ahdh8f4hf4h8

Agreed - I think it would be interesting to see it by state/province (or even better some kernel smoothing) instead of country. Country size is kind of arbitrary, as some countries are very large and others are relatively small.

(Note: full study PDF is on researchgate)

I would take this with a giant grain of salt:

1.) N is in the hundreds of students, all with similar backgrounds - likely a highly specific population. It may not generalize well.

2.) The videos were pretty short: 3-15 minutes long.

3.) The overall effect size is pretty miniscule - it would make little practical difference.

4.) The video topics were roman history and real estate appraisals.

It will be interesting to see if this generalizes (to longer durations, more technical topics, more diverse population, etc.), but given the above I wouldn't change any personal habits based on this study.

In the mid 2000s I worked at company that used Rocky Mountain Basic in a ton of their products (electrical test equipment.) Most of their product lines dated back to 1960s and 1970s, and the software portions consisted of assembler, Fortran, C, and Basic. We were in the middle of a project porting some older Basic code to RMB when the division was shutdown due to 2008 financial crisis. (It effectively killed some parts of the industry for a few years.)

Sadly, a lot of the really old code was much easier to maintain than most of the code I work on today. The company had an excellent engineering culture, and thorough design and documentation was the norm rather than the exception.

I think most of these techniques come down to the relationship between the management and employees. If everyone is honestly trying to solve the problem and earnestly listening to feedback from people above and below, than almost every troubleshooting process works. If the goal is to point blame to another team, then every process will fail.

> Compare that to the Googlers who constantly complain about their employer

The current set of Googlers that are complaining aren't the ones that built the company though. Google's success was built in the late 90's and early 2000s when they really revolutionized internet search and then email. I would argue that since 2010 they have lost this edge - even Bing and DuckDuckGo give better results now for many searches. Many of their interesting projects today are acquisitions.

I'm curious if SV's culture has changed in the last 15 years, and what future impact this may have - at least from the outside it seems like some of the brutal meritocracy attitude that made SV great has died off.

In HS, I would have had more extracurricular activities at least joined some less competitive sports like cross country - there is literally no other time in your life where you will have that kind of free time. I was basically a classic book nerd in HS - very high GPA and class rank but not much else.

In college, I graduated with an Electrical Engineering and Applied Physics degree in 2000. My passion was always software, but at that time almost everyone (in NE Ohio) was giving the advice that electrical engineering would be more lucrative in the long run. I literally read the phrase "the bud is off the blossom" when researching computer science in 1995.

I wound up never doing EE professionally, and have working in software since graduation.

In retrospect, I would have followed own instincts and studied philosophy and math with a minor in Comp Sci. I think the Com Sci minor would be sufficient from a technical side, math would have been a better prep for data science/ML, and philosophy is just good for building mental models and examining your beliefs.

and electrons are just quarks, and quarks are just a state in a quantum field, and a quantum field is just...

> ...endless annotations, dependency injection magic, all sorts of pseudo-types

This sounds like what happens when a bunch of Java/C# developers jump over to python without learning the "python way" - this is more related to the developers than the project

> But nowadays, I don't think modern Python fits any use case real well

Python has effectively taken over the data science / machine learning space. For most use cases, the algorithms are massively more important than the language.

> poorly managed ecosystem and hodge-podge of tools like virtualenv, pyenv, poetry, etc. that never quite become standardized and

This is true, but Java and C# also have many issues in this respect. The move from Java 8->11 is particularly painful - many fundamental libraries related to security or connection handling were not backwards compatible. Many libraries now require multiple branches for different JDK levels. Maven and Nuget are about as good as pip - they all have weird edge cases.

I use both Java and Python on a daily basis - each has their strengths and weaknesses. Java is great if need long running processes with weeks/months of uptime, Python is great for backend data manipulation and analysis.

IBM was a much bigger and more domineering monopoly at it's peak - there was a time where you effectively could not have computers or a datacenter without paying IBM money, directly or indirectly. I wouldn't qualify most of the FAANGs as monopolies (at least not yet) - reasonable substitutes exists for all of them.

The modern market does favor a handful of large players for each market - this is true in almost all sectors though. This not a monopoly though; as long as the top players change over time it is still a working market.

I don't think traditional companies will subsume technology as the article claims - big companies tend to innovate by acquisition, and VERY FEW companies are actually good at producing software at scale. I don't think we've reached phase 3 in the article yet - very few companies have figured out the "magic" of software, hardware, large scale data mining, hosting, etc.

I also don't think tech will have long lasting monopolies, unless government regulation makes it so. Who's worried about IBM or Intel nowadays? Anyone using myspace or vine? All of these were dominant at one point.

I do expect a shift in investor mindset - right now tech companies are not held to the same profitability standards, so the tech stocks are very expensive for even wildly optimistic projections of future growth. It's as if the market is already pricing in a future monopoly for these companies. Given how fragile tech monopolies tend to be, I expect a correction at some point (not a good short though - no way to predict the timing)

Meritocracy exists on a small scale, but always remember that someone wrote the rules, and they generally have an opinion on the outcome. The best moment of the show was when they changed the rules of the glass jumping game after someone figured out a trick to win. Basically the whole RobinHood/GMC debacle. Only approved parties get to manipulate the market.

Most of us are on the winning side of the game right now - central banks have dumped a ton of money into investments and banking, software has very low marginal cost at scale, and companies can be unprofitable for decades if they have a good story. Big Tech has also been immune to regulation and taxes. If the regulatory/legal space or the market want ROI now, then many of us may be the losers in the next round of our economic Squid Game.

Theoretically masks would be much more effective than this - I think the 30-50% number came from a study where they looked at actual non-professional usage of masks. Most people have bad mask hygiene, so there is substantial air leaking around the nose or chin. Many of the makeshift masks people used are insufficient to capture viruses, but they do greatly reduce the "blast radius" of your breathing and coughs.

Right now we don't know the severity or impact of this variant, and several signs point in a bad direction. Maybe in 2 weeks we won't be worried about this variant, but right now being being proactive may slow down or even stop the spread of this variant. If China had instituted a travel ban in the early weeks instead of covering it up we may have never had a global pandemic - contact tracing and quarantines may have been enough to contain it in the early days.

Yeah, I think the article missed the point on this one - travel bans aren't a panacea, but they definitely slow the spread and can buy valuable time at the start of an outbreak.

There's been too much binary thinking by the media and politicians - for example masks aren't 100% effective, but they do reduce the viral load by 30-50% which could be the difference between getting mildly sick and getting a fatal exposure. We shouldn't let perfect be the enemy of better.

This hasn't been possible since the silk road and was created. No one is going back to the pre-modern world.

The trouble is the authorities won't learn anything - many could care less about science or history, they just care about power and influence.

I do agree that more dissent and conflicting expert opinions were needed in this crisis. Especially dangerous is the idea that "if only everyone did X", then all will be well. This is rarely true.

I think the government's role should be to transmit the best possible information available at the moment, give some strong recommendations and incentives (and change them when new information comes to light), but never use force and coercion. It simply doesn't work in the long run.

Are there any experts here that can explain why rates are dropping in Japan, India, and a few other countries? Is it different flu seasons, different treatments, a new strain that is causing delta to self destruct, other? I have read theories but am legit curious.

I have always done "just in time" refactoring - when a big change comes to a module and you're going to have to re-test most of it anyways, then perform the refactor, test the code/module, and then implement the actual change. It helps if you have some long term vision for the product, so you have some idea of future changes as well.

You want to avoid refactoring if no one else is going to test the change, or if your automated test coverage on the module isn't too good. It's too easy to introduce changes that won't be noticed before deployment.

Chances are most code already works well enough even if the design or code is suboptimal - often the system only changes in certain parts and changes there often. If you can identify the parts of your system that are likely to change due to changes in scale or business requirement, than that is a natural area to focus on in future redesigns or refactoring.

Personally, I no longer use the phrase "technical debt" when speaking to management. It's just too vague of a term. Instead I just explain the actual problem and the consequences - too much near duplicate code leading to slow maintenance, non-scalable design choices leading to slow processes, rigid design choices which hamper the team from implementing the upcoming stories, excessive complexity leading making the code hard to change.

Once you get our of analogies and into actual problems, then you can have a reasonable discussion about future tradeoffs. I find even with non-technical managers, debt/wealth is just not a good analogy for code.

Also, often poor code is just part of a larger process problem. You really need to optimize for the whole project, not just code-level to improve the situation.

The most important part of Bayes rule is the concept of Prior knowledge and Bayesian updating - new data should always be combined with your existing knowledge (starting with the "base rate"), based on levels of uncertainty.

Closely related is the concept of conditional probability - how does the probability of something change as you include other information.

It takes some studying to understand probability theory well, but it is very powerful once you start to think that way. Most of the garbage science reporting in the media would be fixed if reporters actually thought this way and honestly applied it.

Toyota is probably on the wrong side of this tech, but it's still a bad short. They can survive a decade or two off of install base and reputation alone, and hybrid tech is still a better proposition for many people until batteries improve. (EV's still have relatively short range per full charge, long charging cycles, and battery range plummets in cold climates.)

With the "Prime" models, Toyota should be able to pivot to an all EV car pretty easily if they needed to.

I think it can work when thinking about accountability and checks-and-balances in processes and roles; for example how would I detect if my CFO is stealing money from a startup, or what if a narcissist became a manager at my company? But I would apply in on a role/position level, not as a starting assumption in new relationships.

I don't think this attitude is helpful - it's going to leak out in your interactions and poison relationships.

You do need to consider the context and background of a person before you can properly understand or trust their opinions and advice. For example, a physicist may be ignorant of how car engines actually work; a mechanic may be ignorant of some of the underlying physics. You really need to get to know people and build a long term relationship to understand if their opinions on a topic are useful and informed.

I don't think there is a simple rule to working with other people - you just need to build a wide body of experiences yourself, and beware of relationships that grow too quickly or where the other person appears to have a hidden agenda.

Honestly, object oriented design is the anti-pattern here. Almost none of the benefits promised in the 1990's and 2000's have come to pass, and for most applications type hierarchies are of little help.

For most applications, having a really good relational data model, and controlling data updates through a service/microservice will give most of the benefits of "object oriented" design without any of the drawbacks. The service will act as a "live object", data integrity will be preserved, data science modeling will be easy, and the application can be designed using whatever tomorrow's preferred platform of the day is, or even some low code stuff for simple applications.

Not only that, the hacker ethos has completely changed since the 2000s. Nowadays most people would defend the holders of intellectual property, where as before most hackers believed "information should be free" and "copying bytes is not stealing".