HelloThere @ hellothere @sh.itjust.works

Reddit refugee

---

Read more

Posts

1

Comments

330

Joined

2 yr. ago

People like to view the problem as a paradox - can an all powerful God create a rock they cannot lift? - but I feel that's too generous, it's more marking your own homework.

If a system can both write text, and detect whether it or another system wrote that text, then "all" it needs to do is change that text to be outside of the bounds of detection. That is to say, it just needs to convince itself.

I'm not wanting to imply that that is easy, because it isn't, but it's a very different thing to convincing someone else, especially a human, that understands the topic.

There is also a false narrative involved here, that we need an AI to detect AI which again serves as a marketing benefit to OpenAI.

We don't, because they aren't that good, at least, not yet anyway.

It's literally a marketing blog posted by OpenAI on their site, not a study in a journal.

Regardless of if they do or don't, surely it's in the interests of the people making the "AI" to claim that their tool is so good it's indistinguishable from humans?

There is definitely some memory holing going on, the bit implying capex cuts were used to fund increases in opex expenditure like pay rises is pure fantasy. Anyone who isn't looking to be lied to knows that the current pay crisis in the public sector is due to a decade of pay cuts, not rises.

In reality, there were situations where councils had to maintain mandatory service provision, and their opex budgets had also been cut in to the bone, requiring further cuts to capex to allow a transfer of cash from one to the other. But we all know that isn't what the Spec is wanting to say.

Now then, while this is possibly the closest I've seen the Spectator come to admitting that austerity is a complete failure, they will always come up short of fully admitting that the reason Britain is fucked right now is because of the economic philosophy they vocally advocated for.

They and their ilk are what broke it.

Nutjobs gonna nutjob, more at 11.

Wat.

The main complaint with buds is that they are comparatively easy to lose, on top of the small size and lots of glue making repairs difficult and heat hard to dissipate.

The ifixit teardown of the FP buds was also scathing.

https://www.ifixit.com/Teardown/Fairphone+True+Wireless+Earbuds+Teardown/146449

The price is due to repairability and better sourcing of materials. Much like cheap chocolate, iPhones, etc, are artificially cheap due to terrible labour practices in the supply chain.

If all someone cares about it technical performance vs price, they will never choose a fairphone.

Their selling point is repairability and well sourced materials - hence the price.

I've had a FP4 since launch, and the USB-C to 3.5 adapter has been attached to my headphones ever since. I've never been in a situation where choosing whether to charge or listen has caused a problem.

If your household income (excluding state pension) exceeds the free tax threshold (£12,500) then you don't qualify for a state pension.

This is laughably cruel, and barely 2k a year above the state pension. In what reality is 12k a year anything other than poverty?

If this is funded via pension savings, you need something in the region of 250-300k to either buy an annuity or have a safe withdrawal rate to have an income of 12k a year.

Assuming a 4% average rate of growth - after charges and inflation - from your 18th birthday until retirement at 67, you need to be contributing the equivalent of 167 quid in today's money, every month, for 49 years, to get that 300k. That may not sound like a lot, but keep in mind a few things:

• the vast majority - 200k - of that value comes from compounded growth, not contributions, making you extra vulnerable to underperformance
• what and when people can afford to save heavily depends on their circumstances, which change throughout their lives. If you went to uni and started contributing at 23 instead of 18, you'd lose 50k just in lost compound effects. Same applies to stopping work to have kids, to support an ill relative, etc.
• according to this recent times article the average pension pot is approx 37k total, so the vast majority of people are no where near close saving even this seemingly low figure. 37k gets you whopping income of ~100 quid a month.

As you say further up, the country spends considerable sums on the elderly when you include the NHS, etc. That figure is not going to decrease if even more people are in poverty. Health costs have this annoying habit of getting higher the closer to death you are, and accelerating that ain't the best idea.

Throwing in the towel just because it's expensive is not the answer.

While there are many valid points in this piece, we need to not get blinded by thinking the state pension is anything more than a minimum living standard, even with the triple lock.

The full state pension is 203.85 quid a week.

For those who rely on it, it's fuck all, which is also why housing benefit, etc, exists.

Wealthy boomers are not wealthy because of the state pension. They are wealthy due to a massive wealth transfer due, almost entirely, to a massive house price boom as well as having interest only endowment mortgages.

Poor pensionsers still exist, and they still live in poverty.

My good fellow, if you believe that cap and floor contracts somehow disproves my point, then you really do need to go back and re-read what I've been saying all along, not just what you think I've been saying.

For the final, final time:

• Both types of generation have the same capex payback considerations, which will be spread over the expected lifetime of the project.
• To be crystal clear, I am not saying the literal cost is identical, obviously the actual Pounds and Pence cost of 1 small onshore windmill is different to 1 massive offshore windmill or 1 CCGT.
• Capex costs do not change day to day due to whether you are generating or not because they are, by definition, unrelated to operation.
• Prices fluctuate throughout the day, and may go up or down between when you start generating and when you stop.
• The input 'fuel' used by solar, wind, and tidal to generate is free, which significantly reduces their Next Unit Cost compared to fossil fuels or storage, which is the key Opex cost when considering whether to generate at a specific time on a specific day.
• Your ability to generate also changes. If you run a solar park and the price is sky high, but it's 10pm in December, tough.
• If the gas you're burning, or the stored power you're realising (batteries, pumped hydro) cost more than the current price you'd be paid, then it is extremely unlikely that you're going to do it.
• Cap and Floor contracts, which specify a maximum and minimum price you can be paid, merely put limits on how much money you can make on each unit sold
• This means that, all else being equal, renewables are able to operate profitability at lower market prices compared to fossil fuels
• As all else is not equal, the viability of a project is based on the projected difference between average costs, including capex payback, and average price.
• These factors must be considered regardless of means of generation - the price of each factor will again obviously be different
• The difference between average cost and average price is your profit margin / return on investment
• The higher the floor relative to projected market price, the easier it is to operate profitability
• The lower the cap, the harder it is
• The closer your projected unit price and projected costs are, the less profit you'll get and the less viable the project is

Again:

• the choice to generate on a specific Tuesday at 1am is based on the price you'll be paid, versus Opex, on that day
• the choice to build the generator in the first place is based on the projected profit from lifetime costs versus lifetime earnings

I will only reply if your next comment actually brings something new to the conversation.

This is a fair point - peaking is more complex, especially if we're considering batteries where their generation cost is going to include probabilistic opportunity costs - ie how confident are they that the price won't fall further and/or if this is the peak of the spot and best time to sell.

But yes, over the decades you'd be looking to run to utility for, you're looking at blended averages to calculate the return.

I explicitly covered this in my 3rd comment - quoted below.

Capex payback is important when businesses are evaluating building new generation. The spot price at 1am on a random Tuesday has nothing to do with whether you're choosing to build new infrastructure. What does matter is average unit prices, over time, not one data point.

I'm done here, you're clearly not reading what I've said if you genuinely believe I've said capex never matters.

• Why are you linking to a definition of economic viability for buildings? The main cost of traditional fossil fuel based electricity generation is the fuel source, over lifetime, not initial construction
• Higher returns due to high average costs certainty attract investment - I said as much in my first comment
• I agree with disconnecting prices from last generator, and that it will reduce inflation, because it would reduce the average price, all else being equal
• A lower average price would extend payback periods which may discourage investment when compared to the current pricing model, again all else being equal. This is a difficult comparison because it would presume an ability to choose between the two payment methods, which you can't.
• We are still talking about average prices, over time, which again is different to spot prices, and both above points are further proof that your original statement of renewables being economically inviable is incorrect.

I'm not the person you originally replied to when you falsely claimed that renewables are only economically viable because of last generator pricing.

I have explained why that isn't the case, how both generation and new capacity decisions are made, the different aspects those decisions consider, and how because their next unit cost is lower due to generation input being free they are able to operate profitability at lower spot prices than are achievable for fossil fuels.

One last time - capex payback is a consideration when building new capacity, yes, but that is based on average prices over decades. It is not a consideration when choosing whether to power up or down on at a specific time on a specific day.

Attempting to simplify this to just capex is wrong.

Context of the situation is important. You can't use them interchangeably.

Capex does not matter when we are talking about choosing to generate using existing infrastructure, because capex amortisation is the same regardless of whether you're generating or not. Choosing whether to generate at 1am on a random Tuesday has nothing to do with your previous capex, but everything to do with your next unit cost. If price is higher than cost, you'll generate, it not you (probably) won't.

Capex payback is important when businesses are evaluating building new generation. The spot price at 1am on a random Tuesday has nothing to do with whether you're choosing to build new infrastructure. What does matter is average unit prices, over time, not one data point.

Their unit cost is zero?

Apologies, I meant to say Next Unit Cost, didn't spot the missing word. As in the cost to produce one more unit of electricity. This measure ignores capex and is just opex plus associated usage costs like amortised wear and tear, depreciation, etc.

It's near 0, but not 0, because the input (wind, sun) is free, whereas gas (for CCGT) is not. As such renewables can sell in to the grid, profitably, at lower prices than fossil fuels can.

The big problem, as you show with the example of cloudy winter days, is disconnecting generation from usage, via storage.

Green gas from anorobic respiration, including hydrogen (or just blue hydrogen, made via electrolysis of sea water) is a useful step for retrofitting, but not something - in my opinion - that should be considered for new builds.