tkihira I was talking today about how it is now a natural assumption that the lifetime employment system is broken, but it is also becoming dangerous to keep the same job forever. If we seek a way of life like programming as an IC for the rest of one’s life, it is difficult to know how we should seek an advantage over young talented people. tkihira The internet is a great way to keep up to date with the world, and to communicate with the world’s top programmers on github. AI will further improve the learning environment. Our industry will basically become a better environment for young people, and there will inevitably be more young and talented people.

kumagi Senior IC, skilled people are solving problems ahead of time like this, and no matter how powerful the weapons used by newcomers become, software will always be complicated to the limit of our cognitive abilities. The day will never come when it’s too easy because it will keep getting easier. image

dmikurube “Software will continue to get more and more complex until it reaches the limits of our cognitive abilities” This really fits my intuition as well. When the evolution of tools makes it easier to solve difficult problems, then it is the business of mankind to keep creating new and difficult problems until the next point where we reach our cognitive limits, so… dmikurube Well, that story and the question of whether we can therefore secure “an advantage over the young and talented” are orthogonal issues, yes. dmikurube But also, as a different topic, I’ve been wondering recently how to solve the problem of “solving a problem ahead of time” and the possibility of the problem not being recognized (and not being evaluated (so How can we solve the problem of “solving the problem ahead of time” and not recognizing the possibility of the problem (and not being able to evaluate it, so we should just burn the problem and then think about it)?

nishio If “systems will continue to get more complex to the limits of our cognitive abilities,” then True, but as to whether software will be representative of “complex systems” in 5 years, I I strongly doubt it.

dmikurube That’s probably true. I don’t think that software issues will no longer be involved, but many will have to deal with issues that involve a lot of different layers (and that’s already rather true even now, not just 5 years from now). dmikurube Even now, the problem of how to create an ecosystem that includes the people involved and the incentives for those involved is much heavier than the problem of (stand-alone) software (in the sense of code). The problem of how to create incentives for the people involved is already much heavier than the problem of software (in the sense of code). How much easier it would be if all the problems could be solved by just writing a few lines of code. dmikurube Time spent writing code and dealing with problems to be solved is rather just a reward or a healing time… dmikurube, it may be interpreted by business people, for example, as a problem of Engineering Manager as a job class in a company. However, this is not the case, but rather a general awareness of the problem as a human being.

kazuho “Continuously increasing complexity to the limit” is a problem that has always existed regardless of the industry (e.g., automotive). In the case of software, I think it means that the change in the cost of labor is small as long as the programmer can recognize it (so it becomes more complex to the limit). kazuho Especially in the case of software, the more users there are, the more costs will be invested in development because the cost of production (replication) is zero (this is not just a matter of being able to do so, but a necessity from the perspective of competitive advantage). (This is not only possible, but inevitable in terms of competitive advantage) As a result, the program becomes more complex until “the programmer’s cognitive limit” where the rate of cost increase increases nonlinearly - The Limits of Cognition


This page is auto-translated from /nishio/複雑さは増大し続ける using DeepL. If you looks something interesting but the auto-translated English is not good enough to understand it, feel free to let me know at @nishio_en. I’m very happy to spread my thought to non-Japanese readers.