[conspire] Fusion [was: Discussion: Using LLMs the Right Way: 10/1/2025 7pm Eastern Daylight time]

[Ron]

Steve Litt wrote on 2025-10-02 14:00:

> If the anti-LLM crowd had foundationed their anti-LLM complaints on the
> environment instead of "dumbing down programmers", they'd have been a
> lot more credible to me.

Boy, that sentiment sure feels familiar to me. I can't quite place my 
finger on where I hear it (ad nauseam) in technology these days.

Somehow years-long shouting to everyone may diminish.


> I keep hoping we'll have practical fusion energy in 10 to 20 years

I used to have that hope, but allow me to deflate yours too:

 > Why fusion will never happen

https://matter2energy.wordpress.com/2012/10/26/why-fusion-will-never-happen/


Basically, any tech advancements that help fusion will also help the 
competition (room-temperature super-magnets, etc.) also grow more efficient.

Which means fusion will have difficulty catching up to renewables.



Also, the financial angle: pay $20,000,000,000 for a bunch of 
electricity in (hopefully) 20 years (maybe) when the lenders can start 
to see a return or... an order of magnitude (or more) less money for a 
renewable solution that starts generating income in a couple years, via 
proven tech like wind farms.

Which are also highly scalable, easier to situate, can co-exist with 
other land use scenarios, and require far fewer highly skilled 
(expensive) workers.


My hopes were dashed by this fine article.

On the other hand, it does look promising for renewables.  And, it was 
written in October 2012 and it is probably more convincing now than it 
was over a decade ago.