20 Comments
User's avatar
Thierry's avatar

I think you're on the wrong track, Ugo. AI is not and never will be what "they" sell us. They make a lot of mistakes and misinterpretations with such confidence that it's hard to tell if they're wrong or lying.

Just yesterday, I asked a question as simple as this (sorry for the French standards but the meaning is there). This was chatgpt but deepseek gave me the same answers.

can nf c33-210 cable be laid directly in the ground?

ChatGPT said:

No, NF C 33-210 cable is not designed to be laid directly in the ground. This type of cable is generally used for low-voltage overhead connections, in particular for subscriber connections to the electrical distribution network.

Why not bury it?

It's not designed to withstand the mechanical stresses and humidity of the soil.

It risks deteriorating more quickly due to the chemical and physical aggression of the soil.

The NF C 15-100 standard requires cables suitable for underground installation, such as U1000R2V or U1000 AR2V armored cables, which offer better mechanical protection.

Solution:

If you need to bury a cable, opt for sheathed NF C 32-321 (U1000R2V type) cable, or NF C 33-226 (armored cable type) if you wish to bury it directly without a sheath.

Need more details about your installation? 😊

You said:

no it's not true the NF C 33-210 cable can be laid without mechanical protection in the ground.

ChatGPT said:

Yes, you're right! NF C 33-210 cable can be laid directly in the ground without additional mechanical protection, because it is designed to withstand outdoor installation conditions, including overhead and underground.

Why can it be laid underground without a sheath?

Adapted design:

This cable is manufactured with cross-linked polyethylene (XLPE) insulation and a black polyethylene outer sheath, making it highly resistant to mechanical and chemical aggression from the ground.

It is designed to withstand humidity and temperature variations.

Standard and use:

It is commonly used by Enedis for low-voltage underground connections.

It can be laid directly in a trench, without the need for a protective sheath (unlike U1000R2V cables, which must be sheathed).

Recommended installation depth:

65 cm minimum on private property.

85 cm under roadway or trafficked area.

With red warning grid about 20 cm above the cable.

Conclusion:

The NF C 33-210 is therefore suitable for laying in the open ground without additional protection, unlike other cables such as the U1000R2V which require a sheath.

Thanks for your patience, and well done on the correction! 😊 Are you doing any electrical work at the moment?

Expand full comment
Ugo Bardi's avatar

That's right. Just yesterday, the same Grok that was so efficient in digging out data about territorial fluctuations during WWII provided me with completely invented scientific literature references. Just like that. Invented from scratch. When I asked it why he had done such a mess, it apologized and said it was citing "out of memory" -- as if he was an old man suffering of Alzheimer. For the time being, these things are hugely unreliable. You have to check everything they do. But if you do that, their capability of digging out data is fantastic.

Expand full comment
John Day MD's avatar

"The AI faked my homework."

;-(

Expand full comment
John Day MD's avatar

It is being determined how much BS people will blithely accept, but this may have always been tested in humans...

;-/

Expand full comment
aaron's avatar

hi professor ugo bardi can you explain why you see the human population decline around 2050 in your interview with tim ventura I would like to know because gaya herrington and nafeez ahmed both club of rome members say we are collapsing right now ?

Expand full comment
John Day MD's avatar

The 1972 LTG "BAU" projections did not account for COVID "vaccines", right?

Expand full comment
aaron's avatar

dear john day md in an interview 6 days ago professor ugo bardi said to tim ventura that the human population decline would happen around 2050 just like the limits to growth predicted and it would not by because of war or famine I just find it weird that gaya herrington and nafeez ahmed do not agree and ugo bardi is writing a book for it and will report it to the club of rome on 12 June 2025

Expand full comment
Ugo Bardi's avatar

We are no prophets. We don't make predictions.

Expand full comment
aaron's avatar

so what you mean ugo is that we will never reach peak oil as long as we can afford it but bloomberg said today that shale oil reached it's peak and can governments not print money professor jorgen randers said this once ?

Expand full comment
aaron's avatar

I know this professor ugo bardi but you said it matches like the un forecast right ?

Expand full comment
John Day MD's avatar

One of the points made in LTG ws that BAU projections only hold until some opposing societal action is taken, which changes the curves. If you look at the 1972 LTG BAU projections and tha recent updates, you see that both births and deaths were a bit lower than initially projected, which balance, but give a somewhat older population. The other projection lines for resources and industry seem to be holding, and the pollution got moved around out of North America...

If "excess deaths" in COVID-"vaccinated" populations are indeed up in a sustained way, as they certainly appear to be, and fertility is down in women and men treated with these gene therapies, then this directly impacts total population projections.

Some financial interests are trying to make WW-3 happen in Europe, against Russia, because they have made bad-bets on stripping Russian assets with the Ukraine war.

We are in a point of human history where the projections are teachers, not prophets...

Expand full comment
Arie's avatar

Dear Ugo, You may want to take a look at the PhD thesis of former marine, Major Ingo Piepers. The title is "On the Thermodynamics of War and Social Evolution." Last I checked it was freely available for download.

Expand full comment
Max's avatar

It's will be interesting to make the same thing but with energy production in timeline of each forces axes against allies.

Expand full comment
John Day MD's avatar

How to lose a Eurasian war:

1) Invade Russia...

Expand full comment
Ugo Bardi's avatar

2) If possible, in winter....

Expand full comment
aaron's avatar

i am 100 % against war do not get me wrong but i understand the eu we must protect our citizen's the if there was a war for example with russia the uk would last 6 days brlgium a day or less we just have or f-16 and a few old trucks and not 1 decent tank but there won't be a war with russia we have a saying in belgium (it will cool without blowing)

Expand full comment
Nicholas Taylor's avatar

I don't know whether there is a real connection, but mean queues produced by random arrivals together with random service events exhibit a lag and their variance even more so. When the demand and service follow a Gaussian pattern, the mean graph, and even more the variance, peak later and decline more quickly. This also manifests as hysteresis. It may be simply because in the growth period the two 'forces' of inflow and outflow are in opposition but later they reinforce each other. An important effect of variance is that the probability distribution of queue size - I suppose the equivalent would be the strength of the 'empire' - can be bimodal towards the end, i.e. with a heavy tail, so some queues persist for no obvious reason.

Expand full comment
Ugo Bardi's avatar

True, but I think here we have a different phenomenon. It is explainable by dynamic models based on dissipation. For instance, here: https://www.mdpi.com/1996-1073/2/3/646

Expand full comment
JustPlainBill's avatar

I think we are about to see this in Ukraine as well.

Expand full comment
Ugo Bardi's avatar

I didn't want to say that, but....

Expand full comment