I can’t really disagree. Sabine is right that they’re similar situations on the surface in that the both represent large investments for extremely incremental gains, but AI takes the cost and grift to a whole different level while offering gains that have laughable value in comparison to even a single small step forward in our understanding of fundamental physics.

I suppose there are benefits to gig work like Instacart in that you get to pick your own schedule, but what is the benefit of driving for Amazon? You’re a contractor, but Amazon picks your hours?

If the state of open source phones are anything to judge by, we will have open source cars at some point, except the foot brake isn’t working yet, so you’ll have to use the hand brake for now. Cars and phones both take a lot of resources to develop, and maybe you’ll be able to “de-Stellantis” your car at some point instead of going fully open source, but judging by the recent steps Google has taken to weaken de-Googling, I’m not sure how long that would last either.

Not really my area of expertise, but this article lays out her perspective on this for anyone who isn’t aware: https://www.scientificamerican.com/article/the-world-doesnt-need-a-new-gigantic-particle-collider/

TL;DR - Many times the cost of the LHC and unlike the LHC, the gains are likely to be incremental instead of revolutionary. The same funding could do much more good elsewhere.

To your point, agreed that even small, incremental gains for science are more valuable than what we are likely to get from AI.

I have started using Copilot more lately, but I’ve also switched from plastic straws to paper, so I’m good, right?

I enjoyed Sabine’s analysis in another video that continuing to make increasingly larger models with more compute is about as effective as continuing to make larger and larger particle accelerators. Come on, bro, this million km Gigantic Hadron Collider will finally get us to the TOE. Just one more trillion, bro.

Yikes, I have a Samsung smart TV. Guess I’ll be wrapping it in tin foil after reading this.

I use Copilot at work and overall enjoy using it. I’ve seen studies suggesting that it makes a dev maybe 15% more productive in the aggregate, which tracks with my own experience, assuming it’s used with a clear understanding of its strengths and weaknesses. No, it’s not replacing anyone, but it’s good for rubber ducking if nothing else.

I use Linux for almost everything, but I do have some important software that only works on Windows, so my solution is dual booting Windows 10 with a different static IP than the Linux partition, with the Windows IP blocked from the internet in the firewall.