It took me a few thought cycles to put down that question.

In my understanding, we, as a civilization, are on a high wire number over a pit of knifes. The AI craze is full throtle as we speak, with conflicting indicators that it may/will cause the worst economical crash ever - becoming third in less than twenty years, where each was always reported as an unprecedented and irrepeatable event - while at the same time infiltrating companies in ways we are still to discover how ruinous may prove and spreading over the internet like a grotesque parasite, rendering services unsuable and leaking sensible data.

If the great AI companies tank (I can only hope) and the services they provide - whatever those may be - faill, how much is it going to affect companies? There is really no sense of consensus; some say it was the best thing ever, others are registering neglictable improvements, others failed miserably and had to return to the original state of affairs.

Then there is the constant attack from agents on the net, rendering websites, more or less important, more or less useable or stealing, destroying or corrupting valueable data.

I've reached a point where I am willing to forego internet and even computers. Even technology in a wide sense. At best, I may keep an air gapped machine, to use for writing and printing documents. But will others do it? And if so, will 8t happen in an organic fashion or will things devolve to a point where something cracks and things just get ugly? Because people are tired. Tired of datacenters popping up like poisonous mushrooms everywhere, destroying everything.

People are getting tired. Everywhere.

you are viewing a single comment's thread
view the rest of the comments
[–] 8 points 1 day ago* (last edited 1 day ago)

Unless "to some degree" is doing an extreme amount of heavy lifting here, I don't expect so.

In general, it's a lot easier to build complex systems when error doesn't propagate through the system, and that happens with analog systems.

With a digital system, a given component has precisely-defined behavior. 1+1=2. And if you manufacture many instances of that component, they all can have the same, precisely-defined behavior.

With an analog system, there are tiny errors that accumulate. 1+1=2.0001, say. Then that tiny error gets passed on to the next component, and maybe the error there magnifies this error.

That doesn't matter much for some simple systems, but it becomes really problematic for complex ones.

Most complex systems we make are made out of simpler parts. Being able to cut that error out of the picture when specifying the behavior of a component or reasoning about it in the context of a larger system is helpful.

And in general, we want to be able to build complex systems.

Digital systems do force quantitization, and that is a limitation on precision. There will probably always be limited use of the occasional analog component, like, oh, a lens, say. Situations where you don't have a chain of many analog components.

But I don't think that we will return to a point where people made simple devices with few components, because the ability to make complex systems is very powerful. And complexity and analog systems don't play well together.

  • source