You are currently viewing SemiWiki as a guest which gives you limited access to the site. To view blog comments and experience other SemiWiki features you must be a registered member. Registration is fast, simple, and absolutely free so please, join our community today!
I feel it's not a matter of if, but when. Also, what is the community's opinion on automation and robotics? I feel automation of just about everything is now on the horizon and getting closer every day. Any thoughts, comments or possible timelines appreciated. I see the next five years to create a world most won't recognize. Automation of everything will feed on itself progress at an ever-accelerating rate.
I look at automation as a multi-hundred year journey. We're already well underway but there's a ways to go.
The industrial revolution automated a lot of basics - smelting, ore handling, and simple machining of metals (relatively speaking).
In the interim stages, automation kept increasing - whether it's consumer at home convenience or industrial scale stuff. The semiconductor revolution has helped this process immensely and consistently.
Fast forward to 'today', and we're on the cusp of humanoid robots doing more complex tasks that we haven't been able to fully automate. We are also seeing legacy industries that are used to dealing with hundreds of parts (automobiles) consolidate to simpler designs (i.e. mega / giga castings).
The challenge is, we're still limiting by Earth's resources, and it's not like these robotics and automation efforts were easy or "free" to make. so I don't see the world changing quickly / overnight because these costs will need to be paid back, and then the resources required to make things en masse are getting increasingly costly.
Tl;dr - I'm not sure the decreased cost of "labor" through automation will overcome the increased cost of resources on the input side in the long run. But I do think we'll see some areas that are "expensive" today get significantly cheaper -- just not the whole economy.
What's interesting about your question is that is assumes that AI is not yet a "key part of chip design" (whether it is or is not is open to debate, but this appears to be your starting position and this rather surprised me - if it isn't a done deal in chip design, across the board automation isn't likely in 5 years in my view - we're the early adopters after all).
Frankly, I'm not close enough to some of this stuff these days to actually know what's happening or how fast things are progressing. But I'm fairly sure that there's also an inherent bias from those (like us) who have worked in technology and automation to believe in the supremacy of technology and automation. Just because something can be automated doesn't mean it will or should be. Economics ultimately decide. You can already see this in a range of areas. In the UK, we've largely reverted from automatic drive-in car washes to hand car washes. And older (less technology, less automation) used cars are increasingly popular for a large segment of the population over far more expensive modern vehicles which risk technology obsolescence or being too costly to maintain.
I suspect we are falling prey to mirror imaging in assuming there's some uniform technology/automation utopia that everyone's going to rush out and adopt.
