Cameron Cooke

Hyperactive Omniheurist, easily confused.

[Cameron Cooke @CGCooke]

@flexibledragnet @grok Is David using the term counterfactual correctly?

[Cameron Cooke @CGCooke]

@flexibledragnet @grok Is David using the term ‘counterfactual’ correctly?

[Cameron Cooke @CGCooke]

@flexibledragnet Do you have the ability to understand counterfactuals?

Because all the evidence seems to suggest that you don’t. And that would explain a lot.

[Cameron Cooke @CGCooke]

@flexibledragnet Just looking at the first of the 3 deals gives you 2609 MW of capacity vs the counterfactual.

Signing a 20 year PPA is a choice is decision made by professional engineers.

They have run the numbers, looked at all the options, and made a rational decision.

[Cameron Cooke @CGCooke]

@pbeisel There is nothing to argue, because he’s just wrong.

Even if the gear adds zero drag, its weight alone results in 2.5-3x more fuel burn than what’s he’s claiming.

That’s just physics.

Either you don’t understand the physics, or you don’t want to admit he’s wrong.

[Cameron Cooke @CGCooke]

@swissTom124 @pbeisel @grok What is the mass of an IFC system install on a 737-800? Please do an itemised breakdown by component.

[Cameron Cooke @CGCooke]

@swissTom124 @pbeisel @grok What is the percentage increase induced drag per 100kg for a 65 Ton TOW 737-800?

[Cameron Cooke @CGCooke]

@pbeisel Induced drag is proportional to the square of lift.

Even if the antenna has zero mass, and zero drag, everything else brings you north of 100KG.

For TOW of 65 tons, that’s a 0.3% increase in induced drag.

Stuff you just know, if you actually study engineering.

[Cameron Cooke @CGCooke]

@flexibledragnet @roobar1984 Google is investing in a developer of Solar, Batteries, Wind (Clean power) in addition to Nuclear (Clean power).

Are you forgetting the NextEra and Elementl PPAs from this year?

How does this contradict the claim that including nuclear in the mix reduces total cost?

[Cameron Cooke @CGCooke]

@flexibledragnet @roobar1984 @PaulMan77121652 @simonahac @ShackelWill ‘Pro electricity economics’ implies:
1. Letting consumers choose
2. Valuing consumer choice over a theoretical model.
3. Knowing the difference between Capex and Capital flows.

[Cameron Cooke @CGCooke]

@roobar1984 @flexibledragnet @PaulMan77121652 @simonahac @ShackelWill It’s listed as $228B USD in the Frank Knight report.

However, it’s just a number you would know off the top of your head if you closely followed the DC space.

[Cameron Cooke @CGCooke]

@flexibledragnet @roobar1984 @PaulMan77121652 @simonahac @ShackelWill You have confused Capital Flows with Capex.

Capital flow:
• Real estate investment
• M&A transactions
• Buying or selling DC companies

Inflated by the purchase of (Australian) AirTrunk by Blackstone

Capex: How much is actually being invested building DC’s. (Over $200B)

[Cameron Cooke @CGCooke]

@roobar1984 @flexibledragnet @PaulMan77121652 @simonahac @ShackelWill They want to buy as much cheap *clean* energy as they can get their hands on.

I’m not sure why you are struggling with the idea that clean power includes nuclear.

Of that nuclear helps stabilise Europe’s (Including Ireland & Denmarks) grid.

[Cameron Cooke @CGCooke]

@flexibledragnet @roobar1984 @PaulMan77121652 @simonahac @ShackelWill Also ~10% Nuclear. I’m not sure you keep confusing pro nuclear with anti clean energy.

[Cameron Cooke @CGCooke]

@flexibledragnet @roobar1984 @PaulMan77121652 @simonahac @ShackelWill Would you mind sharing a source for that graph?

MSFT’s US based DC capex was north of $20B USD for 2024.

The global DC spend was north of $200B USD in 2024.

[Cameron Cooke @CGCooke]

@roobar1984 @flexibledragnet @PaulMan77121652 @simonahac @ShackelWill You keep going back to arguments saying that a theoretical model says X.

While ignoring the behaviour of (in your words) ‘hyper rational’ consumers.

As an engineer, I’m used to dealing with conflicts between my models and reality. But I know the reality is what counts.

[Cameron Cooke @CGCooke]

@roobar1984 @flexibledragnet @PaulMan77121652 @simonahac @ShackelWill Latency isn’t an issue for LLMs. Models take 5+ seconds to fully generate a response so 100’s of ms is meaningless.

Thats why all the training and inference is in the US ☢️.

Latency is so little of a problem that Meta and Apple don’t even bother with DCs in Australia.

[Cameron Cooke @CGCooke]

@roobar1984 @flexibledragnet @PaulMan77121652 @simonahac @ShackelWill I’m glad that you agree that Including nuclear in your mix, minimises risk and reduces your overall power cost.