ryan_b

Karma: 39

ryan_b 27 Sep 2021 19:50 UTC
−2 points
0 ∶ 0
in reply to: Gregory Lewis’s comment on: Clarifying the Petrov Day Exercise
I generally think EA- and longtermist- land could benefit from more ‘professional distance’: that folks can contribute to these things without having to adopt an identity or community
I am curious: if you believe professional distance is a good thing for EA, then what is your explanation for EA not being a solved problem already? The existing networks of professionals had all the analytical knowledge required; save for the people who joined up straight out of college pretty much everyone currently working in EA was already in those networks.

ryan_b 3 Sep 2021 19:32 UTC
3 points
0 ∶ 0
on: What we learned from a year incubating longtermist entrepreneurship
I’d be interested to hear more about the thoughts behind this key lesson:
The LE Project should be divided into several different organizations/efforts, rather than one incubator.
This makes sense to me in light of the different tasks and operational requirements that different purposes are likely to require, but I noticed a theme running through the rest of the report of uncertainty. This included things like funders being uncertain of downside risk; uncertainty about what actions a person with funding should take; an expertise/experience bottleneck for longtermism and especially the combination of longtermism and entrepreneurship.

What do you think of the relationship between the constraints and having multiple orgs in the space?

ryan_b 9 Apr 2021 13:34 UTC
4 points
0 ∶ 0
in reply to: aogara’s comment on: EA Debate Championship & Lecture Series
I appreciate this nuanced comparison between formats. I increase my estimation of the goodness of the EA debate project, because this introduces a new pathway to victory:

The creation of an EA debate format, which can be designed against the backdrop of all the poisoned formats to encourage skills and norms more like those we want to develop.
It does feel like the general period for research before the debate would work well for these topics. Do you know if any format has a mechanism for awarding points on the grounds of agreeing with the opponent’s points, or otherwise acknowledging when arguments or data weigh against you?

ryan_b 7 Apr 2021 20:14 UTC
9 points
0 ∶ 0
in reply to: Habryka’s comment on: EA Debate Championship & Lecture Series
...that video is deeply absurd, and I am caught between revulsion and laughter. They are clearly way past the point where there is even a gesture at real usefulness; this might be the most heavily gamed thing I have ever seen.
However, if the video is an example I am confident that this activity can’t do any real harm to our objectives. Who could do this with even the pretense of thinking they would talk to people in the real world that way? It’s just a verbal luge.

ryan_b 25 Mar 2021 16:17 UTC
2 points
0 ∶ 0
in reply to: FCCC’s comment on: Politics is far too meta
I think talking about political feasibility should never ever be first thing we bring up when debating new ideas.
I think this is much closer to the core problem. If we don’t evaluate the object-level at all, our assessment of the political feasibility winds up being wrong.
When I hear people say “politically feasible” what they mean at the object level is “will the current officeholders vote for it and also not get punished in their next election as a result.” This ruins the political analysis, because it artificially constrains the time horizon. In turn this rules out political strategy questions like messaging (you have to shoe-horn it into whatever the current messaging is) or salience (stuck with whatever the current priorities are in public opinion) or tradeoffs among different policy priorities entirely. All of this leaves aside enough time to work on fundamental things like persuading the public, which can’t be done over a single election season and is usually abandoned for shorter term gains.

ryan_b 20 Mar 2021 21:06 UTC
1 point
0 ∶ 0
in reply to: Vicky Clayton’s comment on: Improving Institutional Decision-Making: a new working group
It definitely is, mostly because there are so few successful projects to point to. Most of the work has been identifying what failed projects have in common, and then there are a few shining counterexamples against which they can test. It currently looks like the core insight is that planning needs to shift from controlling things to accounting for things you cannot control: lots of stakeholders (because many are attracted due to the sheer size of the project); black swans (in multi-year construction there is likely to be a bad storm but no telling when); the economy; etc.

ryan_b 18 Mar 2021 22:33 UTC
6 points
0 ∶ 0
on: Is Democracy a Fad?
I think this argument is harmed by imposing a democracy or dictatorship framework; while I understand the need to simplify, this obscures details that would be useful to us.
Dictatorship is pretty firmly anchored in fascism and communism, which depended strongly on effective centralized bureaucracies and rule of law to work. The kinds of things which could be accomplished by dictators of this era was simply beyond the scope or precision of all but a few rulers in the premodern eras.

I think following the thread of economic arguments is very valuable. In the industrialization section you mention the divisibility and immobility of land as being a factor; similar to this line of thinking is the condition of pastoral populations on the Steppe or Great Plains. The capital in these societies was all wrapped up in animals which were fully mobile, which makes oppressing people very difficult. Among the Mongols the position of khan was decided by election at the kurultai, where people voted with their feet in a literal fashion: in regional elections people joined the camp of the person they supported (along with their animals).
But economic arguments aren’t where the value lies per se. There was a type of kingship practiced in certain tribes in Africa which we would interpret as having a fundamentally religious function. Though the obvious trappings of power were there, like wealth and wives and servants, the actual function of the king was to be sacrificed in the event anything really bad happened. By this I mean droughts, famines, pestilence, and other natural occurrences over which the king had no hope of control. Now this method was never widespread, but I think if we sum up all the similarly-different-from-dictatorship traditional systems we will capture a huge chunk of the historical record.
To get around this problem, I think in the future the focus would benefit from shifting a level down, to mechanisms of power and the conditions needed to exercise them. Economics is pretty good at pointing this out, but as mentioned in the post we shouldn’t neglect the cultural elements like tradition or religious practices.

ryan_b 11 Mar 2021 16:56 UTC
2 points
0 ∶ 0
in reply to: konrad’s comment on: Is pursuing EA entrepreneurship becoming more costly to the individual?
And if you don’t have much status, no one will recognize a worthy attempt in case of failure.
This is a fairly harsh indictment of community norms. It directly implies there is nothing different about EA norms in this dimension relative to society at large, which is kind of a problem because there are well-known areas with superior norms; a well conducted trial reflects well on lawyers even when they lose.

Doesn’t make it wrong, naturally. But if true, it seems like it would definitely merit specific attention from the group.

ryan_b 10 Feb 2021 20:41 UTC
1 point
0 ∶ 0
in reply to: MichaelA’s comment on: Notes on “The Bomb: Presidents, Generals, and the Secret History of Nuclear War” (2020)
they seem to be saying/implying things more like “The US doesn’t need any nuclear weapons, and having them has no benefits at all”
Does Kaplan speak at all about how the US would otherwise deter a USSR invasion of our European allies? Things like alternative proposals for deterrence, or arguments that NATO was superior in an immediate conventional conflict, or similar?

My understanding of the beginning of nuclear deterrence policy is that it was the cheaper option compared to maintaining a sufficiently huge air and transport capacity to respond rapidly to a Soviet land invasion in Europe. It feels to me like his understanding of—or position on—these details would be a factor in interpreting his position on nuclear weapons.

ryan_b 10 Feb 2021 18:37 UTC
1 point
0 ∶ 0
on: Books / book reviews on nuclear risk, WMDs, great power war?
Another book which should be added to the list is The Great American Gamble: Deterrence Theory and Practice from the Cold War to the 21st Century. This book is notable because it covers the case for deterrence pessimism.

It does this because Dr. Keith Payne, the author, was a student of Herman Kahn’s, who was the other side of the deterrence conversation from Schelling. The central argument is that strictly offensive deterrence is not very credible and so poses an unacceptably high risk; the core difference in policy recommendations is to take deliberate steps to reduce the damage for your side, which they argue simultaneously reduces risk and increases the credibility of deterrence.

As a consequence, I think it speaks very directly to EA concerns about nuclear war.

ryan_b 21 Jan 2021 18:45 UTC
1 point
0 ∶ 0
in reply to: EdoArad’s comment on: Megaproject Management
This talk is great, and hits the exact same points as the paper. Would it be alright with you if I put the link to the talk in post with the other resources?

ryan_b 21 Jan 2021 16:32 UTC
1 point
0 ∶ 0
in reply to: Aaron Gertler’s comment on: Megaproject Management
Investigating the field more deeply this year is going to be one of my hobby projects, but my early impression is that a big part of the claim is that when you do regular project management things wrong, the penalty at least scales. It also looks like the cost of doing the analysis right, such as reference class forecasting, doesn’t come remotely close to scaling with the needs of the project.
I’m glad about this, because I was worried at first the whole inquiry might be useless except to people in a position of responsibility. Instead, it looks like there will be a lot of methods that are always a good idea but also scale really well. Bonus!

Megaproject Management

ryan_b20 Jan 2021 18:55 UTC

22 points

6 comments5 min readEA link

ryan_b 7 Jan 2021 21:55 UTC
2 points
0 ∶ 0
on: Improving Institutional Decision-Making: a new working group
I suggest looking into megaproject management. It grew out of regular project management, and focuses on the problem of very large projects because they go so badly so often. The problem they want to solve looks very big; they estimate that projects of this type, excluding economic stimulus programs and defense procurement, account for ~8% of global GDP.
I wrote a summary of one of the intro papers on LessWrong. The person driving the field appears to be Bent Flyvbjerg, who is at Oxford, though more researchers are involved.

ryan_b

Me­gapro­ject Management

Megaproject Management