Author Archives: Zewen Liang

About Zewen Liang

Eric Weinstein’s view on Peter Thiel

Peter does a better job figuring out who the generative heterodox thinkers are. And then he realizes very often that they don’t know the best way of putting forward their ideas. He’s also got another idea that academicians don’t understand, which I call “Maximumly compress minimally distorted”, where he takes some ideas and he is willing to give up a tiny amount of accuracy to make it extremely compact so that when unpacked, it tells you a ton. And because academicians over focus on special cases, he very often runs into the table where someone would object to the idea that his principle isn’t exactly accurate.

Peter is an insanely difficult person for most of us to model, because I don’t think it’s even consistent session-to-session. Peter is so driven by the search for something new, that, whatever it is his thought lasts, is now incorporated into “Well, that’s old. What do we go from here?”

One of the reactions that people had was “Wow I have never seen this side of peter.” The delta between the private person and the cartoon character the media portrayed has gotta be one the largest delta of any other human beings.

Peter is a intensely moral thinker – and how the moral element of his thought interacts with the cognition and strengthens it. (this comment is actually from Tyler)

Episode 16 of The Portal

Atomize the work and team

Small team is incredible. And this totally strikes me and I can’t agree more since I had the privilege working on such teams before:

“There’s nothing more fun than having a very fast moving team where everyone’s rolling in the same direction. That feeling of power, and that feeling of excitement, and that feeling of energy.. There’s nothing like that in my life, before and after.”

Eric Schmidt said in the Tim Ferriss Show

When you think of why a company is the size it is (company size of 5 v.s. 500 v.s. 50,000), an interesting question might be raised: what decides if a large group belongs to a company internally, or should that group be external like a service provider?

Aside from some obvious reasons (e.g. enough capitals to do so), one theory from the economist Ronald Coase is that the limitation comes from the transaction costs – if external transactional costs outweigh internal transactional costs, it’d be done internally; and vice versa. So it is also possible that the external transactional costs are high at a given time in the history, but could completely change to the other way around, because of the “API economics” invented during the information age. For example, you might not need to build a dedicated engineering team within your company to develop payment technology, since external companies like Stripe can provide a much powerful and easy-to-use payment API. Everything becomes “plug-and-play”.

Another well known issue on a bigger group is efficiency – process is created and is inevitable – to make sure everything comes together consistently. But process is what you follow when you don’t have a better idea. And things get worse when there are too many people wanting to influence every decision. The progress becomes stagnant..

Also, in terms of Psychological safety, the scale of the team matters because proximity and intimacy matter.

“With my family, I’m a communist. With my close friends, I’m a socialist. At the state level of politics, I’m a Democrat. At higher levels, I’m a Republican, and at the federal levels, I’m a Libertarian.”

Naval Ravikant

In a bigger group, people tends to feel less psychological safety, and it’s challenging to be authentic. Eugene Wei explained one of the biggest challenges for Facebook, as asocial media, is it runs into the diseconomies of scale. By nature humans are very good at maintaining different identities for different groups. Very few of us have cultivated an identity for that entire blob of everyone we know. But as the Facebook’s newsfeed munches all things together. It’s a situation one might encounter in the real world only a few times in life, perhaps at one’s wedding or funeral, but now it happens to be the default mode on Facebook’s news feed.

The most interesting systems, like cells in a human body, are those that are cross-connected in ways that allowed parts to be interdependent on each other – not hierarchical.

(to be continued..)

Changes in asset ownership evolves tech

Reading through Ben Thompson’s What is a Tech Company, one of the most fascinating aspects I’ve found in the evolution of software comes from the changes of ownerships – who (users or companies or 3rd parties) owns what (hardware, software, data, etc.).

(Old) IBM – Sell to you the hardwares (computers/servers) we made, so you can install the softwares we produced. (Heavy operational costs for IBM, like assets maintenance and IT training. But the bill is on customers.)

(Old) Microsoft – You have your own computers/servers; now you need to install the on-prem softwares we produced so it can work on your hardware. (Arguably the most effective/lucrative business model because of zero marginal cost for Microsoft – distributing softwares doesn’t cost anything.)

Salesforce – We run the softwares we produced in the servers we own. You don’t need to own your servers or softwares – they are off-premises. All you need is to use any web browser to access it. (The definition of SaaS. Has fixed cost for running servers etc. but again this is covered by up-front paid customers.)

Atlassian – Sell to companies without even talking to them. Freemium, Easy-to-pay, and Pay-As-You-Go, from anywhere by anyone. (An evolution from SaaS but it gives emphasis on the bottom-up approach and the democratization aspect.)

Software creates ecosystems.
Software has zero marginal costs.
Software improves over time.
Software offers infinite leverage.
Software enables zero transaction costs.

As a separate note shown above, it is also interesting to see what criteria Ben Thompson uses to define a tech company.

Reading books as news feeds

This year I’ve been running an experiment of reading physical books as though they are feeds on Twitter.

It means that, I read 5~10 books in parallel within a given time — read one chapter in a book when I want, then jump to another book without finishing the previous one.

Reading in a bundle helps you connect dots if there are any. Yet a bundle of books are just a small set of consumption of all mediums. Today people have a great mix of diversified types of mediums: podcasts, youtube videos, social media feeds, magazines, favorite news sites, blogs via RSS, subscribed contents distributed via personal emails, subject matter discussions via corp emails, PPT proposals from your colleagues, etc. People might read less physical books, but I believe we actually “read” more today.

Another thing I adopted is to feel less guilty of not finishing a book — I rarely finished any books I recently touched. Instead of reading a book in order to finish it, maybe a valid measurement is how hard you think during the time you read it.

If someone you trust recommends a great book, you may buy it and put in your bookshelf. There’s no need to feel guilty of letting this book idle either. Since you still absorb information from across all other available mediums.

Employee performance metrics

It is common to hear that employees at Facebook get more burnouts compared to those at Google. Facebook does not necessarily have more projects per person – and I agree with what TechLead has said – it’s more about how the systems are set up, particularly when it comes to the criteria of how the companies evaluate their employees.

The performance metrics at Google are Complexity, Impact, and Leadership, while Facebook’s are Impact, Better Engineering, People, and Direction.

I’ll skip Google. But in Facebook, the Impact and Better Engineering sound a bit of a paradox. Because one wants to “break things” to ship as many features as possible in order to make product impact quickly, but could possibly sacrifice the engineering quality.

The People and Direction is more or less a paradox too – employees can be trampling on each others’ works to lead with a new direction of their owns for the same project. But being too aggressive like this could make them lose on the People metrics. So they learn to be passive aggressive.

It’s the constant battle with the paradoxical performance system that can make Facebookers feel more stressful.

Blame game

Reminds me of Malcolm Gladwell’s episode Blame Game – the 2009 Toyota sudden accelerate scandal was overwhelmingly a matter of human error. In fact, a human factor expert said everything about sudden acceleration looks like a problem with the driver, not the car. We just couldn’t admit it.

Qualities for co-founders and leaders

Paul Graham suggested to look for co-founders that are “animal” – “someone who does what they do so well that they pass right through professional and cross over into obsessive.” Basically “a salesperson who just won’t take no for an answer; a hacker who will stay up till 4:00 AM rather than go to bed leaving code with a bug in it; a PR person who will cold-call New York Times reporters on their cell phones; a graphic designer who feels physical pain when something is two millimeters out of place.”

If he/she is a coder, 1) was the person genuinely smart? If so, 2) could they actually get things done? And finally 3) do they have unbearable personalities that we stand to?

Regarding to spotting the managerial talents. Ben Horowitz thought there are two skills that don’t normally go together. It’s a rare thing.

First, system thinking. Most people are not system thinkers – meaning they cannot think about: ok if I change this here it’s gonna affect things over there.

Second, can you actually see the people in your organization – do you know who they are, as opposed to talking to them like they are you. Do you understand their motivation; what they would think about something if they weren’t in the room, and you are making a decision; can you interpret them well enough so that as though they were there, there. Can you understand the implications through the eyes of people who work for you. You might not be able to articulate something, but they can articulate it for you the way you would have done it better.