Hacker News Reader

Comment by greendave | Hacker News Reader

"In a grotesque way, Florida’s disastrous UI system represents the successful completion of a project: They intended to build a UI system that failed those who needed it, and they very much did."

This is the lens that needs to be applied more generally. Just because a system is not achieving its stated goal does not necessarily mean it is failing - it often means that the true goal is simply something different. To take an obvious example, the US healthcare system is not failing in its primary mission, which is to transfer wealth from the population at large to a specific group. It is in fact spectacularly successful.

Andreesen's article entirely ignores this. Things are not the way they are by accident. Ignore the reasons why and you are either being disingenuous or ignorant.

View on HN

Strong concur - put simply, i think many people should recognize bad UI, in any form, as just an elaborate abstraction of hardball haggling.

If a company makes it HARD to get hired, it's probably to get you to accept an undermarket rate or sweeten the deal.

If a boss makes it confusing to understand how the promotion process and compensation process works, it's probably to keep you from getting promoted and keep you in your seat to do your job and not complain

This misses the point.

It could be well argued - or even simply stated - that, yes, US healthcare system has in fact primary mission to benefit health of US population. How'd you argue it's not the case? I, for one, know this is my idea of the system. Will you prove that those who proposed the system had other ideas? Can you get in their heads? Can you get in the heads of those implementing the system? They'll perhaps bring enough words trying to convince you otherwise - and some would be sincere.

The problem is that a big system is a collective work - and different agents have different goals in mind. Sometimes they can have compromises - agreements on goals, partial or final. Sometimes they can trick each other into doing something which benefits some agents' goals, but not others. You can't make a big system realistically without many agents - so how can you make sure they all work with similar goals?

We talk a lot about possibilities, but sometimes not enough about intentions - and especially verifiable goals. Politicians can promise many things, but it could be expensive to elect somebody who's stating something meaning something completely different - and this idea holds on various levels of public (not only public - see e.g. office politics) decision making.

> It could be well argued - or even simply stated - that, yes, US healthcare system has in fact primary mission to benefit health of US population. How'd you argue it's not the case?

No, that's the entire point greendave was trying to make. The system's goal has been perverted or coopted. This is a logical result of a corrupted system.

Relevant imagery: zerg infestation.

> The system's goal has been perverted or coopted.

I think that's a good description. We can use this model to work on fixing the problem.

I don't think you should judge a system's purpose by the stated intentions behind it, you judge it by outcomes. By observing the outcome of a system (or business, or legislation), you can draw a conclusion about its actual purpose.

EDIT: milesokeefe's Wikipedia link says this better than I could.

I built a car to drive around. Car broke at a nearest intersection. We, however, not talking about car' intentions - we talk about my intentions, and my intentions weren't to stop that soon.

> I don't think you should judge a system's purpose by the stated intentions behind it, you judge it by outcomes.

I do judge by outcomes. I also judge purpose - which is something which was planned, desired for before the system was built - by how I can find that purpose. If somebody tried to save a drowning person by throwing him a rope, but missed and the person died, that doesn't mean that somebody was trying to kill the person. If a system doesn't produce effect A, that doesn't mean it's not trying - as system's creator intended.

The Wikipedia article is ok, if we differentiate intents of a system - which could be a special thing for non-humans - and of system creators.

This is not a machine that normally does one thing, but in an exceptional case fails and does something else. We're talking about a system that in normal operation, during steady state, does X and has outcome Y. That is its purpose.

Complex multiparticipant systems don't have goals, other than to perpetuate their own existence. The health care system isn't conscious. It's the collective interaction between a number of constituent organizations: health insurers, private hospitals, doctors, taxpayers, government, patients, employers, etc. Some of them may have individual executives making decisions over part of the system, but if those executives' personal goals are significantly misaligned from "transferring money from the population to their organization" then the organization that pays them would've long since gone under. Hence the observed behavior.

Similar evolutionary dynamics can be observed in all complex systems - life, biomes, religions, community, government, law, capitalism, tech companies, etc. Any system that does not optimize for its own survival will be replaced by systems that do. And the way to align that with the continued benefit of the humans that the system was supposed to benefit is to make systems "default dead", and then hinge their continued existence on them providing the benefit that they were created for. Unfortunately this is easier said than done, because another human evolutionary impulse is the drive to create systems that will continue to exist and benefit the person who created them (though not necessarily stated collaborators) without any additional effort.

Moloch! Solitude! Filth! Ugliness! Ashcans and unobtainable dollars! Children screaming under the stairways! Boys sobbing in armies! Old men weeping in the parks!

Moloch! Moloch! Nightmare of Moloch!

https://slatestarcodex.com/2014/07/30/meditations-on-moloch/

It's hard to fix broken societal incentive structures when you're operating from within the society itself. That doesn't mean it's not worth trying though and sometimes you can change the incentive structure and fix the behavior, or you can still have some success within it and then leverage new found power to fix things.

I think societal/cultural optimism matters here too - since believing change is possible helps fuel the coordination that can actually make change possible. I think it's an advantage the US has over more culturally pessimistic countries (and that Silicon Valley has over other more culturally pessimistic regions).

>It's hard to fix broken societal incentive structures when you're operating from within the society itself.

"The master's tools will never dismantle the master's house."

as @wpietri would say, posiwid

https://en.wikipedia.org/wiki/The_purpose_of_a_system_is_wha...

Thanks. Had not seen that acronym before.

> To take an obvious example, the US healthcare system is not failing in its primary mission, which is to transfer wealth from the population at large to a specific group. It is in fact spectacularly successful.

The goal of every business is to transfer wealth from the population to the business. It is a capitalist[0] society's job to ensure a system that aligns those goals with society's. When people refer to the healthcare system failing, they mean the system is not doing a very good job of delivering on society's goal's.

The abstract healthcare system itself doesn't have any goals. Doctor's, hospitals, and insurance providers all have goals, but many times they no only conflict with society but also eachother.

[0] - In a non capitalist society you are trading the "aligning firms with society" problem for the far more difficult "aligning individuals with society".

Reply on news.ycombinator.com