It is odd, isn't it? The effect sizes seem ridiculous*, but there's nothing obviously wrong with that study (aside from the sample size). Cochran has blogged about oxygen before as well. To compile some of the relevant papers:

• Winter et al 1975, "The anesthetic effect of air at atmospheric pressure"
• Scholey et al 1999, "Cognitive performance, hyperoxia, and heart rate following oxygen administration in healthy young adults"
• Kajtar L, Herczeg L, Lang E. 2003. "Examination of influence of CO2 concentration by scientific methods in the laboratory" In: Proceedings of Healthy Buildings 2003, 7–11 December 2003. Singapore:Stallion Press, 176–181.
• Kajtar L, Herczeg L, Lang E, Hrustinszky T, Banhidi L. 2006. "Influence of carbon-dioxide pollutant on human well-being and work intensity" In: Proceedings of Healthy Buildings 2006, 4–8 June 2006, Lisbon, Portugal:Universidade do Porto, 85–90. (requested these two)
• Chung et al 2008, "A study on the effects of 40% oxygen on addition task performance in three levels of difficulty and physiological signals"
• Satish et al 2012, "Is CO2 an Indoor Pollutant? Direct Effects of Low-to-Moderate CO2 Concentrations on Human Decision-Making Performance"
• Law et al 2014, "Relationship Between Carbon Dioxide Levels and Reported Headaches on the International Space Station" (I wonder if the ISS has the same kinds of 'sick building syndrome' that terrestrial buildings do and so the relationship may be confounded? On the other hand, the 'chits' mentioned sound like quasi-experiments.)
• Stafford 2015, "Indoor air quality and academic performance" (Texas natural experiment)

The problem for me is that while it makes sense that since we run on oxygen and the brain uses a lot of oxygen (the whole 'BOLD' thing etc), more oxygen might be better, it has the same issue as Kurzban's blood-glucose/willpower criticism: if the brain needs more oxygen than it's getting, why doesn't one simply breath a little more? While sedentary during these sorts of tasks, you have far more breathing capacity than you should need - you are able to sprint all-out without falling over of asphyxiation, after all. So there's no obvious reason there should be any lack, even more so than for glucose. And shouldn't CO2 levels closely track various aspects of weather? But as far as I know, various attempts to correlate weather and cognitive performance or mood have turned up only tiny effects. In addition, too much oxygen can be bad. So is it too little oxygen or too much nitrogen or too much carbon dioxide...?

Jessica Taylor for lending me a CO2 monitor so that I could see variability in indoor CO2 levels.

What monitor is that? You could try recording CO2 long-term, especially if it's a data logger. Opening windows is something that's easily randomized.

I did some looking and compiling of consumer-oriented devices a while ago: https://forum.quantifiedself.com/t/indoor-air-quality-monitoring-health/799/40 I was not too impressed since nothing hit the sweet spot of accurate CO2 and PPM measurement under $100. The Netatmo looked decent but there are a lot of complaints about accuracy & reliability (checking the most recent Amazon reviews, still a lot of complaints).

I've been thinking maybe I should settle for the Netatmo. I've been working on a structural equation model (SEM) integrating ~100 personal data variables to try to model my productivity (some current sample output), and it would be nice to have even noisy daily C02 variables (as long as I know how noisy and can use it as a latent variable to deal with the measurement error). Correlation-wise, I think backwards causation can be mostly ruled out, and the most obvious confound is weather, which is already in my SEM.

* taken at face value, with reasonable estimates of how much rooms differ from day to day or week to week, CO2 levels would explain a lot or maybe most of variability in IQ tests or cognitive performance!

Small proportional changes seem unlikely to drive big effects, unless there is some feedback mechanism that is keeping the level precisely balanced.

Such as in the body, dealing with tightly regulated and critical aspects of metabolism like oxygen consumption.

But 1% changes in oxygen should be happening all over the place.

Perhaps they are. You don't know the effect because the existing experiments do not vary or hold constant oxygen levels. All you see is the net average effect, without any sort of partitioning among causes.

In the paper I linked, I think they directly add relatively pure carbon dioxide. And the total concentration is 0.1%. So the concentration of oxygen in the air is not really changing.

Can't you apply that argument to oxygen and nitrogen as well? If you are willing to believe that adding a small absolute amount of carbon dioxide can have large effects on the brain, then I don't see why you would not also be willing to believe that decreasing oxygen (a critical fuel for the brain's metabolism) by a small absolute amount might have large effects on the brain. Injecting CO2 as they do does control for air variables like mold and temperature and humidity, but I didn't see anything about also injecting oxygen and nitrogen to independently manipulate the air composition in all 9 possible ways to disentangle which it is. It could be that CO2 is inert, but by pushing out oxygen and reducing oxygen levels has effects; it could be that CO2 is inert but it's both oxygen and nitrogen, or CO2 is poisonous but is combining with lack of oxygen.

The texas natural experiment seems like an especially convincing complement to the more artificial setting, thanks for pointing it out.

I found it interesting that the anti-mold renovations had such large apparent effects compared to the ventilation and other renovations.

The sample size is small, but given the effect size I don't think it even matters that much. The error seems like less than a factor of 2.

Small study effects go beyond just sampling error, so they are untrustworthy.

I don't see how it can be about oxygen.

I believe the human breathing regulation mechanism (how frequently/deeply you breathe) is driven by CO2 concentrations. So manipulating the CO2 affects the breathing which determines how much oxygen is your body getting.

It won't be hard if the effects are as large as claimed in the original study.

The study shows minor effects at 1000 ppm and pronounced effects at 2500 ppm. I don't think changes in weather would drive your CO2 concentration to these levels.

And if you interpret the effect of weather as mostly open vs closed windows, there is a whole bunch of other factors in play like the balance of indoor and outdoor contaminants, etc.

I am sceptical of these results, anyway, they look too big. And the authors mention another study:

Researchers in Hungary have questioned this assumption (Kajtar et al. 2003, 2006). The authors reported that controlled human exposures to CO2 between 2,000 ppm and 5,000 ppm, with ventilation rates unchanged, had subtle adverse impacts on proofreading of text in some trials, but the brief reports in conference proceedings provided limited details.

which implied ("subtle") small effect size.

It is odd, isn't it? The effect sizes seem ridiculous*, but there's nothing obviously wrong with that study (aside from the sample size). Cochran has blogged about oxygen before as well. To compile some of the relevant papers:

• Winter et al 1975, "The anesthetic effect of air at atmospheric pressure"
• Scholey et al 1999, "Cognitive performance, hyperoxia, and heart rate following oxygen administration in healthy young adults"
• Kajtar L, Herczeg L, Lang E. 2003. "Examination of influence of CO2 concentration by scientific methods in the laboratory" In: Proceedings of Healthy Buildings 2003, 7–11 December 2003. Singapore:Stallion Press, 176–181.
• Kajtar L, Herczeg L, Lang E, Hrustinszky T, Banhidi L. 2006. "Influence of carbon-dioxide pollutant on human well-being and work intensity" In: Proceedings of Healthy Buildings 2006, 4–8 June 2006, Lisbon, Portugal:Universidade do Porto, 85–90. (requested these two)
• Chung et al 2008, "A study on the effects of 40% oxygen on addition task performance in three levels of difficulty and physiological signals"
• Satish et al 2012, "Is CO2 an Indoor Pollutant? Direct Effects of Low-to-Moderate CO2 Concentrations on Human Decision-Making Performance"
• Law et al 2014, "Relationship Between Carbon Dioxide Levels and Reported Headaches on the International Space Station" (I wonder if the ISS has the same kinds of 'sick building syndrome' that terrestrial buildings do and so the relationship may be confounded? On the other hand, the 'chits' mentioned sound like quasi-experiments.)
• Stafford 2015, "Indoor air quality and academic performance" (Texas natural experiment)

The problem for me is that while it makes sense that since we run on oxygen and the brain uses a lot of oxygen (the whole 'BOLD' thing etc), more oxygen might be better, it has the same issue as Kurzban's blood-glucose/willpower criticism: if the brain needs more oxygen than it's getting, why doesn't one simply breath a little more? While sedentary during these sorts of tasks, you have far more breathing capacity than you should need - you are able to sprint all-out without falling over of asphyxiation, after all. So there's no obvious reason there should be any lack, even more so than for glucose. And shouldn't CO2 levels closely track various aspects of weather? But as far as I know, various attempts to correlate weather and cognitive performance or mood have turned up only tiny effects. In addition, too much oxygen can be bad. So is it too little oxygen or too much nitrogen or too much carbon dioxide...?

Jessica Taylor for lending me a CO2 monitor so that I could see variability in indoor CO2 levels.

What monitor is that? You could try recording CO2 long-term, especially if it's a data logger. Opening windows is something that's easily randomized.

I did some looking and compiling of consumer-oriented devices a while ago: https://forum.quantifiedself.com/t/indoor-air-quality-monitoring-health/799/40 I was not too impressed since nothing hit the sweet spot of accurate CO2 and PPM measurement under $100. The Netatmo looked decent but there are a lot of complaints about accuracy & reliability (checking the most recent Amazon reviews, still a lot of complaints).

I've been thinking maybe I should settle for the Netatmo. I've been working on a structural equation model (SEM) integrating ~100 personal data variables to try to model my productivity (some current sample output), and it would be nice to have even noisy daily C02 variables (as long as I know how noisy and can use it as a latent variable to deal with the measurement error). Correlation-wise, I think backwards causation can be mostly ruled out, and the most obvious confound is weather, which is already in my SEM.

* taken at face value, with reasonable estimates of how much rooms differ from day to day or week to week, CO2 levels would explain a lot or maybe most of variability in IQ tests or cognitive performance!

Weather clearly affects people in a lot of ways, do disentangling the CO2 effects will be hard.

Any idea how high will CO2 go in a room in a normal building, say during winter in a well-sealed residential house? Offices and apartments buildings typically have HVAC systems which have standards for air exchange and such, but a single-family house can do whatever it wants to, including turning itself into an airtight box in the name of energy efficiency...

The trainers are responsible for getting M to do what the trainers want, and the user trusts the trainers to do what the user wants.

In that case, there would be severe principle-agent problems, given the disparity between power/intelligence of the trainer/AI systems and the users. If I was someone who couldn't directly control an AI using your scheme, I'd be very concerned about getting uneven trades or having my property expropriated outright by individual AIs or AI conspiracies, or just ignored and left behind in the race to capture the cosmic commons. I would be really tempted to try another AI design that does purport to have the AI serve my interests directly, even if that scheme is not as "safe".

If I imagine an employee who sucks at philosophy but thinks 100x faster than me, I don't feel like they are going to fail to understand how to defer to me on philosophical questions.

If an employee sucks at philosophy, how does he even recognize philosophical problems as problems that he needs to consult you for? Most people have little idea that they should feel confused and uncertain about things like epistemology, decision theory, and ethics. I suppose it might be relatively easy to teach an AI to recognize the specific problems that we currently consider to be philosophical, but what about new problems that we don't yet recognize as problems today?

Aside from that, a bigger concern for me is that if I was supervising your AI, I would be constantly bombarded with philosophical questions that I'd have to answer under time pressure, and afraid that one wrong move would cause me to lose control, or lock in some wrong idea.

Consider this scenario. Your AI prompts you for guidance because it has received a message from a trading partner with a proposal to merge your AI systems and share resources for greater efficiency and economy of scale. The proposal contains a new AI design and control scheme and arguments that the new design is safer, more efficient, and divides control of the joint AI fairly between the human owners according to your current bargaining power. The message also claims that every second you take to consider the issue has large costs to you because your AI is falling behind the state of the art in both technology and scale, becoming uncompetitive, so your bargaining power for joining the merger is dropping (slowly in the AI's time-frame, but quickly in yours). Your AI says it can't find any obvious flaws in the proposal, but it's not sure that you'd consider the proposal to really be fair under reflective equilibrium or that the new design would preserve your real values in the long run. There are several arguments in the proposal that it doesn't know how to evaluate, hence the request for guidance. But it also reminds you not to read those arguments directly since they were written by a superintelligent AI and you risk getting mind-hacked if you do.

What do you do? This story ignores the recursive structure in ALBA. I think that would only make the problem even harder, but I could be wrong. If you don't think it would go like this, let me know how you think this kind of scenario would go.

In terms of your #1, I would divide the decisions requiring philosophical understanding into two main categories. One is decisions involved in designing/improving AI systems, like in the scenario above. The other, which I talked about in an earlier comment, is ethical disasters directly caused by people who are not uncertain, but just wrong. You didn't reply to that comment, so I'm not sure why you're unconcerned about this category either.

The trainers are responsible for getting M to do what the trainers want, and the user trusts the trainers to do what the user wants.

In that case, there would be severe principle-agent problems, given the disparity between power/intelligence of the trainer/AI systems and the users. If I was someone who couldn't directly control an AI using your scheme, I'd be very concerned about getting uneven trades or having my property expropriated outright by individual AIs or AI conspiracies, or just ignored and left behind in the race to capture the cosmic commons. I would be really tempted to try another AI design that does purport to have the AI serve my interests directly, even if that scheme is not as "safe".

If I imagine an employee who sucks at philosophy but thinks 100x faster than me, I don't feel like they are going to fail to understand how to defer to me on philosophical questions.

If an employee sucks at philosophy, how does he even recognize philosophical problems as problems that he needs to consult you for? Most people have little idea that they should feel confused and uncertain about things like epistemology, decision theory, and ethics. I suppose it might be relatively easy to teach an AI to recognize the specific problems that we currently consider to be philosophical, but what about new problems that we don't yet recognize as problems today?

Aside from that, a bigger concern for me is that if I was supervising your AI, I would be constantly bombarded with philosophical questions that I'd have to answer under time pressure, and afraid that one wrong move would cause me to lose control, or lock in some wrong idea.

Consider this scenario. Your AI prompts you for guidance because it has received a message from a trading partner with a proposal to merge your AI systems and share resources for greater efficiency and economy of scale. The proposal contains a new AI design and control scheme and arguments that the new design is safer, more efficient, and divides control of the joint AI fairly between the human owners according to your current bargaining power. The message also claims that every second you take to consider the issue has large costs to you because your AI is falling behind the state of the art in both technology and scale, becoming uncompetitive, so your bargaining power for joining the merger is dropping (slowly in the AI's time-frame, but quickly in yours). Your AI says it can't find any obvious flaws in the proposal, but it's not sure that you'd consider the proposal to really be fair under reflective equilibrium or that the new design would preserve your real values in the long run. There are several arguments in the proposal that it doesn't know how to evaluate, hence the request for guidance. But it also reminds you not to read those arguments directly since they were written by a superintelligent AI and you risk getting mind-hacked if you do.

What do you do? This story ignores the recursive structure in ALBA. I think that would only make the problem even harder, but I could be wrong. If you don't think it would go like this, let me know how you think this kind of scenario would go.

In terms of your #1, I would divide the decisions requiring philosophical understanding into two main categories. One is decisions involved in designing/improving AI systems, like in the scenario above. The other, which I talked about in an earlier comment, is ethical disasters directly caused by people who are not uncertain, but just wrong. You didn't reply to that comment, so I'm not sure why you're unconcerned about this category either.