Well, how much does it take to create a cloud of poisonous gas that is sufficiently concentrated to induce lethality and also spreads reasonably quickly that people can't escape it?

Bret gives the example of the Tokyo terrorist attack that released sarin in several crowded subway cars... and killed 12 people. Grenades would have killed far more people far more quickly... and that's the point: chemical weapons are far less effective at delivering death than explosives are.

Edit: Wikipedia gives 7mg/m³ for a lethal concentration for one of the Novichok agents, 28mg/m³ for sarin.

Dispersal is a well studied problem: https://fas.org/programs/bio/chemweapons/delivery.html

But to simplify, cluster mutinous are more effective vs troops than a single large detonation. Similarly, rather than starting from a single source you want to spread the weapon from a large number of points. By comparison to 7mg/m3 a 1 metric ton weapon could ideally cover a square kilometer to a thickness of 30m and concentration of 30mg/m3.

Notably, exposure time plays a major role in lethality so covering a wider area makes it much harder to escape from.

How long does the gas cloud last, in various wind conditions?

Really depends on local conditions, but as a benchmark https://en.wikipedia.org/wiki/Ghouta_chemical_attack

Estimates of the death toll range from at least 281 people[3] to 1,729.[14] and 3,600 patients displaying neurotoxic symptoms in 3 hospitals supported by MSF[5]

However, it was considered safe enough for the UN to inspect the area the next day.

Please provide a link when quoting from elsewhhere.

Also you're badly overlooking the obvious. In full "...killing 12 people,[1][2] severely injuring 50, and causing temporary vision problems for nearly 1,000 others"

https://en.wikipedia.org/wiki/Tokyo_subway_sarin_attack

That's not the main point in the series on chemical weapons though. The main point is that modern armies are mobile and well equipped ("the modern system") and that chemical weapons are only well suited to stationary, fixed opponents that are poorly equipped. However, if you are fighting fixed position, poorly equipped opponents there are better and cheaper alternatives that pose less of a risk of self-harm (and war crimes).

In both the posts about elephants and chemical weapons, the theme is the same: they are not useful nor practical given the circumstances so the states make the rational decision not to use them.

The wikipedia article for sarin gas is also extremely frightening, but the example Bret gives of the Tokyo subway attack shows that it’s probably less effective than advertised in a modern military context. The bar isn’t “is this stuff deadly”, it’s is it deadly enough to overcome the expense, complication, and incompatibility with highly mobile modern doctrine to make it worth using on a militarily relevant scale over conventional options.

I’m certainly no expert, so I’m not going to go any further with “is this poison poisonous enough”, but that wikipedia article at least doesn’t seem to make the case.

The tokyo attack was unprofessional. A military attack would be exponentially more effective. Sarin is also tame compared to some of the more nasty stuff out there.

How would militaries be exponentially more effective? As far as I'm aware, the military can't magically change the laws of gaseous diffusion.

A major avenue of research in chemical weapons was efficient delivery. https://fas.org/programs/bio/chemweapons/delivery.html

But to simplify, you don’t just release a gas on it’s own you need a dispersal system like an explosive.

Well, yes, that's kind of the point: given the ability to deliver a explosive to your enemy's position, what makes it worthwhile to include (generally expensive and inconvenient to tranport) chemical weapon material rather than additional explosive and maybe some shrapnel. If it's that the area remains dangerous longer, then congrats, you have impaired your own ability to maneuver into or through that area once the enemy retreats.

WWII shows the absolutely mind boggling amount of bombs you can drop on an area without actually killing that many people. After buildings have already been hit once they make repeated strikes less effective. Chemical weapons are pound for pound vastly more deadly, especially as a follow up and more importantly disable people tying up resources for medical treatment and evacuation.

The most recent example is https://en.wikipedia.org/wiki/Ghouta_chemical_attack which killed or incapacitated ~5,000 with a few rockets. It also acted at least in the short term as area denial which can be extremely useful.

PS: In terms of fighting a near parity army, simply using tents separated by sandbag walls makes bombs or rocket strikes significantly less effective. Add even the threat of chemical weapons and things become significantly more difficult. http://armymomstrong.com/wp-content/uploads/2010/02/2-milita...

Military munitions are engineered dispersal devices. They do a far far better job than plastic bags and umbrellas.

Is it nastier than a conventional explosive weapon, though, at least for military purposes?

Yes. A conventional explosive kills in the immediate area. These things pollute, killing for days/weeks/months.

That's not even the main thrust of his arguments against why chemical weapons are useless.