At first I thought exactly that - just why.

When days started getting warm in the beginning of the summer I started investigating and thought that it might actually be a good idea - you could turn AC an hour before you come home to find cool room.

When summer hit I actually found out that you just leave AC on all the time and it just keeps temperature at 22C or so and there is really no need for it to be connected. You don't want to turn that off as it is more economical to keep the room cool than letting it to get hot and then trying to cool it fast. It barely even showed up on my electricity bill at the end.

So all in all - really not needed.

Then I remembered that both my AC and floor heating are connected.

A hacker could easily turn on both AC and floor heating and cause me a few euros of damages in my electricity bill.

Bytecode is obviously not human readable - lets make Assembly. Assembly requires very specific skills - lets make a language that is more human readable. This new language is way more readable but requires very specific skills like understanding computer memory. Lets make a language that takes care of it. Lets make a language that is not compiled but interpreted. Lets make a language where you can just simply write 1 + 2 and computer would just give you an answer and you don't need to define anything.

Lets make the language that is simply a natural language and computer interprets it an writes the program you need itself.

We are comparing because any programming language is just a try to make programming easier.

Have you read the article? You have a science like "human behavior" where your experiments can't be reproduced. Simply because there will be so many variables or variation that your science is as good as a guess from somebody from the street.

Or again sciences where you just make models of some imaginary things like hidden meaning 10th century in literature. Your only test is another 10the century literature meaning searcher. You just clap each other on the shoulder and give each other phds.

In physics/engineering and similar real life instantly shows how wrong you are. You can't really go far in your career believing something that is not true and tested.

In something like "relationship studies" you can live in your ivory tower for decades. Your only feedback loop is your own students and disciples who just want to live in the same ivory tower.

You can have some crazy theory like "human relationship output is dependent on socioeconomic outputs". And discuss that theory for decades, get your phd and stuff like that and your only judges are the same people in the same field.

You will not edit Assembly in this.

Claude will have to edit it and Claude will not whine.

It is the same as autorun.exe on some CD - I put some CD in my CD-ROM and my Windows are compromised.

You can argue that I should not clone unknown repos, I should inspect each and any file in the repo, look for git.exe or any other suspicious binary and etc. I should not put untrusted CDs or USB sticks into my computer.

However the reality of it is that it will not be the user who inspects USB stick in some contained environment, it will be security policy which disables autorun. Same as Cursor should disable this.

We know that smoke itself is really really bad for you. Open fire is bad for you.

Sitting around campfire with a glass in your hand and a guitar might be really really bad for you.

One interesting point the article touches - there was a study concluding that if you quit smoking at 40 your life expectancy basically equalizes with people who did not smoke in their life. It is an encouraging message that it is never too late to quit. Then again it also sends a different message - you can smoke as you wish in your 20s.

Just yesterday I saw an article on Instagram that they are putting smoked meats and sausages and similar products in the came cancerogenic category as smoking. Which again one one hand states that meats are very bad for you. On the other hand it makes smoking not so bad as you would think? Because people are eating sausages and meats cooked on open fire for thousands of years?

Some guy on Diary of CEO states that rice are basically a poison because they are pure sugar. If you want to live healthy you should definitely drop the rice out of your die. Then again a billion of people eat rice every day. What gives?

We want to have all this information because we want informed decisions in our lives. If we are analytical we even want formulas and graphs just like in this article. What we don't want is to give things to chance and genetics.

Different breeds of dogs have life expectancy difference of ~50%. People of course are not so different in their size but we need to always keep this in mind. You can live as healthy as you like but you always have a ticking bomb in your DNA. What I still am not getting is how much living healthy impacts the outcome versus the genetics. It might be a hopeless fight to e.g. stop eating the brisket if your lungs have an expiry date of 55 years.

What is definitely not solved is knowing what you want and what user wants and what the end result is supposed to be. To write the code you need the specs and to write the specs you need to know what you want. And that can only be answered by years of therapy.

If you gave junior dev exact tasks what to do where you will get better results.

Just like with LLM.

I mean it is a tool and you need to understand how the tool works. When there is too little context, where there is so much context so that you are poisoning it, when you are allowing the tool to do patch-on-patch and etc.