sdr

Karma: 271

sdr 5 Oct 2022 9:09 UTC
4 points
2
in reply to: Daniel Kokotajlo’s comment on: Open & Welcome Thread—Oct 2022
Ben Thompson ( https://stratechery.com/ ) , an American industry analyst currently living in Taiwan has a bunch of analyses on this on his blog. In nutshell, the US has a critical infra dependency on Taiwan in high-performance chip manufacturing; specifically, TSMC has a 90% share of 7nm, and 5nm chips. This is critical infra, for which the US does not have good (or even close-enough) substitues. Based on both these economic incentives, and Biden’s own statements, the US is extremely likely to reply to Chineese aggression against Taiwan with military force.

sdr 19 Aug 2022 7:11 UTC
9 points
0
on: What’s up with the bad Meta projects?
Cross-posting some thoughts:
Facebook’s metaverse strat is focusing heavily on capability / platform, and not content / single-awe-of-moment. To them it’s possibly okay if vrchat wins at the expense of horizon worlds, _just as long as majority of peeps access it via quest_ which they do: https://metrics.vrchat.community/?orgId=1&refresh=30s ← quest users now outnumber pc ones 1:2.
Consider the apple & appstore fiasco, whereby apple can basically, in one OS update, kill retargeting by introducing privacy popups into apps at os level, kneecapping the entire ad industry (single major reason for FB’s this quarter _decline of revenues for the first time ever_); and unilaterally decide, that everyone who takes payments for digital services, and has an an app on ios (read: entire B2C SAAS market) now has to pay 30% to them. _and make it a reality_ on pain of removal from app store. _and it works_.
Basically, FB wants to position itself into the same capability / platform play -they control the device, they can dictate terms for everyone building on top of it.

sdr 12 Jul 2020 3:41 UTC
3 points
in reply to: johnswentworth’s comment on: Kelly Bet on Everything
Oh darn, you’re right. Thank you!

sdr 11 Jul 2020 21:54 UTC
3 points
on: Kelly Bet on Everything
I’m running simulations to get a feel for what “betting Kelly” would mean in specific contexts. See code here: https://jsfiddle.net/se56Luva/ . I observe, that given a uniform distribution of probabilities 0-1, if the maximum odds ratio is less than 40/1, this algo has a high chance of going bankrupt within 50-100 bets. Any thoughts on why that should be?

sdr 23 Jun 2020 13:32 UTC
0 points
in reply to: Mati_Roy’s comment on: SlateStarCodex deleted because NYT wants to dox Scott
-

sdr 30 May 2019 4:44 UTC
6 points
in reply to: Said Achmiz’s comment on: What is required to run a psychology study?
In the context of customer development for product research, yes. For good questions on that, see eg the book “Mom test” by Rob Fitzpatrick, and lean customer development field in general. This was solving for the general question “will developing x be paid for”; being wrong on this particular question is expensive.

sdr 29 May 2019 7:57 UTC
23 points
on: What is required to run a psychology study?
In the name of supporting people actually doing stuff:
- Scott’s IRB Nightmare comes from the circumstance of polling taking place within the context of privileged patient-provider interaction, which is covered by HIPAA, which requires somewhat stringent data handling. If you are not a doctor, and you’re not asking your patients in the hospital, this does not apply to you.
- Yes, you are allowed to “just go out and ask a whole bunch of people stuff”. People can, actually, give away whatever information they feel compelled to do so. People are allowed to enter (mostly) any trade. People are free to do stuff.
- For people <18, you need parental consent.
- There are, like, hundreds of tools to do this -both finding people, and nailing the questions. Google Survey samples currently best across US (specifically, it had predicted the 2016 election results successfully). This is good, if you have a specific hypothesis, that you want to ask from 1000++ people.
- The more quantitative you get, the less signal it carries, at higher precision. Survey & stats criticism generally comes from attempting to determine “things about humanity in general”, which is also (somewhat) useful, but requires N > very large, and _very_ methodical sampling / experiment formulating / etc.
- Generate qualitatively, validate quantitatively. Vast majority of effort goes into actually locating the hypothesis. Before building a research thesis in your room, go out and do the simplest thing first. Talk with people, like, in-person. There’s a learning curve prior to being able to formulate meaningful hypotheses.
- Ask yourself, what rent does answer to a specific question pays. What does it say of reality if it turns out to be A vs B? How does that interact with neighbouring things?
  - And: what, specifically, do you wish to achieve here? Specifically, some qualitative answers to some of the questions above from Bay Area people, along with some synthesis, would be extremely informative (to me at least).
- A good starting point for this might be cultural anthropology, but instead of getting a book, here’s an MVP: get a tape recorder,ask 50 of your friends the questions above, then put answers into a spreadsheet, and a synthesis into an lw post. This is extremely informative for eg measuring local shifts in the overton window, finding common ground (and grounds shifting); and is sorely missing.
  - Why in-person? People who persistently fill out textareas on web pages are heavily biased in income, and mental illnesses; generally, people don’t do that. Being in-person, you’re raising the interview against personal reputation, which bridges the addressibility gap, and makes a much wider variety of people’s voices accessible.
  - To avoid pet-theory issue: Ask open-ended questions (eg the relating to job/ambition ones above are good). Don’t lead, capture the raw stuff
- Do this simple thing first, prior to embarking on specific hypothesis formulation; and post the results!

sdr 23 Mar 2018 2:09 UTC
6 points
in reply to: Raemon’s comment on: LW Migration Announcement
Not grandparent, but browsing through my private notebook for potentially breaking links, eg:
http://lesswrong.com/r/discussion/lw/deg/less_wrong_product_service_recommendations/6yry ← which is one specific advice (and a good one at that) vs https://www.lesserwrong.com/r/discussion/lw/deg/less_wrong_product_service_recommendations/6yry ← which is 404. This actually do have a high impact both on other sites linking specific comment threads, and by extension, on SEO in general (linked page with content changed to empty).
( Relatedly, https://www.lesserwrong.com/non-existing-page returns HTTP 200 instead of 404, which is more wrong, than http://lesswrong.com/not-existing-page )

sdr 4 Jan 2018 5:38 UTC
8 points
on: Superhuman Meta Process
Hello my values a decade ago, it’s so nice to see you publicly documented! In retrospect & in particular, the level of paranoia imbued here will serve you well against incentive hijacking, and will serve as a foundational stone in goal stability.
There is one particular policy here, where my thinking has changed significantly since then; and I’d love to check against Time whether it makes sense, or has my values been shifted:
| Reject invest-y power. Some kinds of power increase your freedom. Some other kinds require an ongoing investment of your time and energy, and explode if you fail to provide it. The second kind binds you, and ultimately forces you to give up your values. The second kind is also easier, and you’ll be tempted all the time.
Contrast with:
| Optimization never stops. Avoid one-time effort if at all possible. Aim for long-term stability of the process that generates improvements. There is no room for the psychological comfort of certainty.
So, the operative word above is “freedom” (personally, I’ve used “possibility space maximization”), and it’s super useful to run a conceptually exhaustive search across surface-y options . But.
You probably have goals of interest, that you wish to achieve (eg “long-term future of humanity”). Some of these might require banging at stuff for an extended period of time. You have behaviours (eg your meta-policies), which you do for an extended period of time. Whether you recognize it as such, or not, you are also vesting into these; and by way of the forgetting curve, and blog readership, they also require ongoing maintenance. And yes, there might come future technological change which will make them obsolete, and put you into the decision between “your values” & “rolling with changes”.
So, my counter to this is, _Anything which does not take into consideration the passage of time, gets eaten by it._ Your Time is a super scarce resource -probably the scarcest of them all. One way to turn this liability into an asset is by vesting into stuff (projects, startups, skills, people, ideas, what have you), and riding the compounding interest across time. This is, to my knowledge, the only way one can scale scarce resources into epic levels of task-specific utility.
(Relatedly, it seems to me, that there is a sliding scale between the need for change in the face of future changes and vesting into things, that most people tend to shift through as they age. Obvious problem here is simulated annealing being susceptible to fixation on phantom (local) maxima by way of changing environment.)
So, unpacking the desiredata from above, the model I’d offer for consideration is the Affordable Loss Principle, with a side dish of Avoiding Infinite Optimizers:
* The affordable-loss principle: prescribes committing in advance to what one is willing to lose rather than investing in calculations about expected returns to the project. Key to affordable loss policies is generation of Next-best-alternatives, such so when it comes to move, there is something to seamlessly move forward to.
Or, in the wise words of Zvi: https://www.lesserwrong.com/posts/ENBzEkoyvdakz4w5d/out-to-get-you
- Get Got when the deal is Worth It.
- When you Get Got, do it on purpose.
- But, You cannot afford to Get Got if the price is not compact. (Sufficiently advanced optimizers will eat your time, attention, and resources for breakfast _if you let them_ . Don’t. )
In conclusion, I’d suggest that yes, run a freedom-maximizing circle, because it eliminates conceptual blindsight, and there is a lot of low-hanging fruit you can pick up on your way. But additionally, be on the lookout for opportunities that are compact, low-hanging, and compounding across time, such so that linear investments today leads to incremental & compounding utility for tomorrow.

sdr 26 Apr 2017 4:46 UTC
5 points
in reply to: k_ebel’s comment on: Project Hufflepuff: Planting the Flag
Thank you for posting this. I agree, that growing negotiation skills is hard under best of circumstances; and I agree that certain types of newbies might self-identify with the post above.

There is a qualitative difference between people who are negotiating (but lack the proper skill), and the parasites described above:
- Beginner negotiators state their request, and ask explicitly (or expect impliedly) for price / counter
- More advanced negotiators start with needs/wants discovery, to figure out where a mutually beneficial deal can be made; and they adjust as discussion proceeds
- These parasites, in comparison, attempt to raise their request against explicitly stated, nebulous things (or nothing at all): “Would you like to do free translation for me?” - “Cause X is very important, and therefore you, specifically, should do something about it”—“Would you like to build my full website for me in exchange of 1% shares?”
For the record:
- I have attempted education in some cases (1-on-1, no social standings on the line on either parties, being discreet, etc), to no effect, and only resentment from the other party.
- I observe that this parasitic strategy works some of the time, which incentivize existing parasitic behavior to grow until saturation. These are the reasons why I brought this up here in the first place.
- Kindly note, that while there were a lot more evidencing going into this than described above, I am hesitant to disclose more specificities about any of these cases, because the Bay is small (-> personal identification), and discussion isn’t reflective-complete (parasites read this, too; the more I disclose here, the more they can shift their strategies)

sdr 4 Apr 2017 16:55 UTC
0 points
in reply to: Viliam’s comment on: Project Hufflepuff: Planting the Flag
Updated. Re: | if you want to publicly address these people ← if people are addressed offline in public, I suspect you can dress it up with the appropiate social grace. But, we’re talking about behavior here (and entrepreneurs have exploits they’re already proud of, like hackers have hacks, and free riders aren’t actively malicious), and I feel that dressing it up with the same grace would actually backfire by not changing (or even harming) the reward structure of the behavior.

sdr 4 Apr 2017 3:57 UTC
5 points
in reply to: Viliam’s comment on: Project Hufflepuff: Planting the Flag
Agreed. Recommend a non-verbed descriptive noun, and I’ll update the post above.

sdr 2 Apr 2017 3:51 UTC
11 points
in reply to: Raemon’s comment on: Project Hufflepuff: Planting the Flag
(I’m not sure which part of this is “armchair-theorizing-sociology piece”, so let me share impressions:
- The 3 specific examples are all observations: 2 on a CFAR event, 1 on a bay-lesswrong event
- The “people putting other’s needs ahead of their own” comes from 2 persons who both bounced from the Bay for this reason
- The “attempting value-pumping” / lack-of-dealcraft is ubiquitous everywhere where people are Getting Stuff Done; the only novel thing in the Bay is high turnaround / people onboarding allows this to be done systematically
- The “let’s make stuff suck less” → “let’s all of you do my stuff” headfake is a non-profit-special; 2 attempts so far on me
- The part where instead of attempting to “forbid parasiting”, I turn it around and ask “how can we make these parasites profitable?” is a special of mine, and has so far been very profitable, in a number of contexts.
If you see none of these, I am happy for you. )

sdr 2 Apr 2017 1:49 UTC
30 points
on: Project Hufflepuff: Planting the Flag
Fellow Hufflepuff / startupper / business getting-stuff-done-er / CFAR / Bay-arean here. Can we talk about the elephant in the room?
- Geeks, MOPs, and sociopaths in subculture evolution ← describes the idea of the role of parasites in subculture evolution; specifically, that once group-surplus achieves a threshold, it is immediately soaked up by parasites funneling it to agendas of their own
- There are, by my count, at least 3 such parasites in the Bay community; and specifically they position themselves as the broken stair step right at onboarding, making the community feel “impenetrable and unwelcoming”. The way how this happens operationally, is when I admit to some level of operational surplus (language skills, software development, business building), from these specific persons I get immediately asks of “Would you like to do free translation for me?” / “Would you like to build $website-idea$ for me?” / “Would you like to donate to $my-cause$?”. I also notice that they don’t do it this overtly to long-term members.
- Note, the problem here isn’t the ask. We do asks in entrepreneur-topia all the time. The problem is the lack of dealcraft: the asks are asymmetrically favouring the asker, and only offer vague lipservice-waving-towards-nice-things as return.
- Presence of these parasites, and lack of dealcraft by these people reached equilibrium at having ‘a strong culture of “make sure your own needs are met”, that specifically pushes back against broader societal norms that pressure people to conform.’ , because people who have been valuepumped hard enough can not sustain themselves in the Bay.
You are attempting to increase the group-surplus of the community. This is very cool. My pre-mortem says, that any such surplus created by the sweat of your brow will be soaked up by this parasitic behavior, and hence fail to achieve long-term changes in admitted competence of the community.

There might be several ways to work around this problem. I want to be upfront about the evaluation criterias for it:
- not talking, or taking action about this problem will not make it go away;
- parasites’ aim is value-pumping: that is, closing deals in which they get the maximum amount of value with the least amount of work on their own;
- parasites participate in the culture like everyone else; for this reason, any plan you might come up with must be reflection-complete: that is, it needs to work, even if everyone in the community knows that such plan is in motion.
A few candidate solutions which sticks out:
- Level up dealcraft: cultivate, and enforce a culture of mutually beneficial asks.
- Level up quantitiy of dealcraft: elicit members -all members- of their goals / objectives / needs, and focus on coincidence of wants. There’s a pretty cool model of this in the book Wishcraft: “Barn-raising”
- Systematically post-mortem newbs, elicit list of parasites (“was there someone who made you uncomfortable, and describe the exact specificities of the situation”), and systematically intervene in the onboarding process.
Edit note: originally, this post used the word “sociopath”, incorrectly -thanks for Viliam’s comment below for pointing it out- fixed.
What links here?
- The Craft & The Community—A Post-Mortem & Resurrection by Bendini (2 Nov 2017 3:45 UTC; 76 points)

sdr 14 Mar 2017 16:34 UTC
0 points
on: [stub] 100-Word Unpolished Insights Thread (3/10-???)
FAI value alignment research, and cryonics are mutually inconsistent stances. Cryo resurrection will almost definitely happen by scanning & whole-brain-emulation. An EM/upload with a subjective timeline sped up to 1000x will be indistinguishable from an UFAI. Incremental value alignment results of today will be applied to your EM tomorrow.

For example, how would you feel with all your brilliant intellect, with all your inner motivational spark being looped into a rat race against 10000 copies of yours, performing work for & grounded to a baseline, where if you don’t win against your own selves, all your current thoughts, and feeling, and emotions are to be permanently destroyed?

sdr 4 Mar 2017 10:04 UTC
1 point
in reply to: Bound_up’s comment on: 5 Project Hufflepuff Suggestions for the Rationality Community
Very yes. Specifically, a bi-weekly, or monthly thread (similar to the open threads currently) of eg “Pitch your idea”, with the hard-constrain for the topmost comments being 100 words at most at any given time, with optional links leading down the rabbit hole.

Edit: bonus point, but not hard requirement describing your idea in language of “up goer five” to avoid that thing where people compress by using technical words, as opposed to compressing comprehensibly. Like, what we want to achieve here is to serve as a common onboarding point for new people to get introduced to those ideas; as opposed to communicating the Theory of Everything in greek symbols.

sdr 3 Mar 2017 9:11 UTC
3 points
in reply to: Osho’s comment on: March 2017 Media Thread
Sure can: subscribe to gwern’s newsletter

sdr 24 Jan 2017 14:09 UTC
1 point
in reply to: gjm’s comment on: Metrics to evaluate a Presidency
Can you point at the part which you find objectionable?

sdr 24 Jan 2017 13:57 UTC
4 points
on: Metrics to evaluate a Presidency
note: topic text was originally different, and included a recently-elected president’s name; which would’ve ranked on google for related-keywords. Below is unedited comment, asking for that name not to be included

Since “Downvoting temporarily disabled”, I would like to express a very, very strong disapproval of this topic being discussed on lesswrong. Rationale:

1, Politics is the mindkiller

2, It attracts the sort of people who would like to discuss these sorts of things, at the expense of those (including myself), who do not; specifically, by ranking for relevant keywords on Google (with lesswrong’s reputation)

3, There exists almost the entirety of the rest of the Internet to discuss these issues, including rationality-related groups, forums, and mailing lists

4, For a specific case study, we just had a CFAR-alumni discussion group blown up by a topic similar to this, which got 100s++ replies, with no measurable convergence; which strongly implies that no, actually, we do not have a collective-intelligence / social tooling to tackle these issues yet.

For this reason, I -along with all upvoters of this comment- would like to ask for an Admin intervention, specifically by either deleting this post, or modifying it to “Metrics to evaluate a president”; that is, talking about general evaluation criterias, instead of ranking for T-related keywords.

sdr 28 Nov 2016 6:25 UTC
26 points
in reply to: Error’s comment on: On the importance of Less Wrong, or another single conversational locus
Speaking as a writer for different communities, there are 2 problems with this:
- Duplicate content: unless explicitly canonized via headers, Google is ambiguous about which version should rank for keywords. This hits small & upcoming authors like a ton of bricks, because by default, the LW version is going to get ranked (on basis of authority), and their own content will be marked both as a duplicate, and as spam, and their domain deranked as a result.
- “An audience of your own”: if a reasonable reader can reasonably assume, that “all good content will also be cross-posted to LW anyways”, that strongly eliminates the reason why one should have the small blogger in their RSS reader / checking once a day in the first place.
The HN “link aggregator” model works, because by directly linking to a thing, you will bump their ranking; if it ranks up to the main page, it drives an audience there, who can be captured (via RSS, or newsletters); and therefore have limited downside of participation.