Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

Do things that don't scale, and then don't scale

derwiki.medium.com

Post reply on HN
Print view View original

201–210 of 216 posts

Re: Do things that don't scale, and then don't scale

#201

Post by positron26 »

throwawayq3423 wrote: post #156
positron26 wrote: post #124

Earlier quoted context omitted.

Circles was a killer feature. And then I never really used it. In hindsight, any SNS begins as a megaphone and if it can't make statements in culture, it doesn't exist. It's not actually about the individuals. Sure, people want to post things, but they want to be feeding into something that does stuff. I don't think Google ever made that case. It was pure chicken and egg with no ice breakers.

> Circles was a killer feature. And then I never really used it. Then it wasn't a good feature.

The point is that we weren't looking for social media, so killer SNS features didn't matter. It was a better mousetrap for a market that's deep down motivated by mashed potatoes.

Re: Do things that don't scale, and then don't scale

#202

Post by Ar-Curunir »

coldtea wrote: post #184
Ar-Curunir wrote: post #142

Earlier quoted context omitted.

Where does it stop? What if your relatives support, eg, racist policies? What if they support genocide? How personal does it have to be? What if you’re dating a person of another race, and your relatives support keeps making racist comments to you, even after you try to explain why that’s not okay?

> Where does it stop? What if your relatives support, eg, racist policies? What if they support genocide? How personal does it have to be? Unless they act on them, they can support anything they like, it means less than them supporting some team or MCU fandom

What? Your relative Being a racist is less important to you than them supporting the opposing team?

Re: Do things that don't scale, and then don't scale

#203

Post by timr »

dghlsakjg wrote: post #194
timr wrote: post #150

Earlier quoted context omitted.

Vet clinics have a moat (exactly as you're implying): the customer relationship. That relationship, plus optimization on the fixed costs, gives an incentive to acquire the practice, maintain the staff as employees, and collect the operating margin. It's not like you really care if your favored pet doctor works for Brand A or Brand B, and -- let's be honest -- you're price sensitive. No matter what you say, there's a…

As you’ve pointed out, people are price sensitive. One of the substantiated gripes of PE owned vets (and many other things) is that they are more expensive, not cheaper. Price pressure is working against your argument here, not for it.

That is what happens in the later stages of monopoly. In any case, my point wasn't about vet clinics, it was about the business financial properties that lead to monopoly. Even a business model as mundane as the local veterinary practice is undergoing industry-wide consolidation.

Any profitable software business is vastly more interesting for consolidation than a vet practice.

Re: Do things that don't scale, and then don't scale

#204

Post by dghlsakjg »

timr wrote: post #203
dghlsakjg wrote: post #194

Earlier quoted context omitted.

As you’ve pointed out, people are price sensitive. One of the substantiated gripes of PE owned vets (and many other things) is that they are more expensive, not cheaper. Price pressure is working against your argument here, not for it.

That is what happens in the later stages of monopoly. In any case, my point wasn't about vet clinics, it was about the business financial properties that lead to monopoly. Even a business model as mundane as the local veterinary practice is undergoing industry-wide consolidation. Any profitable software business is vastly more interesting for consolidation than a vet practice.

Right, but the nature of veterinary practice don’t lend themselves to monopoly, nor is the excess cash even close to artificially achieving that. This is just what you see in an irrational market as well. All this stops when PE won’t overbid on practices, or when the vet pay is low enough that it makes sense to open your own practice. Many forms of software do favor monopoly. Rolling up all of the gov tech startups has network effects in addition to being an industry that operates on large fixed costs and low variable costs.

Vets aren’t like that, one vet has a limit to the number of dogs they can see in a day. I think it is mostly PE trying to mimic their success with doctors offices (which have massive network effects via insurance), and assuming that it will work the same.

It is very hard to have any monopoly power in an industry where there is no barrier to entry for your most valuable employees, and where, as you pointed out, customers are sensitive to price.

Re: Do things that don't scale, and then don't scale

#205

Post by layla5alive »

jakelazaroff wrote: post #72
bossyTeacher wrote: post #53

Earlier quoted context omitted.

>if you don't mind the politics of using that what exactly are the "politics" of using DeepSeek? Feels weird to single out DeepSeek like that

Not sure why parent is being downvoted here. Even without getting into whether it's possible for technology to be apolitical, many AI companies have explicitly political goals. For example, OpenAI's charter is "to ensure that artificial general intelligence benefits all of humanity". They go on to list more specific political goals downstream from that: https://openai.com/charter/

And you believe them?

Re: Do things that don't scale, and then don't scale

#206

Post by jakelazaroff »

layla5alive wrote: post #205
jakelazaroff wrote: post #72

Earlier quoted context omitted.

Not sure why parent is being downvoted here. Even without getting into whether it's possible for technology to be apolitical, many AI companies have explicitly political goals. For example, OpenAI's charter is "to ensure that artificial general intelligence benefits all of humanity". They go on to list more specific political goals downstream from that: https://openai.com/charter/

And you believe them?

That's neither here nor there. The point is that it's expressly political, so using ChatGPT is every bit as political as using DeepSeek.

Re: Do things that don't scale, and then don't scale

#207

Post by robocat »

timr wrote: post #122
salomonk_mur wrote: post #70

Earlier quoted context omitted.

Building a particular type of company - a tech startup. I'd argue that we need to apply this to companies too. Stop building startups (which start losing money and may never ever make a profit in the process of scaling) and start creating companies (which make a profit ASAP and may never scale).

"Tech startups" that make enough money to be interesting [1], but do not have the property of being a winner-take-all market, are the truly magical unicorns of software. If you have one, you should covet it and guard it viciously, because otherwise someone with more money is coming to eat your lunch [2][3]. OP wasn't talking about this kind of thing, of course. The phenomenon of hobby projects being "startups" has al…

> I define this at least as "supports more than one person's salary in perpetuity,"

Only if you ignore the costs of risk.

Let's say you hear that 90% of businesses are financial failures. What is the profit hurdle you need to exceed before you could declare success?

Zombie businesses (one person just earning a living) are common and often the income doesn't cover the risk premium (e.g. of one year of opportunity cost wages/ladder working as an employee). If you're doing a business, then in theory only the financial returns matter. There are non-financial returns (pride, control, yadda yadda) but those apply more to hobbies than businesses (not that I think anything is wrong with starting a hobby business if you're honest with yourself about motivations).

I think the VC rule of thumb is that one investment needs to do better than 30x return over 10 years to cover the losses on the rest of the portfolio. Therefore as an individual unless you get over 30x the return (versus lost wages) then you are under water? i.e. if you quit a $50k/yr job for a year, then only when you're earning $50k/yr AND have gained $1.5 million more then have you broke even.

Re: Do things that don't scale, and then don't scale

#208

Post by RangerScience »

coldtea wrote: post #182
RangerScience wrote: post #146

Earlier quoted context omitted.

AFAIK, you cut off shitty relatives… …and you usually find out they’re shitty because of how they handle (or don’t) “political” conversations. Pro tip: You don’t get to decide for someone else what’s “just” politics. If someone else says it’s important, while you’re interacting with them, it’s important.

> Pro tip: You don’t get to decide for someone else what’s “just” politics. If someone else says it’s important, while you’re interacting with them, it’s important. It's rarely, if ever, important to any concerned party's day to day life. It's bullshit partisanship instilled into idiotic brains by the media and social media. Just because they say it's important, doesn't make it so.

Eh. While that might be true in the course of say, governance - where IMO it’s important and worthwhile to use scientific methodology (statistics, etc) to establish “importance” -

…your family around the notional holiday dinner table? The personal is what’s important, kinda by definition. The point is the subjective and emotional.

Re: Do things that don't scale, and then don't scale

#209

Post by kylecazar »

derwiki wrote: post #5
kylecazar wrote: post #3

Building things for yourself is fun -- I do it. But the original article was written for startup founders, building companies.

Yep, totally — the OG advice was founder-focused. I just couldn’t resist twisting it a bit, because the line itself is too good not to repurpose.

I didn't mention it in my original comment, but I enjoyed your article -- and agree that small, personal, purpose-built tools are having a moment in the age of LLM's. I like that application of AI.

In a way (as I understood it), what you're getting at is that not everything has to be a business. Businesses definitely do have to scale, but personal software doesn't.

Re: Do things that don't scale, and then don't scale

#210

Post by stouset »

coldtea wrote: post #183
stouset wrote: post #145

Earlier quoted context omitted.

No, sorry. Nazi family members come last. Family members with no respect for the rights and freedoms of others and who encourage and support inhumane treatment of others don’t get a free pass.

Unless your family members are card carrying members of the Nazi party, "Nazi" is just a codeword for "disagree with them" just as "no respect for the rights and freedoms of others" is the same.

Sorry, but absolutely the fuck not.

If someone votes for and cheers on the sending of specific ethnic groups to concentration camps, the removal of rights and freedoms for political opponents, and the advancement of authoritarianism, they are Nazis.

We are way beyond Godwin’s Law at this point and it’s time to recognize and acknowledge that fact.

Post reply on HN