4 ms·

The difference is the prior. There's strong prior reason to believe not paying payroll tax is not going to end well. The theoretical basis is simple, clear, we

The difference is the prior. There's strong prior reason to believe not paying payroll tax is not going to end well. The theoretical basis is simple, clear, well-reasoned, and supported by strong evidence in adjacent realms (i.e. the IRS takes taxes taxes pretty seriously). Whereas the prior on the validity of something like "early-stage founders should live with their customers" is much weaker. Even though the advice for each presented example is based on sample size n=1, the resulting belief can and should be quite different depending on your prior. The real meta-advice is to be able to understand and evaluate your prior for any given piece of advice you receive. This mostly comes down to thinking rationally, being knowledgeable, and critically considering each piece of advice for whether it's worth heeding or not. Which is not so far from the author's message I think ("figure everything out yourself", "I don't mean to say don't take any advice"). (Obligatory shout-out to Bayes Theorem)

nostrademons 7y ago
> Whereas the prior on the validity of something like "early-stage founders should live with their customers" is much weaker. There's a pretty strong theoretical reason for that as well: startups succeed because they make something customers want; the more time the person making product decisions spends with times with customers, the more information they can gather about what the customers want; and the more the customer trusts and has a rapport with the founder, the more willing they'll be to share candid feedback. There are some practical problems that make living with your customers a little problematic if you're not Brian Chesky, namely that they may not want another roommate, you may not want a roommate, they or you may already have a family, or you might just get on each others' nerves. But if you extend the general principle to "spend a lot of time with your users" (and being able to extend to a general principle is a big advantage of having theoretical reasoning), there are a lot of other companies that have found similar success. Facebook was another one where the founder literally lived with his first users. Google, for all intents and purposes, lived with their first users (it's grad school, after all). DropBox, Reddit, and Instacart were their first users. Apple, Stripe, Clerky, Twitch, etc. were in regular contact with their first users.
- Retric 7y ago
  That’s an overly wide definition of living with your customers to be useful. Apple for example was selling people kit computers at the start, they did not really have much interaction with or feedback from their customers. Nor could they do much to adapt their product with customer feedback. I mean if all you mean is have even a tiny amount feedback from a few of your customers. That’s going to include almost every company.
  - meristem 7y ago
    "Spend the relevant type of time with the appropriately targeted type of users"
  - unscaled 7y ago
    Their first customers were microcomputer enthusiasts and both Steves were members of the Homebrew Computer club, and Woz built Apple I with continuous feedback from the club meetings. I would argue that it's moving to the Apple II (were they suddenly lost this direct feedback because of the vast expansion of their customer base well beyond their reach.
Dylan16807 7y ago
There's good reason to believe tons of sociological theories. The problem is figuring out whether the pro or con factors win out, and figuring out if you've nailed down every single factor. Being simple, clear, and well-reasoned in the realm of people's behavior is not enough to give you very good accuracy. The kind of priors you get by reasoning in a vacuum just aren't good enough in some realms. Bayesianly combining those priors with anecdotes doesn't really help. Some ex-post theories are fine but I don't think you've hit upon the way to distinguish them.
- chibg10 7y ago
  >There's good reason to believe tons of sociological theories. The problem is figuring out whether the pro or con factors win out, and figuring out if you've nailed down every single factor. Being simple, clear, and well-reasoned in the realm of people's behavior is not enough to give you very good accuracy. This is all argument in favor of a weak prior, not to ignore the existence and utility of a prior completely. Ideally, you have solid (relevant) empirical data that informs your prior but if you're seriously evaluating "advice" then I'd venture to say that you're probably lacking such data. Imo, the critique to be had of what I said above is that there's no clear general way to form good priors and thus the advice is somewhat empty. But that's just life -- there's no magic bullet algorithm to discover truth and one has to use his/her brain to do the best they can. Many who seem to well in this area tend to like the Bayesian framing of beliefs, and personally I've found it useful to guide my own thinking as well.
  - Dylan16807 7y ago
    You can reframe almost any statistics as Bayesian if you try hard enough. That doesn't mean it's actually helping you. If you have a method for generating 'good' priors and it doesn't work, it does far more harm than good when you use that to analyze this kind of anecdote. Anecdotes are such weak evidence that the chance it pushes your priors into the correct result is super low. You'd be better off not trying. Just accept the need to gather proper amounts of evidence. Or in other words: Some undescribed system of priors could help, but the system you actually described doesn't help. So "use priors" is a bad answer.

nostrademons 7y ago

> Whereas the prior on the validity of something like "early-stage founders should live with their customers" is much weaker. There's a pretty strong theoretical reason for that as well: startups succeed because they make something customers want; the more time the person making product decisions spends with times with customers, the more information they can gather about what the customers want; and the more the customer trusts and has a rapport with the founder, the more willing they'll be to share candid feedback. There are some practical problems that make living with your customers a little problematic if you're not Brian Chesky, namely that they may not want another roommate, you may not want a roommate, they or you may already have a family, or you might just get on each others' nerves. But if you extend the general principle to "spend a lot of time with your users" (and being able to extend to a general principle is a big advantage of having theoretical reasoning), there are a lot of other companies that have found similar success. Facebook was another one where the founder literally lived with his first users. Google, for all intents and purposes, lived with their first users (it's grad school, after all). DropBox, Reddit, and Instacart were their first users. Apple, Stripe, Clerky, Twitch, etc. were in regular contact with their first users.

Retric 7y ago
That’s an overly wide definition of living with your customers to be useful. Apple for example was selling people kit computers at the start, they did not really have much interaction with or feedback from their customers. Nor could they do much to adapt their product with customer feedback. I mean if all you mean is have even a tiny amount feedback from a few of your customers. That’s going to include almost every company.
- meristem 7y ago
  "Spend the relevant type of time with the appropriately targeted type of users"
- unscaled 7y ago
  Their first customers were microcomputer enthusiasts and both Steves were members of the Homebrew Computer club, and Woz built Apple I with continuous feedback from the club meetings. I would argue that it's moving to the Apple II (were they suddenly lost this direct feedback because of the vast expansion of their customer base well beyond their reach.

Dylan16807 7y ago

There's good reason to believe tons of sociological theories. The problem is figuring out whether the pro or con factors win out, and figuring out if you've nailed down every single factor. Being simple, clear, and well-reasoned in the realm of people's behavior is not enough to give you very good accuracy. The kind of priors you get by reasoning in a vacuum just aren't good enough in some realms. Bayesianly combining those priors with anecdotes doesn't really help. Some ex-post theories are fine but I don't think you've hit upon the way to distinguish them.

chibg10 7y ago
>There's good reason to believe tons of sociological theories. The problem is figuring out whether the pro or con factors win out, and figuring out if you've nailed down every single factor. Being simple, clear, and well-reasoned in the realm of people's behavior is not enough to give you very good accuracy. This is all argument in favor of a weak prior, not to ignore the existence and utility of a prior completely. Ideally, you have solid (relevant) empirical data that informs your prior but if you're seriously evaluating "advice" then I'd venture to say that you're probably lacking such data. Imo, the critique to be had of what I said above is that there's no clear general way to form good priors and thus the advice is somewhat empty. But that's just life -- there's no magic bullet algorithm to discover truth and one has to use his/her brain to do the best they can. Many who seem to well in this area tend to like the Bayesian framing of beliefs, and personally I've found it useful to guide my own thinking as well.
- Dylan16807 7y ago
  You can reframe almost any statistics as Bayesian if you try hard enough. That doesn't mean it's actually helping you. If you have a method for generating 'good' priors and it doesn't work, it does far more harm than good when you use that to analyze this kind of anecdote. Anecdotes are such weak evidence that the chance it pushes your priors into the correct result is super low. You'd be better off not trying. Just accept the need to gather proper amounts of evidence. Or in other words: Some undescribed system of priors could help, but the system you actually described doesn't help. So "use priors" is a bad answer.