TaurenHunter

quantile function for the Poisson-inverse Gaussian?

Related: there is a fun to watch YT channel hyper-specialized in diecast car races https://www.youtube.com/watch?v=vRMUZDVwmC0

The author seems to have drawn from

- Buddhism's mindfulness

- Taoism's natural flow/simplicity

- Stoicism's control of own's mind

- Mindfulness teachings from Jon Kabat-Zinn and/or Thich Nhaat Hanh

- Cognitive Behavioral Techniques

- Mihaly Csikszentmihalyi's flow states

- Literature/poetry from Henry Thoreau and Ralph Waldo Emerson

We will need a paper titled '"Considered Harmful" Articles is All You Need' to complete that cycle.

Donald Rubin is kind of a modern day Leibniz...

Rubin Causal Model

Propensity Score Matching

Contributions to

Bayesian Inference

Missing data mechanisms

Survey sampling

Causal inference in observations

Multiple comparisons and hypothesis testing

Reminds me of the "Considered Harmful" articles:

https://meyerweb.com/eric/comment/chech.html

There is at least one MMA fighter who practices that therapy: https://www.mixedmartialarts.com/ufc/machida-explains-the-be...

I kinda like what they did in https://maroofy.com/ in that it lists songs very close to what I specify.

However I would prefer a service that allowed me to tell what I don't like and then use that preference to filter out everything similar to it.

I once visited a local Patel Brothers store and it had what looked like dried cow dung patties on the shelf.

Later I've found out it is used for ritual burning and, since that is difficult to obtain in an urban setting, online retailers are selling it too.

Even found a review on Amazon of a person complaining that it tasted terrible.

I feel like Apple said the quiet part out loud: they intend to replace creativity and all that surrounds it with a consumer-ready wafer-like device, sort of a Soylent Green replacement for human ingenuity.

"...the biggest bottleneck of KANs lies in its slow training. KANs are usually 10x slower than MLPs, given the same number of parameters. We should be honest that we did not try hard to optimize KANs’ efficiency though, so we deem KANs’ slow training more as an engineering problem to be improved in the future rather than a fundamental limitation. If one wants to train a model fast, one should use MLPs. In other cases, however, KANs should be comparable or better than MLPs, which makes them worth trying. The decision tree in Figure 6.1 can help decide when to use a KAN. In short, if you care about interpretability and/or accuracy, and slow training is not a major concern, we suggest trying KANs."

Longest dictator rule so far:

1 Fidel Casto Cuba 52

2 Chiang Kai-shek Republic of China 47

3 Kim II-sung North Korea 45

4 Yumjaagiin Tsedenbal Mongolian People's Republic 44

5 Muammar Gaddafi Libyan Arab Republic 42

6 Paul Biya Cameroon 41

7 Omar Bongo Gabon 41

8 Enver Hoxha People's Republic of Albania 40

9 Mohamed Abdelaziz Sahrawi Arab Democratic Republic 39

10 Francisco Franco Spanish State 39

11 Eamon de Valera Ireland 37

12 Teodoro Obiang Nguema Mbasogo Equatorial Guinea 37

13 Gnassingbe Eyadema Togo 37

14 Jose Eduardo dos Santos Angola 37

15 Robert Mugabe Zimbabwe 37

16 Josip Broz Tito Yugoslavia 36

17 Antonio de Oliveira Salazar Second Portuguese Republic 36

18 Todor Zhivkov People's Republic of Bulgaria 35

19 Ali Khamenei Iran 35

20 Alfredo Stroessner Paraguay 34

That reminds me of what Ricardo Semler did at Semco: distribution of authority, employee participation in decision making, flattening of organization, rotating leadership roles, ...

https://semcostyle.com/ricardo-semler-creating-organizationa...

https://mallenbaker.net/article/inspiring-people/ricardo-sem...

From the abstract:

"Researchers have long associated collectivistic culture with harmony and cooperation. However, the bulk of the evidence suggests that collectivistic cultures compete more, and more intensely, than individualistic cultures.

Collectivists are more likely to see competition as zero-sum, engage in social comparison, and base their self-worth on common standards rather than self-defined goals. This raises a paradox: where does the popular conception of harmony in collectivism come from?

In reviewing prior studies, we find that people in collectivistic cultures tend to use indirect, hidden methods to compete against others. This allows for an outward harmony, without negating competition. We ask whether competition in collectivistic cultures is only stronger when competing with outsiders. Studies reject this speculation.

Rather, people in collectivistic cultures compete more with in-group members and are more vigilant toward classmates and co-workers.

Next, we explore how people from different cultures decide to enter into competition. We find that collectivists’ tendency to enter into prestigious competitive environments might end up harming them.

Finally, we discuss whether there can be versions of collectivistic groups without competition or whether this is a utopian dream."

> Aylo is owned by Canadian private-equity firm Ethical Capital Partners

"... we have suggested that modulation of the ANS (autonomic nervous system) via slow breathing techniques in adjunct to relaxation techniques and sleep hygiene may be a more powerful tool in combating insomnia than the prevailing method of using hypnotics and other pharmaceutical interventions."

For some reason that doesn't work for me - it just times out.

Google cache did: https://webcache.googleusercontent.com/search?q=cache:https:...

Reminds me of a recent post about getting a message in Slack with a link to another app about a meeting in Zoom, ... and a number of steps involving a bunch of other apps each costing $15/month/user. In the end nobody does anything but move small amounts of data through those apps.

Northrop Frye in "The Great Code: The Bible and Literature" has done a related analysis in the 80s.

I think he identifies 2 primary Biblical narratives - U-shaped plot and inverted U plot - and their role in shaping Western literary and cultural tradition.

U-shaped plot is a descent into a difficult or chaotic situation that represents a fall from a higher state, then proceeding through a series of complications and conflicts that further the sense of alienation or suffering until there is a turning point that leads to a rise in fortune or circumstance, with a positive resolution and the protagonist achieving some form of redemption, success, or happiness.

This could be a literal descent, like being cast out of Paradise, or the Prodigal Son, Job, or Joseph of Egypt (thrown in a well, sold as slave, then in prison), or a metaphorical one, such as losing one's way morally or spiritually.

Inverted U plot examples: king Saul, Samson.

These archetypal narrative structures have been reflected and reinterpreted in literature throughout history.

"... we show that sketches for Polynomial Kernel from the randomized numerical linear algebra literature can be used to approximate the polynomial attention which leads to a significantly faster attention mechanism without assuming any sparse structure for the attention matrix that has been done in many previous works. ... we propose an efficient block-based algorithm that lets us apply the causal mask to the attention matrix without explicitly realizing the n×n attention matrix and compute the output of the polynomial attention mechanism in time linear in the context length. ... "

"... we propose a Linear Attention Mechanism which is approximate to dot-product attention with much less memory and computational costs. The efficient design makes the incorporation between attention mechanisms and neural networks more flexible and versatile. ..."

"... we introduce sparse factorizations of the attention matrix which reduce this to O(nn‾√). We also introduce a) a variation on architecture and initialization to train deeper networks, b) the recomputation of attention matrices to save memory, and c) fast attention kernels for training. We call networks with these changes Sparse Transformers, and show they can model sequences tens of thousands of timesteps long using hundreds of layers. ..."

"... We propose FlashAttention, an IO-aware exact attention algorithm that uses tiling to reduce the number of memory reads/writes between GPU high bandwidth memory (HBM) and GPU on-chip SRAM. We analyze the IO complexity of FlashAttention, showing that it requires fewer HBM accesses than standard attention, and is optimal for a range of SRAM sizes. We also extend FlashAttention to block-sparse attention, yielding an approximate attention algorithm that is faster than any existing approximate attention method. ..."

Findings (paraphrased):

1 - students are no longer extraordinary but merely average

2 - employers can no longer rely on applicants with university degrees to be more capable or smarter than those without degrees

3 - acceptance into university is no longer an invitation to join an elite group

4 - estimating IQ based on educational attainment is vastly inaccurate

5 - obsolete IQ data/tests ought not to be used to make high-stakes decisions about individuals

There is another article where he references data from each country showing the increase is not just in USA:

https://www.afterbabel.com/p/international-mental-illness-pa...

In Nordic countries too:

https://www.afterbabel.com/p/international-mental-illness-pa...

Related: Dad Joke APIs

https://www.dadjokes.io/documentation/endpoints/random-jokes

https://api-ninjas.com/api/dadjokes

https://icanhazdadjoke.com/api

"Meta-learning models, or models that learn to learn, have been a long-desired target for their ability to quickly solve new tasks. Traditional meta-learning methods can require expensive inner and outer loops, thus there is demand for algorithms that discover strong learners without explicitly searching for them. ... We argue that population-based evolutionary systems with non-static fitness landscapes naturally bias towards high-evolvability genomes, and therefore optimize for populations with strong learning ability. We demonstrate this claim with a simple evolutionary algorithm, Population-Based Meta Learning (PBML), that consistently discovers genomes which display higher rates of improvement over generations, and can rapidly adapt to solve sparse fitness and robotic control tasks."

Your post resonated with me and you have a well-structured description of this sentiment. Are there any books/articles that you can recommend about this issue?

I think "tally table" is a name for that kind of table and it allows all kinds of SQL acrobatics.

I've read this quote recently of GK Chesterton sounding like a revolutionary as opposed to conservative.

From Orthodoxy:

"All conservatism is based upon the idea that if you leave things alone you leave them as they are.

But you do not.

If you leave a thing alone you leave it to a torrent of change.

If you leave a white post alone it will soon be a black post.

If you particularly want it to be white you must be always painting it again; that is, you must be always having a revolution.

Briefly, if you want the old white post you must have a new white post."