Recently, something left a deep impression on me. It started when someone asked in a group chat if there was a good way to run multiple Python programs in parallel. However, most group members suggested using tmux, nohup, or simply appending & to run multiple processes directly. Listening to these suggestions, I felt none were as good as simply installing MPI, where a single line of code mpirun -np [num] python main.py could solve the problem directly. Professional parallel programs clearly offer better efficiency improvements and resource contention resolution compared to manually spawning multiple processes, and the number of processes can be controlled much better. Unfortunately, at least he himself likely did not adopt this solution. Even when I emphasized convenience, he seemed to have never considered it (of course, this is just my perspective; I can’t know if he searched for it himself, but he did not reply or ask further questions).

Thinking about it, this is quite normal. After all, for a completely unfamiliar program, when existing methods should suffice, how many people are willing to try something new? Including myself. I feel that as time passes and age grows, accepting a new concept might be easy, but the motivation to try a new product or software has declined significantly. Within my own knowledge domain, I am certain that the tool stack I already possess is sufficient to solve current problems. Unless someone specifically points it out or requires me to use it, the motivation to actually use it is not strong enough.

Recently, while reading Sapiens, I had the same feeling. Perhaps future generations will look at us with the same sentiment: that they and ‘we’ are all trapped within our own cognition. This viewpoint seems to be frequently mentioned. To break out of it, one must spend time trying and making mistakes, while maintaining a skeptical attitude towards problems (Are there better methods? Optimal results?).

But in today’s fast-paced era, can we really achieve this? Humans have constantly been ‘forcing’ themselves and the world to speed up. Humans evolved from crawling to walking upright; the narrowing of female pelvises brought great pressure to childbirth, shortening gestation periods and simplifying fetal development in the womb. During the migration of Homo sapiens, whenever they arrived at a location, it inevitably triggered a wave of species extinction, especially for species with long reproductive cycles, as their breeding speed could not keep up with human hunting speeds. The agricultural era changed humans from infrequent hunting to daily farming labor, actually increasing the frequency of labor. The industrial era was no different. The arrival of the tech era is even more obvious: letters that once required careful word choice became instant messages available anytime. While work efficiency improved, it also meant an acceleration of work frequency. Do we really have enough time to experiment and make mistakes? Recently, I read several Q&A posts on Zhihu with similar sentiments. Those who truly settle down to do research from the basics often achieve little, whereas those who chase trends and do one thing at a time see results everywhere. I’m not saying either side is wrong or anything like that, but it feels that while both have their unique merits, the former often doesn’t get the results they deserve. As for myself, I always want to supplement courses and knowledge from the basics, but this is actually impossible. After all, human memory fades; it’s not like a course or something that requires frequent review. It’s easy to read something, get busy for a few days, and when you want to continue, you’ve forgotten almost everything. Now, I just use what I learn immediately. Furthermore, the cost of learning and accepting something new is clearly higher than just cobbling together a solution with existing viable methods. The deadline is there, the unknown is there; why bother? Additionally, there are simply too many new things. Take current AIGC products, for example; the explosion and surge have already exceeded my threshold for trying them out. Unless a product is particularly distinctive, at best, it just gathers dust in my bookmarks folder.

Actually, I have always questioned people of faith while also envying them; it’s quite contradictory. I believe that trusting a virtual, constructed deity, especially praying when facing difficulties, is something I view as a form of mental control, a means to reduce contradictions and suppress resistance. But conversely, this belief brings a delusion of escaping suffering and an unwillingness to give up. A person without faith might have a higher probability of giving up on their own life, after all, without the constraints of religious rules and ethics, a person as an independent individual might just ‘restart’ if they can’t get over something. Yet, this very faith gave birth to science. This sounds absurd, as religion often played a role in the process of discovering science and correcting theories. The process of discovering science often requires skepticism first. Nowadays, this might be called being a ‘contrarian’ or ’troll’. Why does sometimes asking a rhetorical question bring such feedback? (Of course, here I mean genuine questioning, not sarcasm or passive-aggressive remarks.) To a certain extent, it’s because the other party cannot accept, is unwilling to accept, or simply doesn’t want to hear viewpoints outside their own cognition.

It seems that, besides our subjective unwillingness to accept, objective factors also constrain us, trapping us within our knowledge domains. In my view, there are two ways to break free: first, driven by incentives, such as the benefits from game trials or product trials, or if I try something out and can write a Zhihu article, WeChat official account post, or video tutorial to generate traffic and followers; or for research, if I read A and combine it with my field to publish an A+B paper. Second, driven by curiosity—trying everything I don’t understand. Personally, I feel this curiosity weakens with age, which is why I have a quote on my GitHub profile: ‘Stay curious :)’.

However, being trapped in one’s own knowledge domain is not entirely a bad thing. If I use existing tools to create a new, more convenient tool, isn’t that also a transcendent way of thinking? Although the time cost might be higher, all roads lead to Rome; creating one’s own toolchain is not a bad thing.