psmirnov

Reasoning models seem to answer this correctly (tested Gemini pro 2.5, ChatGPT + reason). Grok answered correctly even without turning on "Thinking".

It is a published abstract from a conference, not a complete article. Journals don’t make it obvious but if you click on the link to view the issue you’ll see that the issue title refers to an annual meeting.

Interestingly, controlled nuclear fission preceded the weapons tests (Chicago Pile 1), at least chronologically.

I was going to post to say exactly the same thing. I recently moved Ontario, Canada to BaWü Germany. Yes, the kWh price is 3 times higher here, but I can find a contract where I pay about 9 euro a month for delivery, compared to $40 CAD I was paying (after rebates) every month on various fees back in Ontario. Most months I paid more for delivery than my usage back home. It’s nice to have the majority of my energy bill under my control.

Normalizing prices to USD also makes little sense, as I don’t get paid in USD in either country. Normalizing to daily income makes more sense, even though that’s not a perfect measure of affordability (I pay more for electricity but less for internet/cell service, for example). A more interesting measure to me is how many people struggle to pay an average household electricity bill.

That said, I completely agree with other commenters that Germany’s closing of nuclear plants was a big mistake, if not from the energy pricing point of view, then from an environmental impact perspective. I can see a coal burning plant in the far distance from my office window, and I’d much rather that thing was shut down then the nuclear plants.

The Kerry is a must have for me; it got me through my maths degree. The simplicity of being able to close the cap and toss it in a jacket or jean pocket without worrying about pencil marks meant I always had a pencil on me and could write down things as soon as they clicked in my head.

I agree with almost all of this, however I believe that publishing random seeds is dangerous in its own way.

Ideally, if your code has a random component (MCMC, bootstrapping, etc), your results should hold up across many random seeds and runs. I don’t care about reproducing the exact same figure you had, I want to reproduce your conclusions.

In a sense, when a laboratory experiment gets reproduced, you start off with a different “random state” (equipment, environment, experimenter - all these introduce random variance). We still expect the conclusions to reproduce. We should expect the same from “computational studies”.

This should be higher up. Without any judgement on the author as a person, they live in different circumstances than 99% of the employers, employees or customers in the restaurant business. Knowing this is an important part of understanding their social commentary.

Link please?