, and maybe also reddit, feel more universal. (And maybe Instagram?) Meaning, yes, there are clearly distinct crowds on the same platform, but if you have a strong reputation among “entertainment Twitter”, it still transfers somewhat if you’re trying to interact with someone from “tech Twitter”. Or whatever. It also seems to be an identity that follows you around on the internet more. Whereas, ex. your GitHub social currency doesn’t mean much on Instagram. Is Twitter the closest thing we have to a universal reputation system? Why is it so different? (Maybe bc, like reddit, there are so many use cases within the same platform, and unlike reddit, identity matters more?)


Three iterations of reputation systems:

External certifying authorities- ex. Universities or expert boards. Problematic bc control is centralized

Meritocracy - also problematic bc it’s still easily gameable. If your reputation is determined by your actions, you can always game those actions.

Networks (emerging)- reputation based on network consensus (which also requires an immutable identity! I think?). If enough people agree you are who you say you are, and that you did what you say you did, and you’re the right person for whatever’s required, you’re verified. Downside: Reputation networks are gameable too, in that, if you captivate an audience, you’re hard to topple (networks = easy long-term gains, meritocracy = easy short-term gains?)


Reputation is that which cannot be bought or sold.

Follow-up thought: This is also true for reputation networks. The highest quality networks are those which are earned and non-transferable. The lowest quality are ones that are completely open, second order are those which you pay for (money as a crude metric for success/ability), and best are those that are organically formed based on demonstrated merit (and/or, to enter that network, you have to be referred in and prove yourself). That doesn’t mean all the other networks are bad - just that this principle plays itself out on a network level, not just individually.


Fears around privacy are actually fear of the consequences that come from others knowing your private information. (In other words, we don’t really value privacy for itself, only its implications, i.e. fear of negative repercussions)

Research culture


Mind is blown that there is no Netflix/Spotify for research (journal subscriptions, etc). Only SciHub?!

If you take the very long view, are paywalled journal articles going to die out anyway? Would it be like building Blockbuster? But even if papers are open access moving forward, there’s plenty that will still be historically paywalled, right?

It just seems crazy in the meantime though. Is it that it’s that prohibitively expensive? How much does access cost a university, per user?


Can researchers ever be openly political? They’re more like judges - need to be neutral and truth-seeking (although in reality, no one ever is)


I wonder if one reason researchers tend to cluster together within their discipline, instead of being more interdisciplinary, is bc you have such a deep knowledge of your own field that it’s hard to talk to others about it with the same level of nuance (maybe true for experts in general, ex. medical students tending to marry each other bc nobody else understands your crazy doctor life)


(notes from a dinner with other non-academia researchers)

Measuring progress in research:

“I want to generate insights for this field that I otherwise wouldn’t have. Did I have thoughts that wouldn’t have gotten thought if I weren’t in this field?”

(this reminds me of success metrics in VC!)

Start with a question, then figure out how you’re going to approach the question. Have a “toolset” of research methods (ex. literature review, quantitative analysis, expert interviews), and pick the right one for the problem

Use a portfolio approach for methods: ex. Running pilots with a completion goal

How much should you cater to your direct research community/beneficiaries vs. the broader general public?

“Psychological safety” is important to being able to do good research: mobility, steady income, being able to switch tracks in case you lose funding, what would you do next. Yet a lot of us find ourselves in “lucky” situations. Are there ways to systematize these norms?

“Essential questions”: core questions to a field that are very important but broad enough to be accessible (ex. “What is the role of government”). Good questions to rally around for more generalized discussions


Quote from UX research presentation: “Some of this might be new, other parts will not be a surprise to you, but that’s often the point of research, to help validate the things we were noticing”


(from a convo with a friend)

What is the difference between being a hobbyist vs. a scholar? IMO, that one gets paid full-time and the other doesn’t. Not necessarily about a difference in quality or rigor (although there often is - but there are also hobbyist bloggers who take their research really seriously and apply lots of rigor, and professional researchers who don’t). It’s like a hobbyist software developer vs. professional developer. One is paid and the other isn’t, but an arbitrary title says nothing of the quality of their work.

Therefore: you shouldn’t feel bad being a non-professionally trained anything. A self-taught software developer doesn’t feel bad for not taking CS courses. Why would a researcher feel bad not having a PhD? Does it actually matter whether you’re a “hobbyist” or not? A professional researcher gets paid FT to do what they love, that’s the most important part.


Theory: black swan startup founders do best when they’re younger (need the wide-eyed optimism, bull-headedness and stamina to believe they can do anything), but black swan writers/researchers/academics do better when they’re older (need more life experience to articulate the world and form coherent theories)