a device that is generating concepts

Leave a comment

Scalable Generation naming

There is no standard agreed naming for generations and in the past the naming was a major event driven (e.g. post/WWII) nor the range and we are at GenZ which left not too much to go, it might be a good strategy to come up with a scalable method.

First let’s start with the range as it can give the sense of granularity. As of X/Y/Z the range is 15 years, which might make sense as it is the age of getting to a maturity, thus changed behavioral pattern. However taking into account the exponential technological growth impact on accelerated maturity and rate of significant changes, 15 years sampling might be too gross.

Proposal is to lower it down to 10 years due to the reasons above and since it is easier to multiply.

Now, as of the reference point, we can take 2020, or better year 2000 as the easy baseline to add and remember the multiplications of 10.

The naming convention then would be a scalable order of decade counting from 2000. e.g. children born between 2000 and 2009 would be Gen0, born at 2020-2029 would be Gen2 and so on.


1 Comment

COVID-19 (Coronavirus) post-incubation

[October’20 update at the bottom. Hint – really good prediction so far: Israel Covid Dead[+14days] = 0.82% of Confirmed (rsq = 0.9236)].

Those days, when you have to stay (and work from) home, you realize how interconnected our world is – because someone in China distant province eaten something bad, you and all around you (regardless where you are from) stay at home with all kids for weeks or months, gather food, cancel flights, loose savings… in best case, or life in worst. How can we say after all, that this is not a unified world we are leaving in. Enough philosophy – let’s get to the data.

There is an ambiguity about incubation period of COVID-19 and it is defined between 2 and 24 days. Definition of incubation period is “the time elapsed between exposure… and when symptoms and signs are first apparent“. There is however another important index – the duration of disease. Based on the data we can get the duration after exposure till critical period. Assuming “Total Cases” index is captured (on daily basis) after incubation period in most cases and after latent period (the time from infection to infectiousness) in some, we can run correlation between the “Total Cases” [from here: TC] signal and to “Total Dead” [from here: TD] index signal, capturing Rsq while changing time shift between them. The peak of Rsq would expose the Time-to-critical-point of the disease.

Dataset source: Johns Hopkins University Center for Systems Science and Engineering (GitHub; HDX) captured at 16/3/20

Since China has changed on the way the metrics, therefore the dependency between the TC and DC got duality in behavior as we will see, so I have split the data for analysis.

Here is correlation coefficient between TC[day] and DC[day-i] and how correlation looks at peak for all the world except China:

Total Death [after 12 days] ~ 0.248*Total Cases
Total Death [after 6 days] ~ 0.09*Total cases
Total Death [after 0 days] ~ 0.03*Total cases

You can see that as we go towards peak of death index at 12 days after incubation among captured cases, the multiplier becomes scary-high – almost quarter. Really hope that is not the case and just a transitional distortion…

Continue reading

Leave a comment

Language-driven opportunities – Internet

Merging W3Techs and InternetWorldStats numbers for 2019 Language content dominance and users statistics together, easy to see that there is a lot of potential for Arabic, Spanish and especially Chinese languages content development.

Internet content vs users by panguageOther languages are a long tail of sub-2% users and content availability. Very stable Lingual users breakdown with strong dominance of English (changing from 58% to 55% over almost the last decade).

Salable and convenient language transformation tools and methods could close the shown gaps and leverage potential for the barely English-speaking countries.

Even though Wikipedia is also dominant by English articles, the distribution of languages there is slightly less polarized.

Internet content vs Wiki by languageHere you can see that Swedish articles got outstanding 7.3% [2019] share  for very small amount of users, generated by automatic translation bot, developed by Sverker Johansson, that received a lot of criticism, thanks to poor translation abilities of Google Translate. Something to think about…

Leave a comment

Skills network

Anything related can be a network. Any network got a hidden magic inside. Building a network is unleashing another part of the beauty of this world. And when you do this, when you find the secret clusters, when you connect the lonely entities of existence – then this magic is flowing though this network, enlightening and reviving nodes, breaking out over the invisible strings into the life, connecting everything and everyone.

Our skills are shaped by the templates of education and jobs, by human behaviors and to get a glimpse of this bias we can do a simple experiment.

Let’s build s crawler that is getting LinkedIn skills from many people, connect those skills based on cross-correlation and build a network that is based on highly correlated skills.

Even small amount of people I took (about 150), is giving a network of over 400 correlated skills, connected by 3500 links and lots of interesting insights.


So… Mainly pics and less words.

Parts of the network, related to declared programming language-based networks (per language). Pay attention on differences:languages

Design (Since this is a small network, it is biased by my occupation, so the design is mainly represented by word of semiconductors – doing it widely would expose other meanings of design):

Continue reading

Leave a comment

Annual Feedback quantitative analysis

Yes, I am more than 10 years in Intel Electronics and every year I’ve got the annual feedback (called “Focal”). I thought to look at the nature of the work that I am doing by analyzing the semantic data of those documents. From confidentiality reasons I cannot bring parts of it here, but I can show you some quantitative analysis that I ran on the data.

Among various perspectives, I wanted to look at the verbs that are used there to describe my accomplishments (one out of three components together with “strength” and “areas for improvement”). Out of 30 paragraphs (3 each year) and 4.1k words in total, here is the distribution of all top verbs (overall about 200):


The thought was that verbs of accomplishments are the nature of the work. Now, when I look at it, the direction is amazingly correct – this is what I actually did.

Taking “worked” as a baseline (100%), I have calculated the rate for the rest of verbs relatively.

In addition, grouping verbs by their nature (excluding neutral “worked”), I can tell now, precisely how my work looks like:

nature of work

Know yourself. 🙂


Cellular Automata in excel

Had some time recently while flying from one place to another and played with simple Cellular Automata rules in excel. It is very simple to do – you create a rule within the cell as function of other cells and extend it (you can simply copy/paste) to other cells.


Here, if A2 is equal to B1, we define a value within the cell “Set”, otherwise it remains empty. That’s all.

Now, if we will keep the first row and column empty (let’s call them “boundaries”), while extending the rule to bigger area, we are going to get a beautiful pattern:


Here I took the area of 200×200

The size of the “triangles” are recursive 2N+1 (1,3,7,15,31,63…)

So by very simple rule we have created pretty complex pattern

Now what is interesting that by simple change of the reference cell the pattern complexity can dramatically increase.

If we change the “offset” of the reference cell by one i.e. apply this kind of rule:


We will get the next pattern:


Continue reading

Leave a comment

Network of physical quantities

Had couple of evenings of fever 😦

Took International System of Units (SI units) as a baseline and created a network:



Nodes – Physical Quantities. Size of Node is based on OutDegree Ranking (kind of “Importance”). Colors are based on Subject:


Edges – relation based on Units, expressed in terms of other SI units. Color is based on relation (Blue – multiplier, Red – divider). Width/weight is based on power (log).

Gephi layout – “Force Atlas 2”

You can see that most “important” Physical quantities are basic Length, Mass, Time, but also derived Force and Energy.

Taking a look at the Betweenness Centrality, we can see that Force and Energy are the biggest and then Voltage, Resistance and Charge. Maybe this is why it is so native to learn them first right after the basic physical quantities.

betweenness centrality

Enjoy and feel free to download, modify and correct (surely got some mistakes). If you do, please mention the source 😉

Files: Excel/Gephi

Leave a comment

Annual pattern of Demand for Laptops

As I am working on analysis of Lenovo company (hobby/investment), here is an interesting chart out of the work.

Beautiful pattern of demand for Laptops (relative demand per week, comparing to an annual average of demand) during the last decade:

Demand for Laptops - annual pattern

The pattern is very precise and stable since 2004 during most of the year except maybe some end of the year behavior inconsistency. Though there are expected two peaks of demand at winter holidays, year-to-year  variance during ww48 to ww52 is still high. Extremely high peaks at 2007 holidays, driven by Intel Core 2 Duo products wave.

You can see a clear back-to-school growth towards August from the (-10%) dead-season of Laptops marketing at Spring.

Leave a comment

Annual pattern of jobs searching

Are you looking for a job? Most probably this is not a coincidence. Eventually there is a clear annual pattern in amount of people searching for new opportunities.

Here is a normalized relative pattern of amount of people searching for a jobs for a last decade.

Data is based on worldwide statistics based on millions of people and can be used as a model for Human Resource organizations, headhunters or just people that are part of the trend.

X-axis is weeks (1-52), Y-axis is growth in jobs hunting comparing to the average of the year.

X-axis is weeks (1-52), Y-axis is growth in jobs hunting comparing to the average of the year.

We see a nice and clear behavioral pattern that shows a gradual decline throughout the autumn to the annual minimum of job searches by the end of the year and the slight spike before holidays (people that remind that they need money for presents?). Then enormous spike to the annual maximum at the beginning of the year (motivation to change things in the new year? worldwide layoffs pattern?) and then spring decline followed by growth to a summer “hill”.

Same, but with average profile (Rsq > 0.8):

WW Job Search Profile

What can I say? In my company this profile would be one of base components of HR organization budget. I find it cool.

Leave a comment

Ropes and the Perception of Risk

Couple days ago I had an amazing experience at Kos Island (Greece) and tried a ropes course, where I had to pass through all the stages of the cool facility, going from one column to another over the ropes and bricks with increasing complexity. It looked like this:


Six phases x three wings x two levels i.e. 36 stages of fun and challenge for my equilibrium. Beyond the unique experience, I learned something very interesting from the instructor (on the pic above, he is at the middle stage, waiting for me).

After I have passed several stages, he said to me: “You know that there are three points of support, required for equilibrium here. One is the left leg, another is right leg, while the third is usually the hand.” Indeed, I have mentioned that to feel stable, I had to hold the rope (or anything else available) with (at least 🙂 one hand. “BUT the interesting part is that you DO NOT HAVE TO HOLD the rope. It is enough just TO TOUCH it. Even slightly with the side of the finger”. Obviously I have tried it and it was amazing! I could relatively freely walk while tip of the finger was touching the rope, and at the same time I could not walk from the moment after it was detached. So for me it was no more the holding rope, but just the reference point.

Immediately as an engineer I thought about the negative feedback that my brain required to do this task, I thought about sensing lines from power electronics, I thought about feedback in control system of my brain, I thought about continuity of nerve that is running from my finger to some place in the brain and realized that this third point, required for the equilibrium was not actually the finger itself, but that place in the brain that required this sensing for better control of my body.

Now I need to distinguish my experience from a single line “Tightrope walking” (funambulism) or slacklining where there are a totally different biomechanics and brain control of the balance.


I am talking about purely psychological effect of the fear of the failure where you can “easily” go at the height of half a meter and make not a single step at fifty meters. I am talking about effect that you slowly go two thirds of the path and then run freely when you see the end.

This was very much related to the work I am doing these days on Risk Modeling in Complex Systems generally and on Perception of Risk particularly. I shall release some interesting parts of the work here… By now, to touch the Perception of Risk model, I would like to mention two major components of the risk – Perception of the probability of failure and Perception of the impact of failure. Each of those components got multiple assessment biases and this is a whole separate talk by itself. In this case of equilibrium reference point, the signal in the brain was impacting the part, responsible for mitigation of the failure probability component of risk perception, since the impact of the failure was same. The mitigation was possible because of the constant signal to my brain that in case of loss of balance, would ensure that the hand is very close to the rope to catch it. If this is true, then the brain mapping (if it is going one day possible not in static lab conditions) of walking person at the same height with and without the finger touched, would show the area of brain that is responsible for the Perception of Probability component of risk, while the different heights would expose the area responsible for the Perception of the Failure Impact.

Interesting that once I passed some stage the second time it was much simpler to detach the finger and in some cases even not use the hand at all and stretch out my arms to increase the moment of inertia for better balance. It was nearly impossible for me to do it for the first time per stage due to the lack of risk assessment and perceptional bias (a.k.a. fear)))).