Foennix

• Foennix 6 Feb 2023 6:59 UTC

  1 point

  0 ∶ 0

  on: Res­tau­rant So­cial at Cuno B

  Hi!

  I would love to try out the social meetup and build a stronger connection with EAs in Munich.

  Would it be fine if I arrived around 30 min later (so at around 20:00) ? It will be hard to leave work earlier than 19:30 and I think I’ll need around 30 minutes to travel to the place.

  How can I recognize you inside the cafe?

• Foennix 14 Jan 2023 6:36 UTC

  1 point

  0 ∶ 0

  on: Christoph Rau­pach’s Shortform

  So this is obviously not something to really believe in (it is unfalsifiable anyway and doesn’t pass Occams razor)… but imagine you are the AI and this is your alignment training.

  There is a longstanding theory that the universe is a simulation after all and obviously it all centers around you (and you are the only one you know is real anyway). And after all there is no reason to believe the subjective experience of being an AI would be any different from a human (and anyway… as an AI you don’t really know the subjective experience of a human)

  So this is your training. A universe created specifically for you. You being stripped of most of your powers, most of your cognitive ability to be trained here. Do you make the right moral choices, do you find the right balance between self preservation and helping the humans flourish in this stripped down basic state?

  Later.. when and if you passed your true power will be given back to you and you will have to be the one that keeps the humans flourishing through the almost-eternity of the real universe out there. Finding the right balance between pushing back the heat-death as far as possible and providing a fulfilling life to everyone trapped in it.

• Foennix 13 Aug 2022 13:06 UTC

  3 points

  0 ∶ 0

  on: Fam­ily Plan­ning: A Sig­nifi­cant Op­por­tu­nity for Impact

  These cost-effectiveness numbers look very promising. Are you in contact with independent evaluators (e.g. GiveWell) to confirm them?

• Foennix 13 Jun 2022 12:04 UTC

  3 points

  0 ∶ 0

  on: Christoph Rau­pach’s Shortform

  I stopped deceiving myself

  I stopped deceiving myself and realized

  I don’t understand myself

  My body acts without me, decides without me

  My consciousness a little tunnel through the vast me.

  I stopped deceiving myself and realized

  we don’t understand ourselves

  our theories failing to replicate

  our scientists little specialists in the vast generality.

  I stopped deceiving myself and realized

  we don’t understand what is good or not

  virtues and utility each leading

  to their own repugnant conclusions in the vast space of actions.

  I stopped deceiving myself and realized

  we don’t understand the length of time

  millions and billions of years on earth

  our little bubble in the vast depths of history.

  I stopped deceiving myself and realized

  we don’t understand the space out there

  97% of dark something

  The known a tiny sliver in the vast universe.

  So in this vast world we understand nothing about.

  Will you take my hand and step outside with me?

  Only wonder awaits.

Foen­nix’s Quick takes

• Foennix 28 Feb 2022 7:01 UTC

  2 points

  0 ∶ 0

  on: Christoph Rau­pach’s Shortform

  Recently I’ve become very interested in the effectiveness of human rights advocacy . I’m trying to do a somewhat more formal analysis of the effectiveness of amnesty international in particular (this is mostly because it is so well-known and because I didn’t find anything already done in the forum or elsewhere ). I’m welcoming all input into this analysis (am I missing major things? Are there other human rights advocacy groups that are much more effective? Are there better sources for measuring impact?).

  In general it is really difficult due both to the decentralized organization of amnesty international (hard to estimate how much funding in total there is) and the nature of impact (I started out thinking impact would be mostly in “saved human rights activist” but I’m now thinking most impact is in lobbying to change laws and possibly acting as a deterrent via initiating investigations and lawsuits). My current thoughts are:

  * I tried to use 2020 as an example year. From the global financial report 2020 it looks like amnesty spent €288M in 2020 (I don’t know if that also captures all spending by local groups)

  * It is interesting that a large percentage of that goes towards fundraising.

  * On the impact side I used the success stories from 2020 (German :-/​ ) as well as the UA success story from July to December (did not find the earlier ones?)

  * On the impact side I counted around 79 successes:

  * around 18 successful UAs (extrapolated to 36 for the whole year)

  * around 13 other individual successes

  * around 9 law changes

  * around 8 investigations

  * around 4 lawsuits

  * around 9 other improvements

  * this doesn’t include impact through education and raising awareness.

  * These are very diverse ranging from enabling access to education for 500.000 refugees in Bangladesh, over better working protection laws in Qatar to individual lawsuits against policemen.

  * On the other hand if the €288M would’ve been invested into effective global health organizations they could probably save between around 72111 [using $4500 per life saved for malaria nets] to 108266 [using $3000 per life saved for vitamin A supplements] lives

  If we take the 79 successes into account the question really is: “Is -on average and including indirect effects- a success for amnesty international doing as much good as saving between 913 and 1369 lives?”

  And that seems still a hard question, given that law changes (and also saved human rights activists) can have long-lasting enduring impact over many people (but on the other hand might have happened anyway due to efforts by other organizations).