LESSWRONG
LW

pranomostro — LessWrong

pranomostro's Shortform

pranomostro

This is a special post for quick takes (aka "shortform"). Only the owner can create top-level comments.

Berlin AI Alignment Open Meetup April 2023

pranomostro

Hello everyone,

Once a month there is a meetup that's open for everyone who's interested in AI alignment (yes, that's also you!).

If you're doubtful whether to show up to the meeting, just show up.

I'm fairly open for anyone to attend, because I'm really interested in seeing who here in Berlin is interested in this topic and would like to do something about it :-)

We're meeting at the restaurant Focaccia Mia in Friedrichshain near Aurea.

I'm really looking forward to this, see you around :-)

Berlin AI Alignment Open Meetup January 2023

pranomostro

Hello everyone,

Once a month there is a meetup that's open for everyone who's interested in AI alignment (yes, that's also you!).

If you're doubtful whether to show up to the meeting, just show up.

I'm fairly open for anyone to attend, because I'm really interested in seeing who here in Berlin is interested in this topic and would like to do something about it :-)

We're meeting at the restaurant Focaccia Mia in Friedrichshain near Aurea.

I'm really looking forward to this, see you around :-)

Berlin AI Alignment Open Meetup December 2022

pranomostro

Hello everyone,

Once there is a meetup that's open for everyone who's interested in AI alignment (yes, that's also you!).

If you're doubtful whether to show up to the meeting, just show up.

I'm fairly open for anyone to attend, because I'm really interested in seeing who here in Berlin is interested in this topic and would like to do something about it :-)

We're meeting at Focaccia Mia in Friedrichshain near Aurea.

I'm really looking forward to this, see you around :-)

Berlin AI Alignment Open Meetup October 2022

pranomostro

Hello everyone,

we're a group of ~3 people trying to do independent AI alignment research in (some of) our free time. Once a month after one of our sessions, we have a meetup that's open for everyone who's interested in AI alignment (yes, that's also you!).

If you're doubtful whether to show up to the meeting, just show up.

I'm fairly open for anyone to attend, because I'm really interested in seeing who here in Berlin is interested in this topic and would like to do something about it :-)

We're meeting at Lei's Küche near the tEAmwork offices.

I'm really looking forward to this, see you around :-)

EA Berlin Public Meetup October 2022

pranomostro

Public socializing meetup at Peter Pane. Open to everyone, from people who are completely new to effective altruism to knowledgeable effective altruists who just want to hang out and have a relaxed chat.

Berlin AI Alignment Open Meetup September 2022

pranomostro

Hello everyone,

we're a group of ~3 people trying to do independent AI alignment research in (some of) our free time. This is a special occasion after EAGx Berlin, so we're not meeting up on the weekend.

If you're doubtful whether to show up to the meeting, just show up.

I'm fairly open for basically anyone to attend, because I'm really interested in seeing who here in Berlin is interested in this topic and would like to do something about it :-)

We're meeting at Eden Foodgarden near Hackescher Markt.

I'm really looking forward to this, see you around :-)

The meetup is now cancelled, sorry for the confusion.

Replying toInfra-Bayesianism Distillation: Realizability and Decision Theory

pranomostro4y

Infra-Bayesianism Distillation: Realizability and Decision Theory

One argument for infra-Bayesianism is that Bayesianism has some disadvantages (like performing badly in nonrealizable settings). The existing examples in this post are decision theory examples: Bayesianism + CDT performing worse than infra-Bayesianism.

(And if Bayesianism + CDT perform badly, why not choose Bayesianism + UDT? Or is the advantage just that infra-Bayesianism is more beautiful?)

Are there non-decision-theory examples that highlight the advantage of infra-Bayesianism over Bayesianism? That is, cases in which infra-Bayesianism leads to "better beliefs" than Bayesianism, without necessarily taking actions? Or is the yardstick for beliefs here that the agent receives higher reward, and everything else is irrelevant?

Berlin AI Safety Open Meetup July 2022

pranomostro

Hello everyone,

we're a group of ~5 people trying to do independent AI alignment research in (some of) our free time. Once a month after one of our sessions, we have a meetup that's open for everyone who's interested in AI alignment (yes, that's also you!).

If you're doubtful whether to show up to the initial meeting, just show up.

I'm fairly open for basically anyone to attend, because I'm really interested in seeing who here in Berlin is interested in this topic and would like to do something about it :-)

We're meeting at Lei's Küche near the tEAmwork offices.

I'm really looking forward to this, see you around :-)

Berlin AI Safety Open Meetup June 2022

pranomostro

Hello everyone,

If you're doubtful whether to show up to the initial meeting, just show up.

I'm fairly open for basically anyone to attend, because I'm really interested in seeing who here in Berlin is interested in this topic and would like to do something about it :-)

We're meeting at Lei's Küche near the tEAmwork offices.

I'm really looking forward to this, see you around :-)

Replying toAre there English-speaking meetups in Frankfurt/Munich/Zurich?

pranomostroJun 13, 2022

Are there English-speaking meetups in Frankfurt/Munich/Zurich?

There used to be an SSC/ACX group in Munich that I organized, maybe shoot a mail to ssc_munich@lists.posteo.de to see if anyone responds? (Might take a day or so since I will have to approve you as a sender).

If anything fails, there is a very lovely EA group in Munich that I participated in :-) Information here.

[Rough notes, BAIS] Human values and cyclical preferences

pranomostro

pranomostro, Jayjay, Lucie Philippon

epistemic status: Working notes of three different people on the same question, likely useless/incomprehensible to anyone else

The question

How to find the right abstraction level of human values

Problems in over- or underfitting human values:

We can learn human values by observing their actions and distilling them into a preference relation. This learned preference relation can overfit human values (eg: Humans want to raise their left arm by 2 cm on 2022-05-07 if they’re in some specific place) or it can underfit human values (eg: Humans care only about maximizing money). If our preference relation overfits, we expect to not find some known biases, e.g. the Allais Paradox. There are also both inconsistencies that are... (read 1054 more words →)

Replying toOvercoming Akrasia/Procrastination - Volunteers Wanted

pranomostro7y

Overcoming Akrasia/Procrastination - Volunteers Wanted

I'm interested.

Replying toRecommendation Features on LessWrong

pranomostro7y

Recommendation Features on LessWrong

Just some quick feedback on the "Continue Reading" feature: At the moment, when I read a post in the middle of a sequence, the next recommended post is at the beginning of the sequence, but I would like it to be the post after the last one I read in the sequence. Perhaps this is intentional, but I wouldn't use the feature that way, since I already try to read the posts in sequence (sometimes without being logged in).

Replying toAsk LW: Have you read Yudkowsky's AI to Zombie book?

pranomostroMar 17, 2019

Ask LW: Have you read Yudkowsky's AI to Zombie book?

I am currently reading it, currently in the Quantum Physics sequence. I read it all here on LessWrong, I did not buy or read the book version. I sometimes skim through the comments a bit, but sadly, the threads have been unraveled a bit and it is hard to follow a conversation. I don't remember any specific occasion where the comments enlightened me in a new way, though they are sometimes interesting. I doubt it is necessary to read them, though.

Replying toThe tech left behind

pranomostro7y

The tech left behind

Plan 9 from Bell Labs comes to my mind (papers & manpages): By the creators of unix, tight integration of networks (better than other systems I have seen so far), UTF-8 all the way down, interesting concept with process-wide inherited namespaces.

It used up way too many weirdness points, though, and was fighting the old Worse is Better fight. It lost, and we are left with ugly and crufty unices today.

Another one that comes to mind is Project Xanadu. It was quite similar to the modern web, but a lot more polished and clean in design and concept. It probably failed because a really late delivery and by being too slow for the... (read more)

Replying toSages in singularity

pranomostro7y

Sages in singularity

You were downvoted, but I think somebody should try to explain exactly why.

Both the idea and the terminology for "sages" is highly questionable. It evokes the idea of mysterious answers and to arguments from authority which are exactly what we would want to avoid.

Could you maybe clarify a bit more what you mean by the word "sage"? It seems like you conflate people who want to solve problems and win with people who want deep insights into the nature of meaning.

Replying toConditional Independence, and Naive Bayes

pranomostro7y*

Conditional Independence, and Naive Bayes

Possible (?) typo: "V is "Two arms and ten digits"" could indeed be meant as "V is "Two arms and ten fingers""

Nevermind, didn't know digits was another word for fingers.

Replying toProductivity: Instrumental Rationality

pranomostro7y

Productivity: Instrumental Rationality

I started tracking my productivity at the beginning of this month, writing a "master plan" in order to know at each moment exactly what I should do next (okay, not "exactly" exactly, but good enough for it to in theory fill more than one day).

I realized how bad it is. Which is excellent.

I'm not sure how much to include in the plan. At the moment it is so big that if I had ultimate self-restraint and would waste not one minute of the day, I would barely get it done. It seems like that's okay, since I have sorted the activities after their priority and I have been improving since I started.

But does anybody have experience in trying out strict/less overwhel plans and observing if (and if yes, how much and in which direction) they influence success?

LESSWRONG
LW

LESSWRONG
LW

pranomostro

pranomostro

pranomostro

pranomostro's Shortform

[Rough notes, BAIS] Human values and cyclical preferences

Welcome to Effective Altruism Munich

pranomostro

pranomostro

pranomostro

pranomostro's Shortform

[Rough notes, BAIS] Human values and cyclical preferences

Welcome to Effective Altruism Munich

The question

Problems in over- or underfitting human values: