Posts

Journal 'Basic and Applied Psychology' bans p<0.05 and 95% confidence intervals 2015-02-25T17:15:09.371Z · score: 12 (13 votes)
0.5% of amazon purchases to a charity of your choice (opt-in) 2014-04-02T01:55:35.707Z · score: 7 (8 votes)
Does model theory [psychology] predict anything? (book: "How We Reason" (2009)) 2013-06-03T03:11:39.898Z · score: 2 (3 votes)
"disfluency" research 2013-05-30T21:33:12.398Z · score: 10 (15 votes)
does imagining +singularity cause depression? 2013-05-30T16:51:10.177Z · score: -11 (16 votes)
Huy Price (Cambridge philosopher) writes about existential risk for NYT 2013-01-30T02:06:32.404Z · score: -4 (11 votes)
central planning is intractable (polynomial, but n is large) 2012-05-31T18:41:09.552Z · score: 9 (13 votes)
SMBC comic: poorly programmed average-utility-maximizing AI 2012-04-06T07:18:10.770Z · score: 9 (18 votes)
DAGGRE group forecasting workshop 2012-03-07T23:26:00.358Z · score: 3 (6 votes)
Some conditional independence (Bayes Network) exercises from ai-class.com 2011-10-20T05:09:59.345Z · score: 9 (10 votes)
Greg Linster on the beauty of death 2011-10-20T04:47:24.711Z · score: 6 (10 votes)
Contrarians judged mad after being proven right (John Hempton) 2011-09-02T06:41:12.500Z · score: 5 (7 votes)
synapse renormalization - another reason to sleep more than minimum-REM 2011-08-21T19:44:37.093Z · score: 6 (9 votes)
(US only) Donate $2 to charity (bing rewards) 2011-08-18T21:16:48.694Z · score: -1 (4 votes)
SMBC: dystopian objective function 2011-06-24T04:03:16.806Z · score: 8 (8 votes)
1-2pm is for ??? 2011-06-16T05:01:04.539Z · score: -5 (16 votes)
Friendlier AI through politics 2009-08-16T21:29:56.353Z · score: 1 (12 votes)
She Blinded Me With Science 2009-08-04T19:10:49.712Z · score: 13 (13 votes)

Comments

Comment by jonathan_graehl on How strong is the evidence for hydroxychloroquine? · 2020-04-05T21:49:03.274Z · score: 4 (2 votes) · LW · GW

[I know you didn't advocate this, just saying:] If we had the option to wait and obtain rigorous proof, we would prefer that. We don't have that option. Concurrent with obtaining more certain information, it should be used (and is being used) off-label at safe doses in combination with azithromycin to ward off secondary infections.

Comment by jonathan_graehl on What is the impact of varying initial viral load of COVID-19? · 2020-04-05T21:45:06.879Z · score: 3 (2 votes) · LW · GW

Your 'Zvi' doesn't make sense: if we know it works, everyone should get a low dose ASAP. But we don't know it works. I favor Hanson's approach to discover how well it works, and then reevaluate.

Comment by jonathan_graehl on Zoom's security is not that bad · 2020-04-05T21:30:52.818Z · score: 4 (2 votes) · LW · GW

Are you aware that Chinese nationals worldwide are often asked to collect intel or perform ops for CCP? Do you think the disproportionate stories of industrial espionage are just disproportionate reporting? Are you aware that CCP requires its citizen companies to routinely violate users' privacy?

Why does it make you angry that xenophobic tendencies contribute to skepticism of reliance on Chinese software/servers? How is that at all relevant to a rational assessment?

Comment by jonathan_graehl on Iceland's COVID-19 random sampling results: C19 similar to Influenza · 2020-03-28T20:06:07.053Z · score: 14 (5 votes) · LW · GW

That Iceland's currently 1% infected as of now doesn't say anything about how infected it would be after a few weeks of no-special-controls measures (comparable to folks' behavior in a regular flu season). This is the beginning of this virus's worldwide course. It's dishonest to compare a snapshot now with the accumulated total of a whole flu sason.

Comment by jonathan_graehl on Iceland's COVID-19 random sampling results: C19 similar to Influenza · 2020-03-28T20:02:26.153Z · score: 4 (2 votes) · LW · GW

If it's very contagious (it is), the damage could easily become 50x current. It's true that as and if we learn outcomes per infection are not as bad as feared, we will relax. While we should be skeptical of hype, we need to act aggressively early on until we know more about how to treat and how important it is to slow or limit the spread.

Comment by jonathan_graehl on Iceland's COVID-19 random sampling results: C19 similar to Influenza · 2020-03-28T19:58:05.301Z · score: 9 (5 votes) · LW · GW

How many times more contagious (if uncontrolled) and critical/fatal (without hospital overcrowding) is it than a typical flu?

Diamond Princess indicates at *least* 2x on both counts IMO. I think it's a bit shady to say that 2x is ' well within the range of uncertainty ' as if that means something.

I hope it's only 2x worse; I believe 5x on contagion and 3x on severity pre-overcrowding.

Comment by jonathan_graehl on Should we all be more hygenic in normal times? · 2020-03-17T07:34:56.089Z · score: 4 (2 votes) · LW · GW

When can we be assured a pandemic isn't looming?

Comment by jonathan_graehl on Vaccine... Help? Deprogramming? Something? · 2019-12-27T23:45:46.756Z · score: 5 (3 votes) · LW · GW

Vaccines sometimes kill people. Several serious diseases that killed many more people, we're told, are a much smaller risk now. At some point, you'd think people would want to selfishly avoid vaccinating so much. And that's what we see happening. There's a lot of rationalization going on.


Comment by jonathan_graehl on Matthew Walker's "Why We Sleep" Is Riddled with Scientific and Factual Errors · 2019-11-16T21:33:49.106Z · score: -1 (6 votes) · LW · GW

Author advocates biphasic sleep totaling 6.3hr/day.

Author claims http://jcsm.aasm.org/ViewAbstract.aspx?pid=31409 doesn't support 'smaller testes size for men who sleep less'. It does.

Still, I appreciate that he helped me repair some erroneous beliefs I picked up from Walker's book.

Comment by jonathan_graehl on Matthew Walker's "Why We Sleep" Is Riddled with Scientific and Factual Errors · 2019-11-16T21:32:35.228Z · score: 11 (15 votes) · LW · GW

Walker probably feels justified in exaggerating or is simply biased in his investment of accuracy-checking effort) akin to climate scientists - feeling the public won't update its policy as close to the optimal setting unless they lie.

Comment by jonathan_graehl on Whence decision exhaustion? · 2019-06-28T23:00:19.682Z · score: 11 (4 votes) · LW · GW

Feeling of powerlessness - tiring.

Fear of suffering if you choose wrong (plus bonus regret), inability to gather info enough to ensure outcome - tiring.

Feeling of scarcity - that you can't just lock in some "if it turns out I'm wrong, I win anyway, I have this insurance ..." security, but instead must really bear an emotionally crushing loss - tiring.

Is an omnipotent omniscient tired out by decisions? No.

Comment by jonathan_graehl on Micro feedback loops and learning · 2019-05-28T22:41:31.024Z · score: 2 (1 votes) · LW · GW

It's not a good book, except it says: focus/visualize on the result you want, not on verbal coaching cues. That can be good advice, but I object: when there's time+space to consciously plan some part of your gameplay, verbal (self-)coaching is fine.

Comment by jonathan_graehl on What are reliable ways to make a statement in such a way that I will be able to prove in the future that I had made that statement? · 2019-04-24T18:53:48.590Z · score: 6 (4 votes) · LW · GW

Proving you made the statement at a given time is as simple as getting a trusted signed timestamp or inserting it in some blockchain-like ledger, but that's not even close to making yourself accountable for predictive accuracy.

Be sure to publish *all* your predictions so we don't get file-drawered (except on you as a person which we probably can't help).

A common technique is to publish a secure hash of your prediction rather than the text (in case you want to avoid it being self-fulfilled or anti-fulfilled or otherwise traded on) (crypto signed w/ your identity, too).

But if we don't see a stream of plaintext reveals and a means of identifying all such hashes you've published, we might suspect you of planting both positive and negative predictions.

Most people prefer to publish their prediction (+reasoning) clear-text because they want to persuade and they want credit for being smart before the verdict is in.

Comment by jonathan_graehl on What are effective strategies for mitigating the impact of acute sleep deprivation on cognition? · 2019-03-31T21:48:43.248Z · score: 4 (4 votes) · LW · GW

Sleep deprivation is cumulative over the span of weeks. Being short 30 min each day for two weeks is disastrous. Almost no one is near top capability with even 7 hours. Memory formation and recall are especially limited on low sleep, even if you drug to overcome the lapses in attention. Physical health is also severely harmed (look at how attractive someone is after 'beauty sleep' vs deprived), perhaps mostly via poor diet choices but honestly why would evolution not layer on physical garbage collection processes when mental ones are already underway ... further, a bunch of micro naps doesn't give you the same concentration of deep sleep as the last 3 hours of an 8.5 hour bout would (8.5 is my ideal; 8 is tolerable).

Comment by jonathan_graehl on Population Aging as an Impediment to Addressing Global Catastrophic Risks · 2018-10-26T16:57:06.410Z · score: 2 (1 votes) · LW · GW

Similarly, people who don't have children and don't realistically hope for extreme longevity have a counterproductive voice in politics. I'd trust intelligent+loving (high investment) grandparents to invest wisely in the future before I'd trust an environmentalist (by way of illustrating the progeny-dependent long-term 'stakes in the game' criteria, not to derail us into boring political territory; similarly, uninvolved sires don't get any credibility).

Comment by jonathan_graehl on Harper's Magazine article on LW/MIRI/CFAR and Ethereum · 2014-12-29T01:44:54.898Z · score: 1 (1 votes) · LW · GW

https://www.youtube.com/watch?v=W5UAOK1bk74 - shoot, Vassar does really wear slightly-too-large suits. I'll assume that he's A/B tested this to give best results?

Comment by jonathan_graehl on Harper's Magazine article on LW/MIRI/CFAR and Ethereum · 2014-12-29T01:34:56.019Z · score: 0 (0 votes) · LW · GW

You can (or could) watch EY debating (e.g. w/ that presumptuous jaron lanier guy) over videoconference and like many less-polished speakers he has some visible tics while searching for a thought or turn of phrase while feeling under the gun + not wanting to lose his turn to speak.

Comment by jonathan_graehl on Harper's Magazine article on LW/MIRI/CFAR and Ethereum · 2014-12-29T01:29:58.003Z · score: 1 (1 votes) · LW · GW

I liked the excerpts gwern quoted and see truth (and positive things) in most of it. "Hydra-headed" for EY's writing seems inapt. If you refute one of his essays 3 more will spring up in response?

Not sure what Vassar thinks is 3 in 1000 people - exploring+building boldly? Leadership?

Almost running a red light while buzzed+chatting. Hm. Well, I'm sure we all try to have a healthy respect for the dangers of killing and being killed while driving cars.

Comment by jonathan_graehl on Too good to be true · 2014-07-11T22:54:57.015Z · score: 1 (1 votes) · LW · GW

Yes, that's suspicious. Good instinct. I'm sure there's some bias against publishing a marginally-significant result that's got a low (outside the framework of the paper's statistical model) prior. I'd bet some of the unlucky ones got file-drawered, and others (dishonestly or not) kept on collecting more data until the noise (I presume) was averaged down.

However, you might be missing that on an iso-P contour, false positives have diminishing effect size as sample size increases.

Comment by jonathan_graehl on "Stupid" questions thread · 2014-04-22T21:57:37.022Z · score: 0 (0 votes) · LW · GW

Unambiguous mistake or ambiguous parallel construction? I agree w/ your parse, on grounds of the indisputable goodness of truckloads of money.

Comment by jonathan_graehl on AALWA: Ask any LessWronger anything · 2014-04-01T23:52:27.634Z · score: 1 (1 votes) · LW · GW

Thanks for answering that as if it were a sincere question (it was).

"Maybe this universe has invisible/anthropic/supernatural properties" is a fascinating line of daydreaming that seems a bit time-wasting to me, because I'm not at all confident I'd do anything healthy/useful if I started attempting to experiment. Looking at all the people who are stuck in one conventional religion or another, who (otherwise?) seem every bit as intelligent and emotionally stable as I am, I think, to the extent that you're predisposed to having any mystical experiences, that way is dangerous.

Comment by jonathan_graehl on Lifestyle interventions to increase longevity · 2014-02-28T19:28:47.038Z · score: 1 (1 votes) · LW · GW

I snore when I'm very tired and sleeping on my back (when my jaw relaxes down in that position it's harder to breathe even through nostrils). Any cheap advice for that (besides don't do it)?

Are there harmless allergy meds that would be worth taking for better sleep when I have mild nasal congestion from seasonal pollen etc?

Comment by jonathan_graehl on Lifestyle interventions to increase longevity · 2014-02-28T19:24:49.429Z · score: 0 (0 votes) · LW · GW

Fair point, but how long does it take to eat+digest (cooked or uncooked) 100 calories of spinach compared to 100 calories of whole milk? How much does it cost? Etc.

I agree that you shouldn't count the vitamin-fortification of milk as part of the value unless it turns out that milk is an especially good transport for what's added to it.

Comment by jonathan_graehl on Lifestyle interventions to increase longevity · 2014-02-28T13:40:01.097Z · score: 0 (0 votes) · LW · GW

I don't know anything about testicular cancer, but are self-exams useful for breast cancer? I know that the data argues against mammogram-everyone-annually + the ensuing unnecessary surgeries caused by not-harmful tumors or other false positives - no increase at all in life expectancy and presumably there's a significant psychological (and $) cost.

Comment by jonathan_graehl on Open Thread for February 18-24 2014 · 2014-02-20T23:11:18.293Z · score: 1 (1 votes) · LW · GW

Congratulations.

I think there's some support for the idea of trying to lose weight slowly, without cutting caloric intake too much more than it takes to see some progress (tricky when to see it you have to average over several days, or, for women, a month)

Comment by jonathan_graehl on Open Thread for February 18-24 2014 · 2014-02-20T23:09:13.161Z · score: 1 (1 votes) · LW · GW

You're right to have low confidence in our winning-ness. If we were winning so hard, why would we be so often theorizing about what it takes to win?

Reading and writing well means never having to admit that you didn't do any research before weighing in.

Comment by jonathan_graehl on Amanda Knox Guilty Again · 2014-02-19T23:59:43.856Z · score: 1 (1 votes) · LW · GW

http://www.youtube.com/watch?v=joZV0XARmvA - "Is Amanda Knox Guilty?" - NBC produced docu for BBC.

Summary: Guede might be guilty (I don't know), Knox+Sollecito lied to try to get out of trouble, there's interesting DNA evidence with severe technical problems:

  1. a lot of DNA from Solecito (likely making him the one who handled it) on a bra clasp that wasn't collected from the scene until 46 days later. Unfortunately by then there was plenty of incentive by then for authorities to falsify evidence to bolster their extremely weak case.

  2. A tiny trace of Kercher's DNA on a knife in Solecito's apartment - also collected after incentive to falsify (they cheated by running the test when they weren't supposed to due to insufficient amount of material, at the very least).

  3. faint bloody footprints of the right size in the bathroom. not strong evidence of killing and apparently might not even be blood (they didn't collect any, just have illuminator dye photos which can trigger off bleach too).

The rest seems like comparatively unreliable evidence to me. DNA of you, even drops of blood, in your own bathroom? Big deal. Accusations from the convicted killer (Guede)? People pressured by police lied to try to get out of trouble? No surprise. I believe ~90% that neither Knox nor Sollecito killed her or helped cover the killing. Most concerning to me are the reasons given by the authorities - it's mostly pretty lame ("there must have been 3 attackers! there weren't many defensive wounds! kercher knew karate! not even superman could do that alone. and amanda covered her ears!")

Comment by jonathan_graehl on Amanda Knox Guilty Again · 2014-01-31T18:08:54.949Z · score: 2 (2 votes) · LW · GW

video interviews of Knox - http://www.theguardian.com/world/2014/jan/30/amanda-knox-prison-meredith-kercher-murder

Comment by jonathan_graehl on AALWA: Ask any LessWronger anything · 2014-01-28T16:37:13.979Z · score: 1 (1 votes) · LW · GW

Other than wanting more sex, did you notice your mind changing?

I also wonder if late puberty extends the pre-adult skill learning window (adults supposedly can't learn as much or as well).

Comment by jonathan_graehl on AALWA: Ask any LessWronger anything · 2014-01-27T22:56:43.322Z · score: 0 (0 votes) · LW · GW

What's your favorite amount of testosterone? Why? Would the optimum shift according to purpose?

Comment by jonathan_graehl on AALWA: Ask any LessWronger anything · 2014-01-27T22:50:26.326Z · score: -2 (4 votes) · LW · GW

I think there are god-like things that interact with humans

Crazy people and trolls exist. Some of them are eloquent.

So why do you talk about it at all when it just makes you seem crazy to most of us?

Are you looking for confirmation or agreement in others' hallucinations? Or perhaps you suspect your kind of experiences are more common than openly expressed?

I assume I'd take seriously your crazy experiences if they were mine. Is there anything at all you can say that's of value to someone like me who just hears crazy?

Comment by jonathan_graehl on A critique of effective altruism · 2013-12-02T23:15:26.216Z · score: 1 (1 votes) · LW · GW

from the perspective of 100,000,000 years later it is unlikely that the most critical point in this part of history will have been the distribution of enough malaria nets

I read this as presuming that generating/saving more humans is a worse use of smart/rich people's attention and resources than developing future-good theory+technology (or maybe it's only making more malaria-net-charity-recipients and their descendants that isn't a good investment toward those future-good things, but that's not likely to figure, since we can save quite a few lives at a very favorable ratio).

I wonder if you meant that it's a worse use because we have more people alive now than is optimal for future good, or because we only want more smart people, or something else.

Comment by jonathan_graehl on What Can We Learn About Human Psychology from Christian Apologetics? · 2013-10-22T02:02:48.966Z · score: 2 (2 votes) · LW · GW

'convert open-minded skeptics, and ...' ?

Comment by jonathan_graehl on Book Review: Basic Category Theory for Computer Scientists (MIRI course list) · 2013-09-24T09:18:22.756Z · score: -2 (4 votes) · LW · GW

I read it a few years ago and didn't enjoy it at all except for an early section describing some interesting categories. I'm a great programmer and familiar with functional programming but not much Haskell. I worked examples, followed proofs, and generally understood the material.

I view as cargo-cult any recommendations to read this book for someone who wants to program (AI or otherwise).

Comment by jonathan_graehl on What Bayesianism taught me · 2013-08-12T19:06:48.932Z · score: 2 (2 votes) · LW · GW

We should unpack "banish talk of X" to mean that we should avoid assessments/analysis that would naturally be expressed in such surface terms.

Since most of us don't do deep thinking unless we use some notation or words, "banish talk of" is a good heuristic for such training, if you can notice yourself (or others can catch you) doing it.

Comment by jonathan_graehl on Seed Study: Polyphasic Sleep in Ten Steps · 2013-07-27T21:15:02.322Z · score: 0 (0 votes) · LW · GW

Yes, the comments by "a reader" and Michael Turner helped me situate Wozniak's (supermemo polyphasic skeptic) point of view. Thanks for the link.

Comment by jonathan_graehl on Why Eat Less Meat? · 2013-07-27T04:32:07.411Z · score: 0 (0 votes) · LW · GW

I have genuine uncertainty as to the nature of farmed chicken suffering - enough that I'd say it's bad to create your average meat-farmed chicken - otherwise I'd be right there with you at 10^20 or something similarly ridiculous.

The suggestion to genetically engineer suffering-knockout chicken seems a good one (though I'd have some residual uncertainty even then).

Comment by jonathan_graehl on Why Eat Less Meat? · 2013-07-27T04:27:45.425Z · score: 0 (0 votes) · LW · GW

Yes, I took 1.5yr from another comment, which which I guess might be for egg layers or the natural lifespan. I really should have specified lifespan in the poll.

Comment by jonathan_graehl on Why Eat Less Meat? · 2013-07-26T05:06:16.880Z · score: 0 (0 votes) · LW · GW

Median answer - of 100 factory chickens (so 150 chicken-suffering-years) : 1 human QALY - impresses me.

Quite a few people take animal suffering pretty seriously. It must feel odd to have society's rules so far removed from that - like serious abortion-is-murder believers.

Comment by jonathan_graehl on Why Eat Less Meat? · 2013-07-26T04:59:14.218Z · score: 1 (1 votes) · LW · GW

Ok - I didn't see your "Note" at first. I'm not sure what you mean. Presumably your answer would be indifferent or yes, though. Otherwise, could you explain?

Comment by jonathan_graehl on Why Eat Less Meat? · 2013-07-26T04:57:14.206Z · score: 3 (3 votes) · LW · GW

I was thinking of the average human. So 1 part you, 20 parts friend, 10 parts family, 50 parts colleague, 6 billion parts stranger. Of course it shouldn't matter, since I said economic constraints don't apply. Assume everyone gets a QALY and 6 billion times your answer in chickens are farmed.

Comment by jonathan_graehl on Why Eat Less Meat? · 2013-07-25T07:43:31.674Z · score: 1 (1 votes) · LW · GW

Sure, and if the purpose of a group is to reduce animal suffering and voluntary changes in individual consumption patterns are the most effective route, then the likes/shares are presumably accompanied by those people using less farmed animal products.

Comment by jonathan_graehl on Why Eat Less Meat? · 2013-07-25T03:36:56.992Z · score: 1 (1 votes) · LW · GW

I don't know about you, but my explanation for being leery is: what Facebook groups do I expect to encounter? Answer: those that devote a large amount of effort to promoting themselves. (I also expect to encounter Facebook groups that are popular/worthy, but note that the anthropic reason I gave first applies no matter whether the group is actually good). Be skeptical of things that come to your attention through Facebook - at least beware privileging the hypothesis.

I agree that awareness promotion can be good, but another instinct tells me that Facebookers love to conclude that the best thing they can do is share/like/etc. - it's like finding the cheapest way possible to feel like a good person.

Comment by jonathan_graehl on Why Eat Less Meat? · 2013-07-25T03:34:11.570Z · score: 0 (0 votes) · LW · GW

Yes, I realized this as I as making a sandwich and came back to say so :) I'll leave my mistake unedited as a warning to others. -.0001 means what I said but with "prevent the creation" being "create". The sign changes the sign of one of the items in the exchange.

Comment by jonathan_graehl on Why Eat Less Meat? · 2013-07-25T02:36:47.431Z · score: 0 (0 votes) · LW · GW

No, I meant that someone who answers -0.001 would prefer removing 1000 human QALY in order to prevent the creation of a single factory farmed chicken. Though I wouldn't expect any such answers. It's a question about a trade.

Comment by jonathan_graehl on Why Eat Less Meat? · 2013-07-25T02:35:14.837Z · score: 1 (1 votes) · LW · GW

Yeah, if your answer would be infinity, just answer "yes" to the other poll. I noticed this too :)

Comment by jonathan_graehl on Why Eat Less Meat? · 2013-07-25T02:33:34.031Z · score: 1 (1 votes) · LW · GW

Be wary of Facebook groups whose consensus is "it's most important to promote awareness at this stage".

That said, I like the group/concept. It's interesting to ponder, and a welcome counterpart to "reduce farmed animal suffering".

Comment by jonathan_graehl on Why Eat Less Meat? · 2013-07-25T02:17:53.368Z · score: 0 (0 votes) · LW · GW

Sorry about the grammatical ambiguity. "No" means you'd rather the chicken never existed, not that you'd rather the universe never existed. I just mean roughly that you prefer the chicken not exist.

Comment by jonathan_graehl on Why Eat Less Meat? · 2013-07-25T02:16:15.660Z · score: 2 (2 votes) · LW · GW

Do you think factory-farmed-chicken-lives are worth living? That is, if you could create infinitely many of them at no material cost would you do so? Please don't consider the economic value of chickens; suppose this marginal chicken has no practical use whatsoever. Further, it's not an option to create them and then transport them to chicken-rescue pleasure-domes.

[pollid:545]

Comment by jonathan_graehl on Why Eat Less Meat? · 2013-07-25T02:10:38.214Z · score: 1 (1 votes) · LW · GW

Ignoring economic/environmental cost, how many chickens would you create and breed into factory-farming suffering, in exchange for one additional QALY? That is, you wouldn't make the trade unless it took fewer than this number of farmed chickens.

[pollid:544]

(answers may be very small (less than 1) if you value avoiding chicken suffering more than healthy human life-years) or even negative if you'd give up human lives to create more suffering chickens.

(If you think factory-farmed chickens have lives worth creating, please don't answer the poll, as your answer of infinity will throw off the average - you can vote "yes" or "indifferent" to the poll below this instead; this poll is mostly for people who answer "no" to it)

(I don't claim that chickens can actually be traded for human QALY - I still haven't gotten the ritual exactly working yet).