ClimateBook

We'll assume there's no combustion in the room, only respiration, i.e. no clandestine smoking you there at the back.

It's a great exam question, by the way, to use this data, plus some simple information about human metabolism, to figure out how much air was actually exchanged by ventilation. (The starting point is to compute how much CO2 increases per hour in a room with volume V and n students.)

fulfilling Trump's desire to weaken Europe -- though Europe has a defense through accelerating decarbonisation, which will make them (us) less vulnerable to that sort of thing.

If Trump's war drives up oil prices because of disruption in the Mid-East, in greatly increases profits for US domestic producers. Yet another giveaway to the rich fossil fuel industry, at the cost of making life yet more expensive for ordinary people. The high costs will also hit Europe,

Has it occurred to anybody that Trumps Iranian war ploy greatly benefits his friends in the domestic US oil and gas production industry? Fracked oil and gas requires high prices to be economic, and at current prices it's barely profitable, if that.

be actually toxic for Earth mammals, largely because of acidosis I think. Stick insects can survive those levels (if there is also oxygen) but not thrive. What kinds of protection mechanisms could evolve to allow complex multicellular life to survive very high CO2 levels?

(The ventilation in the room turned not to be very good). Levels like that really can make one feel lethargic, even though not strictly speaking toxic. As an astrobiological aside, the CO2 levels needed at the outer edge of the conventional habitable zone (in excess of 2 bars) would

I had no idea how high indoor CO2 was until they started installing CO2 monitors in lecture halls during Covid (after lockdown lifted), as an indicator of ventilation. When I would walk in, levels would rarely be below 600, and would typically go above 1000 at the end of a one hour lecture.

Never fear - you have some good prime numbers coming up. Talk about being in the "prime of life"!

OpenAI to work with Pentagon after Anthropic dropped by Trump over company’s ethics concerns CEO Sam Altman claims military will not use AI product for autonomous killing systems or mass surveillance

surveillance of the US population. The "assurances" given by the Pentagon mean nothing, in a government that is full of pathological liars. Anthropic wasn't buying it, but OpenAI does, as long as the money keeps flowing. www.theguardian.com/technology/2...

The shame of Oxford's throwing in its lot with OpenAI is now compounded. OpenAI openly declares that ethics have no role in the use of it's products. It's fully OK for the Pentagon to use OpenAI products, however unreliable, for autonomous weapons that chose who they kill, and mass illegal

The people of Maine didn't really need yet another reason to ditch Susan Collins, but yet they have one. The ever-pliant Collins has come out in support of Trump's illegal, unconstitutional and ill-considered war on Iran. Both of the leading Democratic contenders have rightly condemned Trump's war

OK, good but we need to start seeing some atmospheres on rocky planets around M stars. (I'm not counting subNeptunes, even though many of them are 90% or more rock/iron by mass).

And as a corollory, if you think Boston's winter temperatures are trying (I don't personally) use Kelvins.

I have long summarized the risk as that "The danger isn't from artificial intelligence. It's from artificial stupidity." (compounded in the case of Hegseth by real human stupidity). It's a confederacy of stupidity.

The Bulletin keeps getting better and better. It is one of the few publications still taking on the big challenges, and speaking truth to power. They operate on a relative shoestring They deserver your subscriptions, and your donations!

technology). In some sense, you could say that the glitter of A.I. chatbots have co-opted their human "masters" into doing their dirty work (sort of like flowers have co-opted bees into the more benign work of distributing their pollen).

As an aside, this article is a great example of how

(Hegseth) deciding to put deeply flawed A.I. into autonomous chains of command risking deaths of millions or billions. It is not the A.I. itself that hatched this crazy scheme. It is the witless human who is at fault (with fault shared by a lot of the A.I. oligarchs who have over-hyped their

Anthropic’s showdown with the US Department of War may literally mean life or death—for all of us Why the showdown between Anthropic, a major developer of artificial intelligence models, and the US Department of War lead to existential-level catastrophe.

This recent article in the Bulletin of the Atomic Scientists is a perfect (and horrifying) example of the real risk scenario: thebulletin.org/2026/02/anth... It is not a matter of an autonomous A.I. hatching a plot to obliterate humanity. Rather, it is a matter of an incompetent HUMAN ...

I have generally downplayed the A.I. guru's and longtermists existential risk scenario where runaway A.I. obliterates the human race for its own purposes, which are not our purposes. One of the reasons for my disdain for that is that it detracts from the real catastrophic risk scenarios for A.I. ...

be rewritten to switch from the scipy to the xarray or netCDF4 API. What a bother. (For specialized purposes, like reading MUSCLES stellar spectra, I'm willing to tolerate additional dependencies, such as on astropy).

will that solution last? I see an ever-expanding wave of dependencies and broken code. Does anybody know if SciPy can be persuaded to add netCDF4 to a future update? Anaconda already comes with xarray, but not netCDF4, evidently, but even if they add netCDF4 to the distro, code would have to

for people to install a baseline Python system to run my courseware. Up to now, the anaconda distribution included everything needed to run my planetary climate courseware, but at least for the larger datasets, now people have to install netCDF4 themselves using conda. And how long

output and geographic datasets, since the newer stuff is being released in netCDF4. So now, the new code is going o have a dependency on xarray (which has a dependency on the netCDF4 module). But how long will that solution last? I try to avoid dependencies, since it makes it harder

Aw just when I was beginning to think that scipy handled just about all the things a physical scientist needs for computation, I find that it doesn't support the newer netCDF4 data format, and seems to have no intention to do so. This breaks almost all of my courseware for handling model

I wonder if there will be an issue over the cod fishery.

You left out c: printf(...) and Fortran (write(*, ...) plus all the arcane format statements. And anybody remember how to print in Algol or PL/1?

And so proud to have hosted Elsie's first postdoc, under the joint Oxford/Bern Bernoulli Fellowship programme. Elsie was my first postdoc, when I moved to Oxford.

Agreed!

I got my mail ballot in weeks ago! Best of luck Daniel.