khafra

Karma: 3,777

khafra Jul 24, 2024, 7:30 AM
2 points
0
in reply to: mishka’s comment on: On the CrowdStrike Incident
It would specifically be impossible to prove the Crowdstrike driver safe because, by necessity, it regularly loads new data provided by Crowdstrike threat intelligence, and changes its behavior based on those updates.
Even if you could get the CS driver to refuse to load new updates without proving certain attributes of those updates, you would also need some kind of assurance of the characteristics of every other necessary part of the Windows OS, in every future update.

khafra Jul 22, 2024, 6:26 AM
4 points
0
in reply to: Gunnar_Zarncke’s comment on: Why Georgism Lost Its Popularity
I feel like it’s still Moloch to blame, if a sufficient bribe to landowners would cost less than indefinitely continued rent-seeking.
I don’t have any calculations to offer in support; but I would generally expect an individual landowner’s time preference to be lower than society’s as a whole, so I suspect this is indeed the case.
So the actual reason is that landowners don’t want to be seen taking a bribe, because that would involve acknowledging they have been knowingly rent-seeking since 1879; and the government doesn’t want to openly bribe them for moral hazard whatever; so even though everyone would be better off by their own lights it can’t happen. And that’s fairly moloch-flavored.

khafra May 8, 2023, 10:15 PM
12 points
on: Open Thread—Jan 2022 [Vote Experiment!]
Twitter has announced a new policy of deleting accounts which have had no activity for a few years. I used the Wayback Machine to archive Grognor’s primary twitter account here. Hal Finney’s wife is keeping his account alive.
I do not know who else may have died, or cryo-suspended, over the years of LW; nor how long the window of action is to preserve the accounts.

khafra Apr 19, 2023, 12:22 PM
2 points
in reply to: devansh’s comment on: [Book Review] Why Greatness Cannot Be Planned: The Myth of the Objective
Or A*, which is a much more computationally efficient and deterministic way to minimize the distance to finish the maze, if you have an appropriate heuristic. I don’t have an argument for it, but I feel like finding a good heuristic and leveraging it probably works very well as a generalizable strategy.

khafra Apr 5, 2022, 11:11 AM
5 points
on: What an actually pessimistic containment strategy looks like
Iran is an agent, with a constrained amount of critical resources like nuclear engineers, centrifuges, etc.
AI development is a robust, agent-agnostic process that has an unlimited number of researchers working in adjacent areas who could easily cross-train to fill a deficit, an unlimited number of labs which would hire researchers from DeepMind and OpenAI if they closed, and an unlimited amount of GPUs to apply to the problem.
Probably efforts at getting the second-tier AI labs to take safety more seriously, in order to give the top tier more slack, will move back AI timelines a little? But most of the activities that my brain labels “nonviolent resistance” are the type that will be counterproductive unless there’s already a large social movement behind them.

khafra Jan 19, 2022, 12:14 PM
3 points
on: “Can you keep this confidential? How do you know?”
For personal communications, meta-conversations seem fine.
If you’re setting up an organization, though, you should consider adopting some existing, time-tested system for maintaining secrets. For example, you could classify secrets into categories—those which would cause exceptionally grave harm to the secret’s originator’s values (call this category, say, “TS”); those which would cause serious harm (“S”), and those which would cause some noticeable harm (“C”). Set down appropriate rules for the handling of each type of secret—for example, you might not even write down the TS ones unless you had a very secure safe to store them in, or verbally discuss them outside of protected meeting rooms; and you might not do anything with the S secrets on an internet-connected computer. Anything above C might require a written chain of custody, with people taking responsibility for both the creation and destruction of any recorded form of the information.
You would then have to watch for mutual information in your communications, and see that no combination of the information that you publicly released could cause a large update toward one of the secrets you were keeping. You’d also want to think of some general steps to take after an unplanned disclosure of each type of secret.
It may not sound like the most efficient way to do things, but there’s some pretty high Chesterton’s Fences around this kind of policy.

khafra Dec 2, 2021, 1:14 PM
4 points
in reply to: Adele Lopez’s comment on: Biology-Inspired AGI Timelines: The Trick That Never Works
The answer I came up with, before reading, is that the proper maxent distribution obviously isn’t uniform over every planck interval from here until protons decay; it’s also obviously not a gaussian with a midpoint halfway to when protons decay. But the next obvious answer is a truncated normal distribution. And that is not a thought conducive to sleeping well.

khafra Nov 25, 2021, 2:19 PM
2 points
on: Open & Welcome Thread November 2021
I’ve used Eliezer’s prayer to good effect, but it’s a bit short. And I have considered The Sons of Martha, but it’s a bit long.
Has anyone, in their rationalist readings, found something that would work as a Thanksgiving invocation of a just-right length?

khafra Nov 18, 2021, 6:39 PM
6 points
in reply to: philh’s comment on: Ngo and Yudkowsky on alignment difficulty
Robin Hanson said, with Eliezer eventually concurring, that “bets like this will just recover interest rates, which give the exchange rate between resources on one date and resources on another date.”
E.g., it’s not impossible to bet money on the end of the world, but it’s impossible to do it in a way substantially different from taking a loan.

khafra Nov 25, 2020, 9:41 PM
8 points
on: Open & Welcome Thread – November 2020
I built a thing.
UVC lamps deactivate viruses in the air, but harm skin, eyes, and DNA. So I made a short duct out of cardboard, with a 60W UVC corn bulb in a recessed compartment, and put a fan in it.
I plan to run it whenever someone other than my wife and I visits my house.
https://imgur.com/a/QrtAaUz

khafra May 27, 2020, 2:41 PM
2 points
on: [Repost] Movable Housing for Scalable Cities
Note that Mortal Engines—that steampunk movie with the mobile, carnivorous cities—was released halfway between the original publishing of this essay and today.
Given the difficulties people have mentioned with moving high-density housing between and through cities, maybe we need small cities on SMTs ?

khafra May 1, 2020, 2:31 PM
0 points
on: Interview with Aubrey de Grey, chief science officer of the SENS Research Foundation
These were some great questions. I doubt a few of the answers, however. For example:
My estimate of how far off LEV is with 50% probability started out at 25 years 15 or so years ago, and is now 17 years, so let’s use round numbers and say 20 years. Those estimates have always been explicitly “post-money”, though—in other words, when I say the money would make 10 years of difference, I mean that without the money, it would be 30 years. I think $1B is enough to remove that factor of 2-3 that you mentioned in the previous question, i.e. to take it down to around 1, because it would add a digit to our budget for 20 years. That factor is already coming down, and I expect that it will continue to do so as further progress is made at the bench, which is why I average the benefit out to a factor of 1.5 (i.e. ³⁰⁄₂₀).
Aubrey de Grey admits to drinking four pints of beer a day, and I believe his total ethanol consumption is much higher (via evidence which is strong to me, but not to you). He’s 57, and looks older than many in their 70s. The evidence may be ambiguous on the longevity effects of <2 drinks per day, but it’s quite clear on 4 or over.
This doesn’t seem like the behavior of someone who truly believes, in the sense of constraining his expected experiences, that his remaining expected lifespan is almost exactly the time to LEV. I don’t know what the real timeline to LEV is, but Dr. de Grey acts like he believes it’s well over 30 years.

khafra Apr 7, 2020, 12:48 PM
2 points
in reply to: Dmitry Vaintrob’s comment on: Research on repurposing filter products for masks?
The 100% efficacy for a middle filter layer that’s had a saltwater + surfactant sprayed onto it sounds really good; but I wonder how tight the filter material has to be, for that level of efficacy. I also wonder how much air resistance the salt coat adds.
A HEPA filter + carbon would be less restrictive if the carbon part were salted than if the HEPA filter itself were salted, but that might not deactivate all of the virus.

khafra Apr 6, 2020, 6:47 PM
2 points
in reply to: jmh’s comment on: What is the impact of varying infectious dose of COVID-19?
If virus exposure mid-illness worsens your symptoms, doesn’t that mean being indoors is harmful? it would be far healthier to spend as much time outdoors as possible? Perhaps on a net hammock if you have to lie down, so your face isn’t lying on a cloth full of the virus you’re exhaling? Surely this effect would be so large that clinical studies would have noticed by now, people recovering much faster when they’re not in a hospital room, or in a room at all.
On a gears-level, it seems like illness severity would be heavily dose-dependent until the virus replication rate has outpaced the amount you could reasonably inhale.
If so, if you have a specific event that you’re concerned may have exposed you, it might be worthwhile to sleep outside for a few nights, weather permitting.

khafra Apr 3, 2020, 3:01 PM
2 points
in reply to: Roko Jelavić’s comment on: The horse-sized duck: a theory of innovation, individuals and society
How many dimensions is inference space? How many duck-sized horses do we need, to have a ²⁄₃ chance of taking those steps? And are they being modeled as duck-sized monkeys with typewriters, or are they closer to a proper mini-Einstein, who is likely to go the correct direction?

khafra Mar 30, 2020, 11:38 AM
4 points
in reply to: James_Miller’s comment on: Will grocery stores thwart social distancing, and when should I eat my food stockpile?
I live in a hot region, and have a car parked outside. I’ve been putting non-heat-sensitive packages in there for a day, since interior temperatures should be going above 130F / 55C, and easily killing any viruses.

khafra Mar 27, 2020, 12:41 PM
2 points
on: March 25: Daily Coronavirus Updates
Disinfection guidelines are 70C for 30 minutes. I’ve read elsewhere that 27C deactivates the virus, but never seen that claim attached to logs per hour. Has anybody seen quantitative data on covid survival rates in human-survivable temperatures at various humidities?

edit: found some stuff for the last SARS: if you go to 100F / 48C *and* 95+% humidity, you will kill 2 log10 in 24 hours. If you lose humidity *or* temperature, you’re back to the baseline of 1 to 0 logs in 24h.
https://www.hindawi.com/journals/av/2011/734690/

khafra Mar 26, 2020, 2:42 PM
2 points
on: Bayesian Inference with Partially Specified Priors
Is the described process different from Dempster-Shafer ?

khafra Mar 26, 2020, 2:19 PM
3 points
on: Do 24-hour hand sanitizers actually work?
For the object-level question, Wei Dai linked to this study showing benzalkonium chloride (and a few related chemicals) ineffective against enveloped human coronavirus (although this was one of the common cold variants).

khafra Mar 25, 2020, 12:07 PM
6 points
on: Adding Up To Normality
This is good, but I’d add a caveat: it works best in a situation where “normal” is obviously not catastrophic. The airplane example is central to this category. However lift works, air travel is the safest method of getting from one continent to another ever devised by humanity. If you take DMT and finally become aware of the machine elves supporting the weight of each wing, you should congratulate them on their diligence and work ethic.
The second example, morality under MWI, veers closer to the edge of “normal is obviously not catastrophic.” MWI says you’re causally disconnected from other branches. If your good and bad actions had morally equivalent effects, you would not anticipate different observations than you would under “normality.”
As lincolnquirk pointed out, Covid and other long tail events are diametrically opposed to the “normal is obviously not catastrophic” category. Instead of the object-level belief being changed by a discussion on aerodynamic theory, it’s being changed by the plane suddenly falling out of the sky, in a way that’s incompatible with our previous model.
So, I’d tweak your adage: “promise yourself to keep steering the plane mostly as normal while you think about lift, as long as you’re in the reference class of events where steering the plane mostly as normal is the correct action.”
What links here?
- Adding Up To Normality by orthonormal (Mar 24, 2020, 9:53 PM; 85 points)