zrezzed

I’m disappointed this is the first comment on this post.

gwern’s writing (including on nicotine) was formative for me; it showed me how and why the internet the was important: it let me read good, well written thinking I had never seen from the NYTs or my parents.

I first saw a link to gwern.net on HN. And I trusted the NYTs as an institution then, and do to this day… and I’m sure I clicked through, and took the gwern post seriously in part because comments weren’t universally negative.

You can point to bounded trust problems, or talk more about how “The Media Very Rarely Lies”…

But please don’t take up the first comment on a gwern post to cheap shot the NYTs

Long-form slop. Mmm…

Ah yes, the Backrooms of UX hell.

With the right procedural generation and some decent flavor, that could be genuinely horrifying!

This matches my understanding, thanks. I thought I was going crazy reading other comments.

I’d bet the author does understand.

I’m fairly certain subsequent runs even with the same prompt will make selections with different entropy. Even with low temperature, there’s no reason a good enough model couldn’t start such a list with “Right” 20% of the time.

I agree it’s equivalent, and that’s a great way to think about it.

But… I wouldn’t be surprised if humans answered closer to the LLM results than 80:20. I’d actually be surprised if humans converged very close to the right result.

Would be a fun mechanical Turk experiment to run.

I think of it as "the first time anyone has not lost". Until that crash run, every game had ended with a top out.

> turn the door handle, and unplug the rack.

There are door handles that you cannot turn, and racks you cannot unplug: ones buried deep in the Cheyenne Mountains and hidden far away in the Siberian tundra. They are protected by unfathomably powerful systems, with the support of countless people and backed by the whims global economic power.

I say this as someone who likely agrees with you. I think the power of the real world, that of companies, and governments and militaries should decrease our concern with AGI gaining power itself.

But I don’t think it’s as obvious as pointing to the fragility of software. Our human systems are fragile too, and subject to manipulation in not-so-different a way as data centers. I think you should not be so quick to discount the voices of many smart people shouting.

Sure. Most of my experience is in California, but I've found the grading differences from region to region to be just large as from out of state or international.

I climb 5.11c+ comfortably in the gym. I can barely lead a short Yosemite 5.8 outdoors; it's terrifyingly more difficult. That's a bit on the high side of a delta, but still typical.

I would say the difference is 75%+ in the scales simply not being the same. For the same grade, the holds would be far smaller, positions much more technical, often far more sustained, etc. The other ~25% is the environment is much more challenging, even with perfect weather at sea level: route finding is harder, you have to manage gear, leading is scary, etc.

I've done a few high Sierra climbs. Grades there are little softer, and my limit is probably ~5.7

If you care about strongly about privacy... choose your email provider accordingly.

If Outlook is your company's email provider... the third parties are the concern of your IT department.

If you're not hit by one of those two conditions... why are you using Outlook? This feels like unwarranted outrage bait.

> Dismissing a multi-millenia multi-cultural fundamental norm as "indoctrination" severely downplays the collective learning of uncountable generations.

I agree in this case, but only because I think marriage is an institution that we’ve learned needs (sometimes strong) active endorsement and it’s pro-social.

My view is that cultural norms are at worst tyrannical and at best the essence of being human.

> Social iconoclasts dismiss organized religion, marriage, sexual norms, social duty, gender roles, and cultural traditions.

This is just true. But this is all of us: we all have a relationship with society around us and dismiss parts of it. As we should.

Uh, do you remember what youtube comments were like 2016-ish and earlier?

It seems obvious to me the overall experience of youtube comments is much more dependent on youtube's choices than people's behaviors.