Boston-based, NAO Co-Lead, GWWC board member, parent, musician. Switched from earning to give to direct work in pandemic mitigation. Married to Julia Wise. Speaking for myself unless I say otherwise. Full list of EA posts: jefftk.com/ânews/âea
Jeff Kaufman đ¸
Thatâs a box fan CR box
A filter with PC fans isnât a âCR boxââa Corsi-Rosenthal box is specifically a design based on box fans. For example, Wikipedia, US Davis, Clean Air Crew, and the Corsi-Rosenthal Foundation mention only box fans.
Good point! Iâm used to houses with the older systems (steam, forced hot water) that are common in the Northeast and wasnât thinking about this effect.
Efficiency in terms of fraction of removed particles wouldnât decrease, but because CFM will decrease efficiency in terms of CADR will too.
Ceiling fans are generally reversible, so you could have the ceiling fan blowing air up and mount the filters for airflow in the opposite direction. I havenât tested this configuration though!
There is actually this study from the national labs that indicate filters hardly need replacement at all. I think it is even worse than razorblades that grow dullâfilters last forever!
That study is looking at nuclear facilities, but Iâm not sure it generalizes to environments with more particulates in the air. In a dusty enough environment your filter will surely get clogged up to the point where youâre not able to move much air through it!
Sure!
I do have my question about your design (and GPT o4-mini-high is pretty skeptical), wondering whether the fan really draws all the air through the filters.
When I tested the prototype it worked well: https://ââwww.jefftk.com/ââp/ââceiling-air-purifier
Now that I know more about how these fans move air, I think it would work even better if the filters extended slightly lower.
BetÂter Air Purifiers
Sure!
We all really appreciate the work youâve done for the Forum over these seven years!
Let me show you what the Forum 1.0 looked like seven years ago. ⌠It didnât work on mobile
And now time for quibbles! The mobile implementation was a bit of a hack, and there were occasional elements that were way the wrong size, but it did work! (#65, initial effort). You can play with the archive; hereâs the old Forum on 2017-02-12.
The new one is of course way better ;)
@Julia_Wiseđ¸ gets credit for this! My draft had two paragraphs for âSetting aside that some people donât have the economic breathing room to make this kind of tradeoffâ.
PriÂoriÂtizÂing Work
did you work on side projects, do some sort of fellowship, or jump straight to applying for jobs in biosecurity?
I did a lot of reading, and a side project around air filtration (1, 2, 3, 4, 5, 6, 7) that I donât think was all that helpful, but mostly I talked to people in the field about what was missing. I think it helped a lot that I was a bit of a known quantity: Iâd been writing publicly for a long time which let people have a sense of what to expect from me.
why/âhow did you choose the Nucleic Acid Observatory?
Based on my reading and talking with people in biosecurity I thought the NAO was aiming to solve a really important problem and it had a lot of good people but the group as a whole was too academic: not enough experience building things, learning by doing, or moving quickly. This seemed like a project where my skills were very complementary, and I think that did end up being the case.
interested in pivoting to either biosecurity or cybersecurity
Great to hear! I donât have a good sense of what would make sense. The NAO isnât currently hiring, but at some point itâs possible weâll be looking for engineers and for candidates who were sufficiently strong elsewhere not having a bio background wouldnât be a blocker.
I do think working on side projects is often pretty good for getting a sense of whether you like the work and other people seeing what you can do. And with LLMs itâs easier than ever to get spun up in new domains.
I donât really have good advice on how to get into the field, though; sorry!
The social norms of EA or at least the EA Forum are different today than they were ten years ago. Ten years ago, if you said you only care about people who are either alive today or who will be born in the next 100 years, and you donât think much about AGI because global poverty seems a lot more important, then you would be fully qualified to be the president of a university EA group, get a job at a meta-EA organization, or represent the views of the EA movement to a public audience.
This isnât just a social thing, itâs also response to a lot of changes in AI timelines over the past ten years. Back then a lot of us had views like âmost experts think powerful AI is far off, Iâm not going to sink a bunch of time into how it might affect my various options for doing goodâ, but as expert views have shifted that makes less sense. While âdonât think much about AGI because global poverty seems a lot more importantâ is still a reasonable position to hold (ex: people who think we canât productively influence how AI goes and so we should focus on doing as much good as we can in areas we can affect), I think it requires a good bit more reasoning and thought than it did ten years ago.
(On the other hand, I see âonly care about people who are either alive today or who will be born in the next 100 yearsâ as still within the range of common EA views (ex).)
SpendÂing on Ourselves
I donât think impartiality to the extent of not caring more about the people one loves is a core value for very many EAs? Yes, itâs pretty central to EA that most people are excessively partial, but I donât recall ever seeing someone advocate full impartiality.
ďNuÂcleic Acid ObÂserÂvaÂtory UpÂdates, April 2025
I think a week from today is a good amount of review time, but I think the key thing is to be clearer in setting expectations in the future.
(Iâd also recommend, as I wrote above, apologizing to Sinergia for misunderstanding their earlier request and giving them a very short time to review.)
In the future the most reliable way to redact a screenshot is:
Open the screenshot and add the black bars
Save the edited screenshot as PNG
Since PNG doesnât support layers or history, the redacted information is reliably no longer present.
Interesting! I hadnât realized people had started using the term this way!