cjfd

Well, I think people should go with their own judgment. Was a bad parent bad as a person or unable? Children who endure such an ordeal, as I once was, will know the difference. One of the aspects of enduring bad parenting is that one starts to doubt ones own judgment. Once all aspects have been weighed ones own judgment should be taken as accurate. Spouses of bad parents are very often quite blind to what is happening in their own house.

This is a rather stupid caricature.

It is not so black white. It is a dilemma. There is great value in freedom of expression. But fake expression with the explicit intention to mislead is not so good. It would be best if the population would be better at distinguishing truth from falsehood. Unfortunately much of the population seems to be at a level close to cattle.

Sure, we are likely to see soundness bugs in the current lean kernel again. And in other proof assistants too. Regarding the correctness of proofs this is not the biggest worry, in my opinion, though. Both sides of attempting to get faulty proofs in and of strengthening the kernel to keep them out have AI on their side. If AI advances more and becomes smarter it can be used on both sides. And the defending side has the easier task here. I am not sure about the Navier-Stokes proof but in general it sounds highly unlikely that many well-known theorems turn out false.

The bigger worry are question that Terrence Tao is also talking about. I.e., is the proven theorem actually the theorem we are interested in? Are basic definitions of the field stated correctly? That kind of question. There is also still parts of proof assistants that are not the kernel and that can be abused. E.g., abuse the pretty printer/parser to make something look different from what it actually is. Introducing an axiom while typographically hiding that one was added. If I remember correctly I saw an example of the latter thing in the Coq (now Rocq) theorem prover many years ago. I forgot how it precisely went or where I read that.

Perhaps. But are humans going to get what they deserve?

If we accept that AI is going to do all of these things humans will be superfluous and will just be optimized away. It looks a lot like we see the birth of silicon based life by the efforts of carbon based life. Carbon based life will die and it will not even be because AI decided to kill it, it will be because feeding and watering it was less important than other concerns. Bacteria may continue to exist, though.

The question in the title is basically a non-question. AI is quite good at achieving some sorts of stated goals. The easiest way is by exerting the least amount of effort. The least amount of effort ignores ethical concerns. The training around ethical concerns was most likely rather light to start with. If we accept the hypothesis that at some point the AIs are going to be more intelligent than humans it follows that human survival is not a concern and will be ignored as a superfluous concern.

Well, who knows? The trajectory has been set in and it is downwards. After Nero and the year of the four emperors the Roman Empire still had ways to go. So, I would say also the US can potentially recover. On the other hand the saying 'it cannot possibly get worse than this' has also been disproven may times in history. In any case, I like to sum it all up by saying "The US is a superpower attempting to become a third world country".

As a patrician with only a desktop I was able to use this website.

Sure, it is quite easy to install python from source. Usually I am doing it because I need older version for work projects. Of course, the other thing that one needs in that kind of situations is virtual environments. This makes it also quite easy to ensure that you always run the correct python. Create a venv with a specific version of python and once you are in that venv, the 'python' command will automatically run that python version. That way it is easy to use an unlimited number of python version next to each other.

Yes, this sound like it is true. But is it good? I mean, if I accept this idea, I start feeling more like a pet living in the house that is my platform then as a human who creates things. Also, it can be inefficient in that the same program or algorithm ends up being implemented in every language. Also, the article does mention that there is some pain in between programming language, e.g., when writing language bindings. I would push this point a bit further and note that we have too many tools and it is extremely painful to have to know all of them. Nowadays with AI it gets a bit more easy because it can help a lot with unfamiliar tools, but still that sounds like a problem that would not need to be solved if it was not created in the first place. Multiple programming languages tend to multiply the number of tools needed. I have been feeling for decades that C++ is a good language because it can be both low level and high level and can therefore take away the need for multiple programming languages. The world seems to be taking a different course, obviously, and in my opinion it makes people less free and less powerful than they otherwise could be.

Right. Because If I want feature A and it is available in configuration among features A through Z I will always perfectly pick the right letter of the alphabet and I will not at all be irritated by having 26 options that I have no idea about what is the difference between them.

I kept my old laptop (bought in 2007) alive for quite a while too. But last year I finally retired it. I also used archlinux 32 which worked fine for a while. But at some point the breakages really got too bad. I was using xfce4. For a while the xfce4-terminal was broken and would not start. That has the easy workaround of using xterm instead. But there were more breakages and it just started taking too much time. Quite a bit of software is ditching, or has ditched, 32-bit support.

Well, Joel is not saying not to refactor anything. In fact he is advocating improving the code quality in small steps. What is missing there is that in practice improving code quality in small steps requires automated tests . But I guess he can be forgiven given when he wrote the article.