September 17 2020, 15:42

In Istanbul, an interesting trick is widely used: “sorry, your card didn’t go through. Maybe cash?”

Checking into the hotel. Will you pay right away? Okay, let’s. Oh, the card didn’t go through. Alright, I say, then I’ll pay later. I go into the city. Starbucks Reserve. It’s like the coolest Starbucks “in the neighborhood”. Oh, sorry, it didn’t go through. I give another one. “That one didn’t go through either”. I say, I’m patient, I have five more, and now I’ll call my wife, she has.. Hooray, finally, it didn’t go through, they say. We enter a restaurant to eat. We ate. They bring the bill. I hand over the card. It didn’t go through! I give another one. It didn’t go through! I say, no problem, I have five.. A girl comes by: “we have another terminal, maybe this one will work?”. It worked. I return to the hotel. I want, I say, to pay today so I don’t have to worry tomorrow. We tried the same cards, six declines in a row. I say, no problem, now we’ll contact booking, I’ll argue with them, and they will allow me to pay through the personal account. I say, theoretically it might affect the hotel rating, so maybe you can find some terminal that, by chance, will work? They looked – one worked.

The hotel receipt said “Error 14. The number does not belong to a credit card / The card number is incorrect.” (translated from Turkish).

In the cafe — “Error 12. Rejection. Invalid Transaction. Undefined transaction authorization”.

Obviously why. In this country, with the pandemic, small businesses survive however they can, and cash is a classic way to avoid taxes and bank commissions. It has always been like this, I just didn’t notice this particular tactic in my previous three visits.

P.S. By the way, in Istanbul nearly 100% of people on the street — in masks. Here the police fine you for not wearing a mask, the fine — about 9,000 rubles (900 lira).

September 14 2020, 00:45

Check out this fantastic lecture on viruses by Evgeniya Lazareva @[1040949013:2048:Lazareva Evgenya]

If you don’t feel like watching the whole thing and just want to see the part related to what I write below, then check out

– about how the virus manages to encode 4 genes on a super short circular DNA. Programmers will find this especially interesting.

For this Hepatitis B virus, one of the strands (in the ring) is even shorter than the other (1700-2800 nucleotides) – just 3200 nucleotides long. And this “program” encodes four genes (which encode 4 proteins: HBsAg, HBcAg, polymerase, and the gene expression regulatory protein). So, it seemingly does not fit the information for 4 genes at first glance. And you know how they solve this? The same sequence is read in six different configurations of a “floating reading window.” Roughly speaking, the same “bit” participates in six “bytes” simultaneously.

Sapolsky stated in his lectures that the interpretation of this program gives different results depending on the cell’s state. I then started actively googling and stumbled upon an interesting thread on Reddit, describing how an organism understands which cells to make for different tissues using the same program:

https://www.reddit.com/r/askscience/comments/irxncu/how_does_a_cell_know_what_to_become_if_they_all/

There are enzymes that suppress or otherwise activate genes by their presence during the polymerase’s operation, i.e., during the “program’s” reading and interpretation. These enzymes can appear from the external environment (e.g., from the mother in an unborn child) or be produced by the body itself, or they might be the product of the same program. But overall, the real mind-blower is that a program controlled by these enzymes can create enzymes that will control the mechanism in the next cycle, which will also create an enzyme that will control the next cycle where a protein is then created. This part of the process is called “epistasis,” and such genes – suppressor genes or inhibitor genes. If you’re interested in examples, read https://en.wikipedia.org/wiki/Epistasis (in English), where there are many examples.

The mechanism that creates different isoforms of a protein from the same “program” is called alternative splicing. For example, the human body synthesizes no less than 100,000 different proteins, while the number of genes encoding them is approximately 20,000. Thus, inhibiting one of the components in the presence of a hundred others already results in a new protein.

Sapolsky also mentioned in his lectures that these inhibitor enzymes need to be present in progeny cells in their previous concentrations when a cell divides, which is a probabilistic process. If the number of enzymes in the progeny cell is insufficient to suppress some gene during its expression, it begins to produce what it shouldn’t, and some serious problem develops. But this usually does not happen because there are error-correcting mechanisms – cells where something has gone wrong are destroyed.

Imagine what a single mistake in an unfortunate place in the genome could lead to? Thus, a system largely based on probabilities and a “suitcase balance on a needle point” turns out to be very reliable if there’s plenty of time for debugging and clear environmental fluctuations. Again, programmers will find this interesting.

Clearly, designing all this on purpose as we currently envisage, by “trial and error”: once created by chance (at its first level of complexity), and then simply survived in some conducive conditions and perished in 99.9999(9)% unfavorable ones, and 99.9999(9)% did not find any favorable conditions and died off as “species” never achieving the “goal”, and we now think that it was a mechanism specifically created by evolution for such conditions. No, it’s just because there are very, very many organisms, even more conditions, the simplest organisms have fast generation turnovers, and there were 3-4 billion years to work with. Basically, trial-and-error programming works excellently when you are not in a hurry and have infinite resources at your disposal.

But it looks like magic, I swear.

P.S. What else to write on FB first thing in the morning, 5 minutes after waking up.

September 13 2020, 13:45

I’ve already lost this twice and had a hard time finding it on Google again, so let it hang in my little Facebook archive for myself and others.

I made an effort and gathered the needed 3D models of simplified human skull heads, which can be rotated in any direction and used as a reference for drawing custom heads.

https://sketchfab.com/Epidural

There are 13 of them here

Plus, there are some good models from https://sketchfab.com/brushrush

Here are also some good models:

https://sketchfab.com/3d-models/drawing-figure-for-artists-base-head-planes-fafb17bb02664aa29fb006263e4ee186

https://sketchfab.com/3d-models/low-poly-reference-male-head-1bd8d40882ba4749ace512d00467f0cd

https://sketchfab.com/3d-models/low-poly-head-bbeae382b4d5442bb88f68b9476161f4

https://sketchfab.com/3d-models/male-head-iii-35819a6090544762ba91b30ae8c6606c

https://sketchfab.com/3d-models/planar-head-by-oleg-toropygin-7d831ee5b5494ab595c443088e549c59

https://sketchfab.com/3d-models/planes-of-the-head-asaro-a8949dea30ad48f5936c6c5848674858

John’s Asaro head:

https://www.artstation.com/artwork/rBnAG

https://sketchfab.com/3d-models/asaro-head-69a6612dce3c42659b504c8e96769b23

https://sketchfab.com/3d-models/asaro-head-complex-side-f313dcb5ccde429cbd7eee303f139768

Loomis head:

https://sketchfab.com/3d-models/andrew-loomis-head-d2020b07408b4f238004f04d082a3193

https://sketchfab.com/3d-models/andrew-loomis-head-v20-b39985b404f8432496aafdd72398575a

Reilly’s head (wireframe model):

https://sketchfab.com/3d-models/reilly-head-dfff220da27743dc8445abad8d5d5f23

https://sketchfab.com/3d-models/reilly-abstraction-version2-ff00933296be49d69deb75fa07c7255c

https://sketchfab.com/3d-models/v3-reilly-201c917375c143309e650fd6562a50e1

Skull:

https://sketchfab.com/3d-models/january17-head-study-skull-a31c2e8569c04bb2923d001556bcd7ad

And just a useful Pinterest:

https://www.pinterest.ru/search/pins/?q=%D0%9E%D0%B1%D1%80%D1%83%D0%B1%D0%BE%D0%B2%D0%BA%D0%B0%20%D0%B3%D0%BE%D0%BB%D0%BE%D0%B2%D1%8B&rs=srs