There's also cis-male people who will "pass" that SRY test if they take it for some reason...
This is a dumb ass way to try and define the woman's category... which is about the expected result of bigots trying to work backwards from the result they want headlines about.
We are not. The reason we wanted to announce early was so that folks had plenty of time to opt-out now. We've also added the opt-out setting even if you don't use Copilot so that you can opt-out now before you forget and then if you decide to use Copilot in the future it will remember your preference.
Would you be able to comment on https://news.ycombinator.com/item?id=47522876, i.e. explain the legal basis for this change for EU based users? If there is none, you may have to expect that people will exercise their right to lodge a complaint with a supervisory authority.
Why would you expect an engineer to be able to comment on legal affairs? Presumably it was cleared with Microsoft's legal department or whatever GitHub's divisional equivalent is.
More likely they have RCE vulnerabilities known to various governments than intentionally made secret backdoors... which is worse since a backdoor would probably at least only be usable by the county that manufactured it (for example see Jia Tan's attempted backdoor).
1. LLMs aren't "efficient", they seem to be as happy to spin in circles describing trivial things repeatedly as they are to spin in circles iterating on complicated things.
2. LLMs aren't "efficient", they use the same amount of compute for each token but sometimes all that compute is making an interesting decision about which token is the next one and sometimes there's really only one follow up to the phrase "and sometimes there's really only" and that compute is clearly unnecessary.
3. A (theoretical) efficient LLM still needs to emit tokens to tell the tools to do the obviously right things like "copy this giant file nearly verbatim except with every `if foo` replaced with `for foo in foo`. An efficient LLM might use less compute for those trivial tokens where it isn't making meaningful decisions, but if your metric is "tokens" and not "compute" that's never going to show up.
Until we get reasonably efficient LLMs that don't waste compute quite so freely I don't think there's any real point in trying to estimate task complexity by how long it takes an LLM.
Half of the "heavy duty vehicles" (which I believe is roughly similar to the classification you are using) sold in China in December were electric. Between rapidly improving batteries and maturing technology for swapping batteries as a refuelling strategy electrification of trucks is the obvious and inevitable future. They are simply cheaper to operate.
Good for the Chinese. The rest of us do not have the upfront capital to purchase these trucks. And there is still the matter of fertilizer, concrete, bulk chemicals etc. And solar panels. There is a very good reason why solar psnel factories (like JinkoSolar run off coal or hydro and not solar power.
That's one of the wonderful things about automotive infrastructure. You can make gradual incremental changes and slowly improve the entire system. It may not be fast enough or cheap enough, but you can still make it happen.
> The rest of us do not have the upfront capital to purchase these trucks.
We can afford what we can do. We don't need to do what we can afford. If we wanted to build and deploy electric trucks enmasse like China then we could do it, regardless of upfront capital.
> The rest of us do not have the upfront capital to purchase these trucks.
You don't need any upfront capital. Do it when the trucks become due for refurbishment a truck. Then it's almost a no-brainier, as its cheaper convert it to an EV: https://www.januselectric.com.au/
> The rest of us do not have the upfront capital to purchase these trucks.
Isn't this the purpose of a loan? You have a truck with a higher purchase price that adds ~$2000/month to your loan payment but then you save ~$3000/month in diesel.
And you're saving a lot more than that in diesel when it's $5/gallon.
We don't know that. Beijing might have been investing in them as insurance against its not being able to get enough diesel fuel to run an all-diesel fleet of trucks, so countries that are self-sufficient in oil shouldn't just blindly imitate Beijing's move.
We know that because we know how much they cost, how much they cost to operate, and the same for diesel trucks. The technology here isn't a bunch of state secrets.
Here you're just repeating the assertion I called into question ("they are simply cheaper to operate") -- or more precisely you are implying it. Does your not repeating it outright mean you mean to slowly distance yourself from it?
If you have evidence that there is a fleet of electric trucks anywhere (big enough to make a dent in China's transport needs) whose actual total cost proved to be less than a fleet of diesels doing the same work would have cost, then share it. If all you have to offer is words to the effect that "an examination of the relevant technologies by any competent analyst will of course find that the battery-powered fleet would be cheaper", then I repeat my assertion.
I was not in fact repeating the prior assertions. I was explaining why we know they are cheaper to operate. Because we know the costs of both them and the alternative. No fancy deductions needed where we're arguing "well electricity is cheaper than diesel but we don't know how much they use" or something.
I am certainly not backing down from the claim that "they are simply cheaper to operate". That is an absolutely trivial claim that is entirely obvious to anyone even remotely familiar with numbers in this space.
I would note I was discussing trucks that swap batteries - and thus the "paying drivers to wait around while trucks recharge" step doesn't exist. I'll also note all the other costs you are listing are capital costs not operational ones. Broadly speaking for most uses we appear to have crossed the threshold where the total cost of ownership is lower for most tasks, but for some niches (like "ice road shipping") I doubt the buildout is worth it (yet).
To attach some rough numbers, TCO of PRC electric truck (which cost 2x diesel) went from paying for itself in 4-5 years at $60 barrels to 2 years at $100. Diesel increase to $150, it pays for itself in 1 year.
OK, then can you name one deployed fleet of trucks anywhere that uses swappable batteries?
According to an unreliable source that gives fast answers to my questions, U.S. freight companies spent approximately $32 billion to $36 billion on new diesel Class 8 trucks in 2025.
Now are we to believe that these companies and their investors are foolish? That they didn't do calculations and consult experts before spending this money?
Are we to assign more weight to comments here on HN assuring us that electric trucks are cheaper in total cost of ownership than diesel trucks? -- comments that cost the writers nothing but a few minutes of time?
Countries dependent on the Persian Gulf's remaining open to international shipping trade shouldn't just blindly copy U.S. freight companies here: for those countries, any extra cost for an electric fleet might be worth the peace of mind of knowing they will always be able to deliver food, medicine and other essentials to their populations. France for example takes all aspects of its national security seriously and relies almost completely on imports for any fossil fuels it uses. In response it is electrifying as much of its economy as practical (and continuing to invest heavily in nuclear electricity production and renewables).
Many major close loop operations, i.e. mines, heavy industrial clusters, ports where trucks stay on location with ~100% utilization rates have been electric for a few years now, trial started in ~2020. Started with something like 10 pilot cities, now standardized around CATL #75 pack and been mass rollout last few years, there are literally 1,000s of fleets running on battery swap now. Goal is something like 80% of highway freight done by swap stations by 2030.
>Now are we to believe that these companies and their investors are foolish? That they didn't do calculations and consult experts before spending this money?
Or you know smart investors/planners making peace with stupid US energy policy the precludes freight electrification which is vastly more economical if there was state capacity to deploy it economically.
> maturing technology for swapping batteries as a refuelling strategy
This seems like a non-problem to begin with. There are electric semis with a 500 mile range, which at 60 MPH is over 8 hours of driving, i.e. the legal maximum in most places. The same trucks can also add 300 miles of range in 30 minutes, which adds five hours of driving in the time it takes for a typical lunch break. Why do you even need to swap the batteries?
The nitrogen comes from the air - we're perfectly capable of capturing it using renewables.
It's probably one of the last things to be created that way because it's one of the places where methane is used more efficiently than burning it... But fundamentally there's no issue here except energy availability and a short term supply shock.
While there are requirements on making copies in the FDL, I think it is extraordinarily unlikely that a court would find that a company making internal copies would violate the license when those restrictions are just along the lines of "and you must include a copy of the license".
And the FSF would be extremely foolish to ever pursue such a suit, because extremely ordinary non-AI related activities involving working with internal local documents also make copies in a similar way. If OpenAI violated the FDL by doing so then the FDL is a foot gun of a license that companies would be well advised to avoid.
The only suit that makes any sense would be the one against using the FDL licensed documents to train the not-FDL licensed AI... and the judge already rejected that in this case.
The requirements being easy to meet doesn't absolve someone from having to follow them. And the FSF clearly says they aren't pursuing a suit in this case. I'm not suggesting that filing suit here would make sense, rather, that the FSF does care deeply and fundamentally about requiring requiring copyleft restrictions -- they vehemently don't tolerate permissive use under their licenses.
> If OpenAI violated the FDL by doing so then the FDL is a foot gun of a license that companies would be well advised to avoid.
That has been said about a lot of FSFs licenses, and in fact, many companies do avoid them.
> The requirements being easy to meet doesn't absolve someone from having to follow them.
Everybody ignores some GPL requirements. For example the following one:
> a) You must cause the modified files to carry prominent notices
> stating that you changed the files and the date of any change.
Look at any GPL project. Do they have "prominent notices" in all files modified by someone who is not the initial author? For every person who modified them?
How long would be the list of dates alone for files that are often modified?
Another great example of a footgun. Those terms did make more sense in the days before source control when it was common practice to do that anyway, and in many older projects you will find that they were followed.
Ultimately, the terms of the license do matter if someone wants to enforce them.
No, even under such a restrictive license (which I think the GFDL is?) there's no argument.
Their copyright was not violated by anthropic downloading the books, because anthropic had a license to do that.
And their copyright was not violated by anthropic training on the books, because the court found that no ones copyright was violated by doing this. Antrhopic didn't need a license to do this. So the restrictive terms of the license can't prevent it.
I mean they might have an argument for compensation based on "well the settlement Anthropic agreed to didn't exclude us even though they didn't violate our copyright"... but just for the compensation outlined in the settlement.
Nuclear has the same issue as (unpredictable) renewables, it is incapable of cost efficiently following the demand curve. As a result, just like renewables, it requires a form of dispatch-able power to complement it (gas, batteries, etc). Solar and nuclear fill the exact same role in a balanced grid - cheap non-dispatchable power.
Or at least nuclear would if it was cheap, but since its costs haven't fallen the same way that the costs of other energy did... well new nuclear buildout really doesn't have a good role at all right now, it's just throwing away money.
Solar and nuclear complement eachother fine - because their shortfalls (darkness for solar, high demand for nuclear) are mostly uncorrelated... a mix of non-dispatcahble power with uncorrelated shortfalls helps minimize the amount of dispatchable power you need... but batteries have made it cheap enough to transform non-dispatchable power to dispatchable power that nuclears high costs really aren't justifiable.
It's not the lack of oil that enabled this. The west* fought a bloody war to defeat North Korea. We just didn't win (though we did prevent the north from taking the south...). Now you've got a dictatorship protected by their ability to deal devastating damage to South Korea via nukes, huge stockpiles of conventional artillery (and Seoul is within range), etc. Moreover one backed by a superpower (China, and before China the soviet union... indeed these countries are the reason the west didn't win the first war as well).
They could have all the oil in the world and we'd be no more in a position to do anything about it.
*US, Uk, Australia, Netherlands, Canada, France, New Zealand, Phillipines, Tukey, Thailand, South Africa, Greece, Belgum, Luxembourg, Ethopia, Columbia, and South Korea.
The US did not win because the US did not win. Crying about the reasons does not help. Usual FAFO. Does not hurt to think of consequences before starting something
South Korea and its allies did not win - but they did successfully defeat the North Korean invasion of South Korea that started the war. Resulting in 53 million people today who live good lives in a high tech liberal democracy instead of living in abject poverty under the dictatorship that controls the north.
Despite not winning, the consequences of the western nations going to war in this case appear to have been significantly positive. It's really the only war since WWII that I think I can confidently say that about.
No need for political lecture. This was a simple point of win / win not.
>"high tech liberal democracy"
After US involvement South Korea was anything but. It is only since 1987 that some semblance of normalcy had started to appear. Still it is a country practically owned by Chaebols and Hell Joseon work and life culture. Recent temporary martial law with the president's shenanigans does not inspire much confidence either. Call it whatever you want.
When you imply there was "fucking around and finding out", "starting something", and "[negative] consequences" to a war that had positive consequences, and which was a war the other side started, there absolutely is a need to correct that.
Edit: Just noting that at the time I responded the above post consisted entirely of "No need for political lecture. This was a simple point of win / win not". The rest was edited in after the fact.
Yes I edited it later to respond you your point of flourishing society. I should have put it under PS or edited. My fault. The original point still stands.
Edited.
I do understand that comparatively to North Korea SK is of course a huge win for people. However I think they would like to compare their lives with something better than one of the world's poverty and people's abuse champion
This is a dumb ass way to try and define the woman's category... which is about the expected result of bigots trying to work backwards from the result they want headlines about.
reply