I had this impression as well, until I had to troubleshoot some problems I was having with the screen. Did not give it root access, but it run a bunch of analysis on the system and within a few minutes it was spitting out configuration files that I just had to copy in the correct directories.

Doing the same myself would have taken me a day on the arch wiki. I’ve been using Linux for years, when I was on X I was editing the Xorg.conf without looking up the documentation. If you know exactly what the problem is, you’ll fix it faster that way. However, if you don’t troubleshoot many systems often it is unlikely that you have a structured approach to identifying the problem. LLMs can be quite organised in doing that.

Never seen small spherical walnuts.

Tobacco has been used regularly for way longer and in much larger quantities than cannabis. There’s been much more attention on tobacco as it is something that through the past few centuries was widely spread across society.

Widespread usage of cannabis is something moderately recent.

Cannabis is also consumed in smaller quantities: an average user may smoke one or two cigarettes a day, while a tobacco smoker easily smokes 10 to 20. Moreover, tobacco is legal while cannabis has been illegal in most places until very recently when it started being allowed in a few countries. This leads to much more data regarding tobacco available than for cannabis.

Good morning, the list of known carcinogens is quite short. It is quite short because the evidence required to get on that list is quite extensive. This is an article explaining the basics of how that list works: https://my.clevelandclinic.org/health/articles/25081-carcinogens It is not a comprehensive explanation of how it is made, but it explains the basic reasoning behind it.

Knowing a compound is carcinogenic is not enough to link it to something being carcinogenic with high statistical reliability. The reason why that list exists is that statistical evidence is more reliable than scientific evidence. You can build a scientific model to explain why something happens, but that always has some assumptions. A large enough statistical sample has a much higher confidence. We know that pure benzene is carcinogenic, that is why it’s mostly prohibited to use in research and production. Do we know if other products containing benzene are carcinogenic because of benzene? Not really.

We can make an assumption: products who contain benzene and to whom people are regularly exposed are carcinogenic. It does indeed make sense as an hypothesis, however we don’t know if it is true with statistical reliability. It may be true for some mixtures and not for others. This is in my opinion sufficient evidence to link some in general to cancer. Do we have any proof that any kind of smoke other than tobacco causes cancer? No, we do not.

However, the fact that we do not have such evidence does not mean that smoke does not cause cancer. We have strong evidence that smoke causes cancer, however we can not prove it with enough statistical reliability. We can not prove it statistically because this kind of studies is complicated to perform as it requires data that is difficult to obtain. You have to study a large sample of people, tens or hundreds of thousands, throughout their life while they perform one such action. As such you’d need to follow a large group of users who smoke cannabis and are open to receiving regular checkups and interviews. You also need a negative control: a group of people who does not smoke marijuana. However, tobacco and red meat are also known carcinogens; to exclude whether they got cancer from these causes you’d need separate groups to compare to: people who smoke marijuana but not tobacco, people who also smoke tobacco, people who smoke both and eat red meat, people who smoke marijuana, no tobacco but eat meat, people who smoke marijuana not tobacco and no meat. Now add to this that being an hairdresser is a known carcinogens and you’ll start to understand why obtaining a definitive answer is difficult despite the fact that many studies exist.

but a wide swathe of studies has failed to ever conclusively establish a connection between cannabis smoke and cancer.

The list of known carcinogens is quite short. That is mostly because it is difficult to conduct studies with a large enough sample to be sure that something is a carcinogen with high statistical reliability.

Given our current knowledge, it may be argued that eating fast food every day is not bad for you, as there are no conclusive studies linking it to increased death rates.

In the laboratory, most mutagenic compounds are labelled as mutagenic despite the fact that they are not known carcinogens.

There’s plenty studies linking smoke to diseases. Smoke is a known toxic agent.

Cannabis smoke is not a recognised carcinogenic agent. That is different from saying there’s no evidence linking it to cancer.

It’s smoke in the lungs on a regular basis. That’s plenty evidence.

The article discussed in depth how bad technical decisions made Microsoft product bad quality and unreliable. They they propose that this is the reason why Microsoft lost potential contracts with openai.

The Linux foundation, home to some of the best software engineers, and known not to pick up any trend just because it’s new - let’s say they still work with patches sent in a mailing list - reckon that a new tool is being very useful to them to the point that they’re integrating it into their workflow.

People still criticise them because they should know better such tool is useless.

To be fair: take this methodology that was developed years ago but apply these specific cutting edge methodologies to improve it gives you much better code in a day or two than any scientific code I’ve ever seen published by affirmed researchers in the field. You get the code, documentation and tests. Is the code easy to maintain? Most certainly not. Is code published by scientists maintainable? You’re lucky if it even runs. You take that and you have a partially working solution, you spend a week rewriting it and you have a working better methodology that would likely have taken you a year to develop.

The fact that the material is resistant itself does not mean that a layer of atoms of a few nanometers of thickness is scratch resistant itself. I guess you’d have to store it in a protected area and handle it with gloves, which doesn’t feel much more appealing than magnetic tape.

In a drill bit you don’t really care if a few atoms on the external surface fly off, in this case you would.

2 terabytes of data could fit within the area of a single A4 sheet of paper

Unless it can be paper thin this does not look better than magnetic tape. Moreover, would the handling of the material be safe from handling? I reckon you could scratch it pretty easily.

Sure, but I doubt RAM plays a big role in this.

I consider price and technical specifications. I don’t have 200€ to spend on a phone. Most phones I bought were less than 100€ new. What I care about a phone is that it supports two SIM cards.

With such constraints, choice is quite limited unfortunately.

Is it worth having a free device? Indeed. Is it worth spending 4 times the price just for that? Not to me.

As long as your phone model is supported by any custom mod. I have checked compatibility for almost all smartphones I owned, some 7 or 8 through the years.

Not a single one of them was ever supported by a custom mod.

I like to have a separate partition for /home Whatever happens I can wipe root safely and install something else.

Vista is the reason I started using Linux.

Food is very cheap and available everywhere. Clothing is basically free, and clothes are now easily high quality.

his and 90% of people’s material conditions are shit

Other people is richer than you does not mean your conditions are bad.

Thank you for the article, it really was a great read.

I work in the field of drug development, and I am very impressed by what this one guy was able to do. There’s people who worked in the field their whole life and don’t know even half the things he was able to do using chatgpt to plan out a project.

You should indeed read the article, it is quite a nice read and may show how chatgpt used in a proper manner can be a very useful tool.