alyssavance

AI engineer

Wiki Contributions

Comments

FWIW I'm not convinced by the article on Haley, having bad conservative policies != being an anti-democratic nut job who wants to rig elections and put all your opponents in jail. She's super unlikely to win, though.

Most highly educated people lean left, but there really are just very few Stalinists. I quoted a poll above showing that just 8% of Americans would support an AOC hard-left party, and actual Stalinists are a small fraction of that. There's no developed country where tankies get a serious fraction of the vote. See Contrapoints for why communist revolutionaries are super unlikely to take power: https://youtu.be/t3Vah8sUFgI

There are many crazy professors of various stripes, but universities aren't states. They can't shoot you, can't throw you in jail, can't seize your house or business, and are ultimately dependent on government funding to even exist.

The current constitution isn't that old (although 65 years is still longer than most democracies), but with brief interruptions, France has been a democracy for around 150 years, which is far longer than most countries can claim. 

Thanks for the response! Here are some comments:

 - India, Turkey, and Hungary are widely referred to as "hybrid regimes" (https://en.wikipedia.org/wiki/Hybrid_regime), in which opposition still exists and there are still elections, but the state interferes with elections so as to virtually guarantee victory. In Turkey's case, there have been many elections, but Erdogan always wins through a combination of mass arrests, media censorship, and sending his most popular opponent to prison for "insulting public officials" (https://www.bbc.com/news/world-europe-63977555). In India's case, Modi is no doubt very popular, but elections are likewise hardly fair when the main opponent is disqualified and sent to prison for "defamation" (insulting Modi). Rather than being voted out, hybrid regimes usually transition to full dictatorships, as has happened in (eg.) Russia, Belarus, Venezuela, Nicaragua, Iran, etc.

 - Of course nothing is certain, but France's president is very powerful, and this article discusses in detail how Le Pen could manipulate the system to get a legislative supermajority and virtually unlimited power if elected: https://foreignpolicy.com/2022/04/20/france-election-le-pen-macron-constitution-separtism-law-state-of-emergency-referendum/

 - I made a map here of roughly how much electoral support the far right has in each country worldwide (https://twitter.com/alyssamvance/status/1656882958903418880). However, it is tricky to forecast based on this, because far-right parties can appear from nothing and gain a wide support base very quickly (as happened with eg. Chile's Republican Party).

I'm surprised by how strong the disagreement is here. Even if what we most need right now is theoretical/pre-paradigmatic, that seems likely to change as AI develops and people reach consensus on more things; compare eg. the work done on optics pre-1800 to all the work done post-1800. Or the work done on computer science pre-1970 vs. post-1970. Curious if people who disagree could explain more - is the disagreement about what stage the field is in/what the field needs right now in 2022, or the more general claim that most future work will be empirical?

I think saying "we" here dramatically over-indexes on personal observation. I'd bet that most overweight Americans have not only eaten untasty food for an extended period (say, longer than a month); and those that have, found that it sucked and stopped doing it. Only eating untasty food really sucks! For comparison, everyone knows that smoking is awful for your health, it's expensive, leaves bad odors, and so on. And I'd bet that most smokers would find "never smoke again" easier and more pleasant (in the long run) than "never eat tasty food again". Yet, the vast majority of smokers continue smoking:

https://news.gallup.com/poll/156833/one-five-adults-smoke-tied-time-low.aspx

There are now quite a lot of AI alignment research organizations, of widely varying quality. I'd name the two leading ones right now as Redwood and Anthropic, not MIRI (which is in something of a rut technically). Here's a big review of the different orgs by Larks:

https://www.lesswrong.com/posts/C4tR3BEpuWviT7Sje/2021-ai-alignment-literature-review-and-charity-comparison

Great post. I'm reminded of instructions from the 1944 CIA (OSS) sabotage manual:

"When possible, refer all matters to committees, for “further study and consideration.” Attempt to make the committee as large as possible — never less than five."

Eliezer's writeup on corrigibility has now been published (the posts below by "Iarwain", embedded within his new story Mad Investor Chaos). Although, you might not want to look at it if you're still writing your own version and don't want to be anchored by his ideas.

https://www.projectlawful.com/posts/6075?page=21

Load More