HNHacker News
TopNewBestAskShowJobs

morningseagulls

95 karma · joined August 30, 2019

submissionscomments
morningseagulls··on Jeff Bezos’s Master Plan
I like this detour into economics:

>Bezos’s enterprise upends long-held precepts about the fundamental nature of capitalism—especially an idea enshrined by the great Austrian economist Friedrich Hayek[, who] argued that no bureaucracy could ever match the miracle of markets, which spontaneously and efficiently aggregate the knowledge of a society.

>Amazon, however, has acquired the God’s-eye view of the economy that Hayek never imagined any single entity could hope to achieve. [...] With its logistics business—and its growing network of trucks and planes—it has an understanding of the flow of goods around the world. In other words, if Marxist revolutionaries ever seized power in the United States, they could nationalize Amazon and call it a day.

Perhaps we need to break up Amazon because communism? /s

morningseagulls··on The Differentiable Curry
Thanks for posting this.
morningseagulls··on Still Manufacturing Consent: An Interview with Noam Chomsky
>Mainstream exists in a form it is not because Mainstream translates the Truth™, but because there are no better theories yet.

Where have I heard that before? Oh yeah, the string wars in theoretical physics.

>Newton's mechanics is worse than Einstein theory in describing celestial bodies' movements, yet the former was better than anything until the latter appeared.

In its early days, the Copernican model was no better than the mainstream Ptolemaic model at describing the motion of celestial bodies:

For his contemporaries, the ideas presented by Copernicus were not markedly easier to use than the geocentric theory and did not produce more accurate predictions of planetary positions. Copernicus was aware of this and could not present any observational "proof", relying instead on arguments about what would be a more complete and elegant system.[0]

>I have a lot to argue against neoclassical economy, especially about its view of a person as a profit-driven (rational, as they call it) degenerate

I think you're confusing a person with a firm. A person has rational preferences[1] -- another unrealistic assumption -- while a firm seeks to maximise profits.

>The problem is that no heterodox theory provides both an axiomatic basis and empirical framework to compete mainstream economic science.

The idea that an axiomatic basis is necessary for a good theory is classic physics envy[2], and ignores the fact that even physics itself does not have an axiomatic basis, because that is Hilbert's sixth problem, which remains open.[3]

I don't know why this phenomenon is called "physics" envy anyway: it seems like you want economics to follow the model of Euclidean geometry, hence "mathematics" envy, which is even more absurd.

I get that mainstream economics has become more diverse, but neoclassical economics, which I believe is still the dominant school, is based on postulates that are simply unsupported by empirical -- and even mathematical -- evidence.

>What would you propose instead of Neoclassical economics? Nothing?

A critical synthesis of everything that isn't marginalism? An empirical search for a new framework with postulates backed by evidence?

And perhaps economists should learn more mathematics and have the courage to invent their own tools. After all, if you want to carry on with your physics envy, you should know that Feynman invented his own integral[4], which mathematicians derided for a long time for its lack of rigour... until its power became clear, at which point the functional analysts got down to work to extract the deep mathematics underlying Feynman's formalism.[5]

>It's not ignorance to accept theory that is not always true

This is an acceptance of Friedman's discredited positivism.

>it's stupidity to omit theory without having anything better

We have, and had, better: marginalism (aka neoclassical economics) was heavily criticised even during the 20th century. It's stupidity to cling to a theory when there are better ones.

[0] https://en.wikipedia.org/wiki/Copernican_heliocentrism#Early...

[1] https://en.wikipedia.org/wiki/Preference_(economics)#Applica...

[2] https://en.wikipedia.org/wiki/Physics_envy

[3] https://en.wikipedia.org/wiki/Hilbert%27s_sixth_problem#Stat...

[4] https://en.wikipedia.org/wiki/Path_integral_formulation

[5] https://ncatlab.org/nlab/show/path+integral

morningseagulls··on Reverse Derivative Categories
It's rare to see a CS paper, let alone a CS paper with category theory, with 7 authors!

It's fairly mathematical, so I'm going to quote the abstract for some context as to why you might care about this:

The reverse derivative[0] is a fundamental operation in machine learning and automatic differentiation. This paper gives a direct axiomatization of a category with a reverse derivative operation, in a similar style to that given by Cartesian differential categories for a forward derivative. Intriguingly, a category with a reverse derivative also has a forward derivative, but the converse is not true. In fact, we show explicitly what a forward derivative is missing: a reverse derivative is equivalent to a forward derivative with a dagger structure[1] on its subcategory of linear maps. Furthermore, we show that these linear maps form an additively enriched category with dagger biproducts.

[0] This is usually called the reverse mode of differentiation in the automatic differentiation literature: https://en.wikipedia.org/wiki/Automatic_differentiation#The_...

[1] https://ncatlab.org/nlab/show/dagger+category

morningseagulls··on The Real Reasons I Don’t Hire You
>My brother interviews for McKinsey and he says the number one thing they look for in new hires is a sense of understanding of someone else's mental state. He had a good technical term for it, but I forget what he called it.

>It's basically though the ability to not think, "this person is a jerk who is wrong," but "this person has his reasons for reaching his conclusions and although I am angry with him after our disagreement, I see the personal factors that must have gone into it."

Avoid the fundamental attribution error?[0]

[0] https://en.wikipedia.org/wiki/Fundamental_attribution_error

morningseagulls··on Dutch family 'waiting for end of time' discovered in basement
>I'm pretty sure in Australia if you drove 90 seconds from the capital you'd (still) be in the middle of nowhere.

Lol. The capital is in the middle of nowhere, that's why. You need to drive a few hours northeast of the capital to get to somewhere (i.e. Sydney).

morningseagulls··on Dutch family 'waiting for end of time' discovered in basement
>> One time I spent a year where I didn't say anything out loud to anyone.

> How did you get food?

Not the parent, but this isn't hard: self-checkouts are common these days, and even if you take food deliveries, you don't pay the delivery person directly, you pay through the app.

morningseagulls··on Dutch family 'waiting for end of time' discovered in basement
>Curious what particular eschatology drove them to do this, the article doesn't mention any. Now I'm imagining that it wasn't tied to any specific religion or prophecy, but just the concept of time somehow coming to a stop.

New reports[0] are saying the father used to be a member of the Unification Church:

>Local television station RTV Drenthe earlier said the Austrian man, identified only as Josef B, the father and the captive family were all part of South Korea’s controversial Unification Church, dubbed “Moonies” after their late founder Sun Myung Moon.

[0] https://www.theguardian.com/world/2019/oct/17/dutch-police-a...

morningseagulls··on Dutch family 'waiting for end of time' discovered in basement
>The father and six kids hadn't been allowed out for nine years. If this isn't a cult, it's a huge abduction story.

It's now been reported[0] that the father was an ex-member of the Unification Church[1]:

>Local television station RTV Drenthe earlier said the Austrian man, identified only as Josef B, the father and the captive family were all part of South Korea’s controversial Unification Church, dubbed “Moonies” after their late founder Sun Myung Moon.

[0] https://www.theguardian.com/world/2019/oct/17/dutch-police-a...

[1] https://en.wikipedia.org/wiki/Unification_movement

morningseagulls··on Still Manufacturing Consent: An Interview with Noam Chomsky
>Let's take electrical engineering as an example.

I'd usually say you're shifting the goal posts, perhaps muse about how dehumanising and simplistic it is to think of humans as components in an electrical circuit, and move on. Then I thought that it'd be more illuminating to address your analogy, because your reasoning suffers from the same problem that I've been talking about.

>If you model an electrical circuit, you assume that there simply is a voltage source where electricity comes from. That assumption is of course complete nonsense: In reality, electrical energy is not just there, it has to be generated somewhere, for example in a power plant. The electricity also flows through a grid that is shared with many other consumers.

I think it might help to read Musgrave's paper[0] (the abstract will do if you can't get the paper itself) on the three kinds of unrealistic assumptions, in which he attacked Friedman's twist. The assumption you pointed out is the most benign type: a negligibility assumption. It's like assuming that a object falling through air experiences no friction: complete nonsense, but air friction exerts only a negligible effect on most falling objects, unless the object is, say, a feather.

In the case of your example, taking into account the power plant adds nothing to the analysis, because the electrical circuit you're concerned with usually doesn't output electricity. Usually. But in the 21st century, that can change...

>So, according to your requirement to always use the more realistic model, electrical engineers should always model a whole power plant and an entire power grid in each circuit, because otherwise their model would be bullshit.

Firstly, I did not say the more realistic model should "always" be used. The situation in economics is more like the more realistic model is ignored in favour of the incorrect unrealistic model, even when the more realistic model is appropriate. The analogy with Newtonian/Einsteinian mechanics is quite shaky here, since as I've said, Newtonian mechanics is vastly more accurate than neoclassical microeconomics.

To take another example from engineering and continue the theme of relativity, error corrections required for GPS depend on an understanding of relativity.[1] If, as you've previously suggested, the simpler and more easily understandable model should be preferred, then we would not have GPS and allied technologies today.

Secondly, in the 21st century when we have renewable energy, electrical engineers are now required to take into account a power grid, because consumers are now also power plants if they have solar power, for example, and want to feed the excess power generated into the power grid.

The point is this: if the circuit you're modelling is a passive one that doesn't generate electricity, sure, make use of the unrealistic assumption that there's an unknown voltage source. However, if your circuit is also a voltage source, that assumption is then false, and you need to update your model.

The same should apply in economics. Unfortunately, it turns out that markets and economies are far more complex systems than power grids: as soon as you have more than one agent (e.g. humans, corporate entities, HFT algorithms, etc.), the simplistic analysis derived from marginal utility theory fails. So, outside of the laboratory perhaps, the models used in neoclassical microeconomics are probably almost never appropriate for the problems they're treating.

>> Yes, it does.[2]

>Do you know of any empirical studies that show that the models currently used in macroeconomics actually cause problems with real data?

This is a really garbled question. Firstly, nowhere did I assert that economic models "cause problems with real data". I really don't understand this assertion and would welcome any clarifications.

Secondly, I think you misunderstand the point of my argument about the SMD theorem. The point is that, as soon as you have more than one agent/consumer (human, firm, algorithm, what-have-you), they will interact in a non-linear way, giving rise to a nonlinear market demand function that gives a "weird" market demand curve.

This has nothing to do with any effect that the belief in a certain economic model would have on the economic behaviour of an agent. That is an entirely separate question altogether.

Back to your question about whether "weird market demand curves are possible" only "in a few exceptional situations". The SMD theorem basically says no: in all but the most exceptional cases, multiple market equilibrium points can exist.

The implication is that the aggregation problem[3] is not solved by the neoclassical assertion that the market demand curve for a well-defined commodity has the same pleasant mathematical properties -- in particular, that it is monotonically decreasing ("downward-sloping"), i.e. obeys the "law" of demand -- as the demand curve for that commodity of a rational agent.

And the reason why the aggregation problem isn't solved is because, as soon as you have more than one agent or commodity (like in most IRL markets), you get a complex system. And when you have a complex system, you can't draw line diagrams like you do in the textbooks. You need to do modelling and numerical analysis. There's a field called complexity economics[4] that's only beginning to do that, some 40 years after the SMD theorem appeared. Better late than never, but it's still a lot later than the other less "dismal" sciences, which have come to realised that they're dealing with complex systems and have updated their thinking accordingly.

[0] https://onlinelibrary.wiley.com/doi/abs/10.1111/j.1467-6435....

[1] https://en.wikipedia.org/wiki/Error_analysis_for_the_Global_...

[2] https://en.wikipedia.org/wiki/Sonnenschein%E2%80%93Mantel%E2...

[3] https://en.wikipedia.org/wiki/Aggregation_problem

[4] https://en.wikipedia.org/wiki/Complexity_economics

morningseagulls··on Still Manufacturing Consent: An Interview with Noam Chomsky
>Maybe the simpler model is enough to understand what is going on. Maybe a more complex model suffers from over-fitting and performs worse than the simpler model. It really depends.

If only physics departments everywhere would adopt this philosophy! Physics undergrads would then be liberated from the yoke of Einstein's relativity, let alone the tyranny of the mind-blowing quantum field theory with its insufferably complicated Feynman diagrams.

It does not depend. The simpler model in economics is fatally flawed, and is the one suffering from over-fitting (to a straight line, no less) and contributing to the bad reputation of economics as a "dismal science".

>> The SMD theorem guarantees, for example, that most market demand curves will NEVER satisfy the fundamental neoclassical "law" of demand.

>Does it really say that? Or simply that weird market demand curves are possible in a few exceptional situations?

Yes, it does.[0] The exceptional situations are precisely those that neoclassical economics assumes are the norm, namely, that there is only one agent and one commodity in the market.[1]

As soon as you have more than one agent/consumer in an economy, the mathematics will undermine the "law" of demand, because the inter-agent interactions would generate non-linear terms in the market demand function.

And once you have terms of higher order, the demand curve generated by that function will have sections that slope upwards, exactly what is being forbidden by the "law" of demand.

And we haven't yet taken into account the very real fact that multiple consumers cannot possibly have the same preferences. Or that these preferences can change with different levels of income.[2]

In short, neoclassical microeconomics only works in a communist economy of identical clones with identical preferences and incomes.

Oh, the irony.

[0] https://en.wikipedia.org/wiki/Sonnenschein%E2%80%93Mantel%E2...

[1] https://en.wikipedia.org/wiki/Representative_agent

[2] https://en.wikipedia.org/wiki/Homothetic_preferences

morningseagulls··on Still Manufacturing Consent: An Interview with Noam Chomsky
>Mainstream science can't have one, only marginal branches can.

This is like native speakers of a language (say, English) asserting that only non-native speakers of their language have accents. You do, you just speak with an accent that's considered to be standard.

>This kind of intellectual dishonesty could come only from a ideology-loaded follower of some heterodox (and usually pseudoscientific) theory, most likely grounded in some ideology (guessing from value judgments of op like "neoliberal" -- left one) rather than verified by empiric data.

And mainstream (i.e. neoclassical) economics[0] does that, even in the face of mathematical reality, let alone empirical data.[1] The ignorance of the implications of the SMD theorem,[2] because it implies that the aggregation problem remains unsolved and the fundamental breakdown of the "law" of demand, was a profoundly anti-intellectual move.

[0] https://en.wikipedia.org/wiki/Neoclassical_economics

[1] https://news.ycombinator.com/item?id=21269064

[2] https://en.wikipedia.org/wiki/Sonnenschein%E2%80%93Mantel%E2...

morningseagulls··on Still Manufacturing Consent: An Interview with Noam Chomsky
>Watch out for moving the goalposts.

Says the one who's been moving the goalposts in the first place. The point of this thread was about whether or not the Nobel prize in economics was anti-intellectual and ideologically biased. You chimed in with:

>>>The D in SMD was actually awarded the nobel prize in economics.

Which is why I responded, because this was a non-sequitur and an appeal to authority (of the people awarding the prize, and the question was whether or not these people were biased).

>He was awarded the nobel prize after all.

That means nothing in the context of this thread, and is another appeal to authority.

>"Rigorous reformulation" does not sound like skirting at all.

The SMD theorem was a big deal, yet Debreu wasn't awarded the prize for that, but for a rigorous reformulation of Walrasian equilibrium,[0] which has very strong assumptions.

In particular, that rigorous reformulation, known as the Arrow-Debreu model,[1] fails under weaker (i.e. more realistic) assumptions, because of the SMD theorem.

This is skirting the issue.

>Economy (or climate or orbits) exhibit chatoic and complex behaviour and it is not surprising that they can have multiple stable points (look up bifurcation diagrams for a simple representation of chatoic solutions). If I am not mistaken, one can hear such things routinely in university economy classrooms today especially when giving caveats about models being wrong in general. Even when you include game theory and Nash equilibria with respect to economic solutions, there might not be efficient ways to reach the equilibrium points at all!

A few points:

- It's nice and all that caveats are being given, but what tools are being taught to economics undergrads to handle chaotic and complex behaviour? Please supply links to some university course syllabus as evidence (I'm happy to read material in a language other than English).

- If economics students aren't being taught tools to handle chaotic and complex behaviour, then giving caveats is just paying lip service.

- Game theory and Nash equilibria are tools to deal with equilibria, which is what neoclassical economics is comfortable with. To be more realistic, you need to look at what happens in disequilibria, as you've pointed out yourself.

- Chaos theory also showed that equilibria may never be attained at all by a system, so the obsession with equilibria in economics is, at best, unproductive.

>That is to assume that every equilibrium solution you get is tied to a specific policy or a set of them (or even more, an ideology!) or that points can be reached easily which are very strong assumptions. Maybe some ideologies don't exhibit any reachable equilibrium points or pertain only to unstable ones.

Agreed. Unfortunately, neoclassical economics, thanks to its underlying ideology, assumes that equilibrium points can be reached by the "free market" and are desirable.

[0] https://en.wikipedia.org/wiki/Competitive_equilibrium

[1] https://en.wikipedia.org/wiki/Arrow%E2%80%93Debreu_model

morningseagulls··on Still Manufacturing Consent: An Interview with Noam Chomsky
>I think its still fair to say that economics is not a science the way medicine is.

Which is very unfortunate, because like medicine, the practice of economics has the power to influence decisions and affect lives.

morningseagulls··on Still Manufacturing Consent: An Interview with Noam Chomsky
>Protectionism in economic sense are measures that inhibit international trade. Tariffs, selective subsidies and too restrictive immigration policies are good examples. Maybe in Australia they use the word differently

I'm not sure which part of the Australian politics I've described doesn't fit your definition of what protectionism is.

I've certainly gone into great length about restrictive immigration policies: in fact, one of the first pieces of legislation of the Barton government was the 1901 Immigration Restriction Act,[0] the passage of which was a condition of Labor's guarantee of confidence and supply for the minority Protectionist government.

The 1901 election itself[1] was straight from the textbooks: Protectionist vs. Free Trade vs. Labor. Protectionism (Protectionists and Labor) won that election, both in terms of seats and popular vote. You couldn't get a more literal election than that.

As for tariffs and selective subsidies, Australia has that too.[2][3][4]

>The function of unions is decidedly NOT to “protect the labour market”, but to provide organised market power to a group of labourers. A union fulfils its function even if it secures jobs for its members at the cost of say workers of another trade or those who are not unionised. Hence a union can satisfy its members but reduce the overall employment.

You're presuming a fragmented union ecosystem, which isn't quite the reality in Australia. The Labor party was founded as the political arm of the unions; this shows a high enough degree of co-ordination amongst the unions at the very least.

And once you have a party in parliament, you get the power to influence policy-making in such a way as to protect the labour market. A single union may only be able to protect its own members, but a well-coordinated group of unions is well-positioned to protect the entire labour market itself. Labor did that in spades in the early days of federation, and still does to this day.

>At first you claim that the 90s boom was due to protectionist policies. Now you claim that it’s because of personal computers. Which one is it then?

I did not. You wrote:

>>yet the productivity growth in the 1990s was followed by Labour winning majority in the parliament. What am I missing?

I pointed out that you missed the fact that Labor was in power from 1983 to 1996.

I then made the facetious statement that perhaps it was Labor's fault that the 90s boom happened: this was a joke because, clearly, nobody blames a party for a good economy.

I then went on to suggest that the productivity growth may perhaps be more reasonably attributed to the proliferation of PCs, and the attendant automation. There are other plausible factors for the boom: the economic rise of China must surely have been one.

In any case, it's simply not true that the 90s boom was due to protectionist policies. Labor during the Hawke-Keating years was uncharacteristically enthusiastic about reforms, and I think the success of their policies may have inspired the Third Way of Blairite Labour.

[0] https://en.wikipedia.org/wiki/Immigration_Restriction_Act_19...

[1] https://en.wikipedia.org/wiki/1901_Australian_federal_electi...

[2] https://en.wikipedia.org/wiki/History_of_tariffs_in_Australi...

[3] https://en.wikipedia.org/wiki/Parallel_import#Australia

[4] https://en.wikipedia.org/wiki/Australia_Tax

morningseagulls··on Meetup.com alternatives
>When I hosted meetups, paying $20/mo wasn’t the end of the world.

>So just give the $2 back to each attendee in that case, solved.

Under the new model, organizers will be paying $2/mo for every attendee that RSVPs. For meetups attracting a large audience, this easily comes to >$200/mo, which is a 1000% increase.

>I think for meetups where the hosting companies are getting value, ie: recruiting at tech meetups, it’s reasonable for hosts to absorb attendee costs.

>Now, finding cost effective space for large groups was very hard.

Many large meetups are run by volunteers who have to seek out companies to sponsor the space, which you've already acknowledged is hard to do. $20/mo is affordable for them and can be done without seeking out additional sponsorship; $200++/mo isn't.

Ultimately, I think organizers will self-select: those who can afford Meetup.com's model, or who're already charging attendees, will stay; those who can't afford the new charges and don't believe in charging attendees will leave. This dissatisfaction isn't an overnight thing anyway, and the new pricing policy will be a push factor for some to do their own thing.

morningseagulls··on Still Manufacturing Consent: An Interview with Noam Chomsky
>Wait, but unions are not necessarily protectionist.

Please define what you mean by "protectionist".

In what sense are unions not protectionist? Their sole function is to protect the labour market. I'd view unions in Australia as really protectionist: they've managed to secure very high minimum wages, as well as highly restrictive immigration policies that Trump praised to high heaven.[0]

>Also the protectionist government you are referring to the cabinet which was active until 1903

Their immigration policy[1] lasted for decades, before being replaced by the current immigration policy. The Protectionist Party[2] hung around until 1909, when they merged into the Commonwealth Liberal Party. This party was a predecessor of the current Liberal Party:

Protectionist + Anti-Socialist -> Commonwealth Liberal

Commonwealth Liberal + National Labor -> Nationalist

Nationalist + dissident Labor + Australian -> United Australian -> Liberal

So there is still protectionism hanging around in both major political parties today.

>the productivity growth in the 1990s was followed by Labour winning majority in the parliament. What am I missing?

It's spelled Labor, because modern spelling.[3] And Labor was in power from 1983 to 1996 (Hawke[4], then Keating[5]), so that productivity growth in the 1990s... probably Labor's fault? Lol. I think the proliferation of personal computers may have had more to do with it.

[0] https://www.abc.net.au/news/2019-06-27/donald-trump-tweets-s..., https://www.theguardian.com/us-news/2017/mar/01/donald-trump...

[1] https://en.wikipedia.org/wiki/White_Australia_policy

[2] https://en.wikipedia.org/wiki/Protectionist_Party

[3] https://en.wikipedia.org/wiki/Australian_Labor_Party

[4] https://en.wikipedia.org/wiki/Bob_Hawke

[5] https://en.wikipedia.org/wiki/Paul_Keating

morningseagulls··on Still Manufacturing Consent: An Interview with Noam Chomsky
>I’m SUPER interested how protectionist capitalism has helped the working class.

Australia is an example of a protectionist capitalist society: in fact, its first federal government was formed by a party called the Protectionist Party, with the support of the union-based Labor Party.[0] Unions are relatively strong there, compared to the rest of the world, and Australia's immigration policy is pretty tight, which protects workers there from foreign competition. Consequently, workers have some of the highest (minimum) wages in the world, and enjoy relatively excellent working conditions. The economy has also been doing well since the 1990s, despite the GFC and everything.

The EU is probably another example: strong labour laws with stringent labour market tests and excellent working conditions, as well as stringent requirements on a wide range of commodities that has to be satisfied by imports.

[0] https://en.wikipedia.org/wiki/Barton_Government

morningseagulls··on Still Manufacturing Consent: An Interview with Noam Chomsky
>In physics, there is even the metaphor of the spherical cow[1] to make fun of the need to simplify things in order to understand them.

It's a joke to remind us of a sobering fact: that unreal assumptions lead to unreal results.

>Just because a model is simple doesn't mean that it is bad or worse than a model with more complex assumptions.

But if a model with more complex assumptions is a better approximation to reality than a model with simpler assumptions, would you not adopt the former?

What has happened in economics is that people have clung on to, say, Newtonian mechanics, instead of embracing Einstein's relativity.

Even this does not quite adequately express the enormity of the inertia that you see in economics, because at least Newtonian mechanics is right most of the time, whereas neoclassical economics is wrong most of the time. The SMD theorem guarantees, for example, that most market demand curves will NEVER satisfy the fundamental neoclassical "law" of demand.

Let me also point out that your comment is rehashing Friedman's positivism.[0] Even in quantum mechanics, which has for decades been dominated by the positivism of the Copenhagen interpretation, people are now transcending that positivism because they've come to the realisation that quantum mechanics can't be advanced without overcoming the crutch of positivism. And Friedman's assertions about assumptions are unsound anyway.[1]

[0] https://en.wikipedia.org/wiki/Essays_in_Positive_Economics

[1] For a critique of Friedman's argument about assumptions in models, see https://onlinelibrary.wiley.com/doi/abs/10.1111/j.1467-6435....

morningseagulls··on Still Manufacturing Consent: An Interview with Noam Chomsky
>The D in SMD was actually awarded the nobel prize in economics.

Debreu's work on general equilibrium, which was the basis of the award, had to skirt around the SMD theorem:

>Prize motivation: "for having incorporated new analytical methods into economic theory and for his rigorous reformulation of the theory of general equilibrium."[0]

In order for the Arrow-Debreu model[1] to give a unique equilibrium, which is usually desired, some strong (arguably, unrealistic) conditions must be assumed, otherwise there can be multiple equilibria, as guaranteed by the SMD theorem.[2]

In particular, the SMD theorem asserts that any polynomial function can be a market demand function, which fundamentally breaks the so-called "law" of demand in neoclassical economics. This is because most polynomial functions are not monotonically decreasing[3], as required by the "law" of demand, but give rise to demand curves that can curve up or down arbitrarily. As an example of such a function, just think of the curve generated by a generic cubic polynomial.[4]

This "anything goes" situation then leads to non-unique equilibria, since a supply curve can now intersect the market demand curve at more than one point. But since an equilibrium point is supposed to be where welfare is maximised in a society, the existence of non-unique equilibria means that there can be multiple economic arrangements under which social welfare is maximised. Then it's a matter of taste (i.e. "ideology") to decide which equilibrium point a society wants to be in.

[0] https://www.nobelprize.org/prizes/economic-sciences/1983/deb...

[1] https://en.wikipedia.org/wiki/Arrow%E2%80%93Debreu_model

[2] https://en.wikipedia.org/wiki/Sonnenschein%E2%80%93Mantel%E2...

[3] https://en.wikipedia.org/wiki/Monotonic_function

[4] A graph of such a curve is shown here: https://en.wikipedia.org/wiki/File:Monotonicity_example3.png

morningseagulls··on Deep Learning with PyTorch: A 60 Minute Blitz [video]
>To be fair, the main thing you learn doing the cryptopals challenge is to not write your own crypto.

D'oh! Good point though lol.

morningseagulls··on Meetup.com alternatives
>Since you have no contact information for any of your attendees, you can't reach out to them to move them to another platform.

>It would be nice if Meetup could give you the contact information of all your members

Meetup has a spam problem, as greggman2 has pointed out.[0] Your proposal would make that worse.

>This allows someone else to step in and take over. Taking all the hard work you did to create the meetup and giving it to someone else.

Believe it or not, some people think this is a feature. I know of groups where the original organizer(s) had to leave the locality for various reasons, but designated successor(s) to keep the group going.

>allow you to close a group without allowing anyone else to step in.

Yes, that'd be good to have.

>They hold your meetup group hostage if you're an organizer.

Yes, all these contributed to lock-in, but some of these were features that some people wanted in the past, or at least that's my impression from using the site for many years.

[0] https://news.ycombinator.com/item?id=21266422

morningseagulls··on Meetup.com alternatives
>How would I "discover" events on meetup.com or another event service?

Meetup has a search engine that allows you to find events by date, location, or subject matter. So I could

- look up events near me for this coming week.

- look up events within a certain radius from a location.

- look up events based on my interests.

> if I'm looking for an event I Google.

That's the fallback. I think what Meetup provided was a single site that people visit to find local events that fit their interests, which also allowed community-building.

morningseagulls··on Meetup.com alternatives
>However, I feel the discoverability isn't there. The local UX meetup will likely choose one platform. The local frontend developers meetup will choose another.

I think discoverability can be had via Google events search, Facebook, Eventbrite, etc.

Yes, there will be fragmentation, but I know people have found out about Meetup events via Eventbrite, and people have mentioned these on HN as well. So fragmentation is status quo anyway.

>I've always liked how easy it is for a local developer to just stumble across my group.

You could do what some organizers have been doing: advertise your events on multiple channels.

morningseagulls··on Meetup.com alternatives
>Don’t have $2? Really? Even meetups for spiritual teachers who operate as a non profit suggest a $20 donation per event for instance.

One could counter that these people are the ones who've spoilt the market for the rest of us.

And really, it's not attendees who're complaining, it's organizers who believe in keeping their events free for attendees.

morningseagulls··on Meetup.com alternatives
>If a meetup is not even worth $2 to attend, is it really worth an hour+ of your personal time?

Considering it's mostly organizers who've been the most vocal about it, this is a pretty strange question.

morningseagulls··on Meetup.com alternatives
>Cannot relate to the people here who are screaming bloody murder over this. You have to give to get.

Perhaps you can't relate because "the people here who are screaming bloody murder" are mostly not attendees: they're organizers.

And they've been doing a lot of giving ever since Meetup started charging organizers.

morningseagulls··on Deep Learning with PyTorch: A 60 Minute Blitz [video]
>But the OP clearly didn't ask about doing it as a challenge or to understand how it works (which is what cryptopals is about), but actual usage.

But the parent of the comment I was replying to clearly had the former in mind, as a subsequent comment showed.[0]

[0] https://news.ycombinator.com/item?id=21240429

morningseagulls··on A Code Glitch May Have Caused Errors in More Than a Hundred Published Studies
>>> Most professors are unneeded given the size of the problem space.

>most subjects are too deep for the professor to be knowledgable about more than just their specialty

>in many cases, the size of the problem space is much too big for a single group of researchers to have any effect on it.

>most studies do not have the manpower available to do a completely formally-correct ideal study

I agree that the size of the problem space requires a lot of researchers. This is probably why there's so much attention focused on machine learning and big data, since they seem to have the potential to address the problems you've listed here. Of course, there are many technical and ethical issues to be confronted in developing and deploying them.

>If research were _truly_ a priority for humanity, and we really dumped all of our effort into scientific research as a society (i.e. governments and companies both prioritized R&D and gave the scientists enough resources to actually do the jobs properly)

It's not, because the greatest problem for humanity is still subsistence, which requires solving a massive resource distribution problem. There are some governments and some companies that do prioritise R&D and provide enough resources, but they are too few and far between.

>psychological, sociological, medical, and most other fields of research that aren't mathematical, are considered dubious.

>then we might see these fields as "hard science" rather than "soft science".

It seems to me that these fields are "soft" in part because the ethical issues surrounding the surveillance that's needed to collect the data to do a formally correct study are quite formidable.

morningseagulls··on Seizing the Means of Knowledge Production
>The modern product of the American university is generally skeptical of anything resembling a meritocracy

A little-known fact about the word "meritocracy"[0] is that it's a recent invention, coined by Baron Young of Dartington[1] in his satirical work[2] outlining how dystopian a society organised as a meritocracy would be.

[0] https://en.wiktionary.org/wiki/meritocracy

[1] https://en.wikipedia.org/wiki/Michael_Young,_Baron_Young_of_...

[2] https://en.wikipedia.org/wiki/The_Rise_of_the_Meritocracy

← PreviousPage 2 of 4Next →