After receiving the orders that were actually printed from an Internet Explorer 6 only website, and faxed over from another office before being re-scanned in along with a barcode that usually failed to make it over the fax, hence the need to hand-type things. True story (not for JR specifically, but circa 2013)
Hydrogen is not great for airplanes since the extremely low density makes the tanks too large. The best solution would be synthetic hydrocarbons (synthesized using hydrogen) which can outperform fossil jet fuel.
These problems are grossly exaggerated in popular discussions. Hydrogen has been routinely transported and stored in standard steel cylinders for over a century. Most cities originally used coal gas (50% hydrogen by volume) for heating and illumination before switching to natural gas after World War II. What kills the idea is the abysmal efficiency of electrolysis and hydrogen fuel cells. Standard high-voltage DC power lines would be much better suited for getting solar power from Arizona to Alaska.
This is not quite true. The original gas pipes in most cities were built for "town gas" which was produced from coal and is 50% hydrogen by volume. The infrastructure could handle hydrogen just fine, but the low conversion efficiencies make it impractical.
The Bohr radius is the result of a simple classical physics calculation (a common exercise for undergraduates in their first year). It depends only on the electron mass and the fine structure constant which is the strength of the electromagnetic interaction. In the SI system, the speed of light has a fixed value which defines the unit of length.
IQ tests only work if the participants haven't trained for them. If they do similar tests a few times in a row, scores increase a lot. Current LLMs are hyper-optimized for the particular types of puzzles contained in popular "benchmarks".
Companies are optimizing for all the big benchmarks. This is why there is so little correlation between benchmark performance and real world performance now.
Yes, LLMs have become extremely good at coding (not software engineer though). But try using them for anything original that cannot be adapted from GitHub and Stack Overflow. I haven't seen much improvement at all at such tasks.
No shot, their classic engineering ability has exploded too.
The amount of information available online about optics is probably <0.001% of what is available for software, and they can just breeze through modeling solutions. A year ago was immediate face-planting.
The gains are likely coming from exactly where they say they are coming from - scaling compute.
This is highly misleadling. Nobody is setting the price, it is determined by an open market. This naturally drives it towards the price of the cheapest energy source with available capacity (often natural gas). It would be irrational to sell electricity for cheaper than this. If more batteries get deployed, the price will more often get set by battery storage instead.
> In this model, the price of electricity is set by the most expensive source needed to meet demand at any given time. Often, this is gas-fired power plants. Even if cheaper renewable sources like wind and solar are supplying a significant portion of electricity, the overall market price is influenced by the cost of gas.
If you don't cover 100% of the current power usage from batteries, the price will be price of gas plants.
The gas plants could be 1% of given moment, yet still set price
It's not like the price is set for the whole year, though. The price is set each half-hour, so it does matter what percentage of the time the gas peaker plants are necessary to supply the grid. This makes the effect of renewables on prices quite nonlinear: if they can never supply 100% of the grid, then they have zero effect on the average wholesale price. But going from 0% to 99% is a large part of the hurdle, then the transition from 99% to 105% will have a very large effect on the pricing (of course, given the variable nature of renewables, this will get blurred a bit more: currently the UK grid is entirely renewables about 1% of the time. But doubling the renewable capacity will raise that percentage to a lot more than 2%).
> The gas plants could be 1% of given moment, yet still set price
Makes sense. Since no one would build that last 1% (or then, last 10%) of needed capacity due to it being wildly unprofitable. Then you are dealing with rolling blackouts or even worse.
The cost of a watt is not fungible. Reliable electricity is worth many multiples more than an unreliable grid no one can rely on being there when they need it.
This is a very sensible way to structure an electricity market. It’s got to be set at the marginal price, otherwise you mess up incentives of cheaper producers.
Sure, it's a marginal price. It is surprising to me that HN struggles to understand marginal pricing, it makes me more likely to assume when I see such people unhappy with taxation that they probably also don't understand marginal taxation.
Marginal differences have a cliff effect, which is one of the things US Republicans are worried about in the event Trump isn't able to subvert or abolish entirely this year's elections. If you've gerrymandered every seat so that you'll win by 3-5% and then your support collapses 10% across the board then you lose all those seats, not 10% of them. Ouch.
For that 1% in reality it's probably not quite the case, my understanding is that most of the gas plants pay a significant price in terms of efficiency loss and wear on the turbine, for restarts, so e.g. make 10MW for an hour, switch off for an hour, then make 10MW for an hour is 20MWh produced, but incurred a stop-start. The 20MWh might equate to £1000 of gas burned, but the stop-start has an effective price of £500. So you need to charge £75 per MWh to break even. Or, you could sell for £60 per MWh, deliver 10MWh for all three hours, 30MWh, £1500 of gas burned, no stop-start overhead, your overall costs were the same but you got more profit because 30 x £60 = £1800 instead of 20 x £75 = £1500.
Very true about the Republicans. In the special elections since the general we have seen shifts of this level. Unfortunately, I strongly suspect subvert is what's going to happen.
That is exactly what I wrote. Gas plants being at 1% implies that there is no cheaper source with available capacity. Why should anyone sell electricity for less then?
So if I offer my services providing electricity through a bicycle transformer for the cheap cheap price of $1000 per kWh does that mean everyone has to pay that price.
Every 30 mins the UK energy suppliers put in a bid for how much energy they can produce and what price they will do it for. The UK then selects the cheapest N companies to fufill the predicted energy demand. Each company selected is then paid the price of the most expensive supplier chosen which is usually gas.This is a simplification of what the Octopus Energy CEO explains in the link below starts ~1:40.
No. The operative word there is "required". The grid sorts the various providers cheapest to most expensive, then uses all the power from each until they don't need the power anymore, at which point they pay everyone who they did take power from the rate of the highest winning bidder.
If you were offering power at $1000/kWh, you would simply lose the auction.
Imagine the scenario where Alice, Bob, Charlie, and Daniel are each selling power at $1/kWh, $2, $3, and $4 respectively. We need 30 kW of power.
Alice bids 10 kW at $1/kWh. We draw power from her, but we still need 20 kW
Bob bids 15 kW at $2/kWh. We draw power from him, but we still need 5 kW.
Charlie bids 30 kW at $3/kWh. We draw 5 kW from him. We don't need any more power, so Charlie has set the price at $3/kWh
Over the next hour, Alice gets $30, Bob gets $45, and Charlie gets $15. Daniel gets nothing, because he was out bid.
There is a significant fine to be paid by the non delivering supplier. This still happens and that is why there is also an auction for reserve power. Oversupply is fined even higher as that is also bad for grid stability.
Think about this like a market. Suppose yes, there is demand for your power at $1000/kwh.
What is the market pressure here? Suddenly a ton of new capacity in solar, gas, etc, will come online and drive that price down because there will be much more capacity before you reach the point of $1000/kwh purchases.
The alternative is that people get paid at cost of production, which if you think about it is less fair. Why should a gas turbine get paid $67/kwh and a solar cell or battery get paid less? It also means that the market incentivizes more cheaper energy as a rule, because they take profit.
Would you go to the gas station charging $2 above market price just because their costs are higher to produce the gas?
As I understand it (and even if I’m broadly right I’m greatly simplifying) there’s an auction system and if demand is X kilowatts, they line up all the bids to supply in cost order and draw a line at X kilowatts. All successful bidders receive the price bid by the highest successful bidder.
There are rare times in this kind of market where the price does go very high (though not to $1000 per kwh), and those brief periods push average prices up substantially.
In markets where batteries are going gangbusters, they are squashing many of these peaks and thus reducing average prices paid by consumers (though not as much as you’d hope because the majority of retail electricity costs are distribution rather than generation).
reply