Hacker News new | past | comments | ask | show | jobs | submit | tonmoy's comments login

Yes, but I guess the model is optimized for relatively quick response, whereas these techniques are allowing the model to spend more time to generate a higher quality response

I don’t see the issue in that though. If the target was to keep the average emission down across the entire country and if inefficient brand A decided to merge with efficient brand B to keep the average down that seems like it still adheres to the spirit of the law

Seems more like it meets the letter of the law.

The spirit was surely be too accelerate efficiency by ensuring all manufacturers improve. That has been negated; reducing the necessary efficiency for some manufacturers just because others are doing well.

It's like if you allowed multiple people to mix blood samples for a DUI check. Sure, there'd have to be less drinking over all, but some would still be drunk af and the effectiveness of the law would be greatly reduced.


Not a great analogy. CO2 emissions are a global phenomenon, so the average emission level is exactly what matters. Drunkenness is not.

Reasonable criticism of the analogy, but the point was/is that this chicanery weakened the effect of the legislation allowing more CO² than would otherwise have been the case. Thus, the spirit of the law was not respected.

Yeah it's not really "gaming" anything.

Compilers atrophied our skills of writing assembly, calculators atrophied our skills of doing arithmetic and search engines atrophied our skills of recalling random facts, but that enabled us to gain skills in other areas

Given that those parameters are for wheels on a plane that doesn’t have wheels, I would say fixing the parser is the better fix

It is, but who enjoys fixing parser issues...

Even without the recent MAGA wave, Americans on both sides of the party line have always thought that. Even in the corporate world the domestic market is always the primary target and international is an afterthought


I don’t know where you live, but I charge my EV using two phase 240V outlet (same as the ones used by washing machine) and it takes about 8 hours to charge from 30% to full, which is more than enough for 99% of my use cases


Unfortunately, a lot of people here who live in apartments have to wait for any electricity to be added to the parking lot at all, even 120V is generally unavailable. It's definitely restricting adoption.


Yes, but also - people who live in apartments and condos are the most likely not to even own a car in the first place, and if they do, drive it the least.

The people who are doing the most driving are the suburban folks living in houses with big yards and driving 30 miles into work every day.

And that's a larger chunk of the total population in the US than the apartment dwellers who own cars.

Your metro may vary.

Rome wasn't built in a day.

Charging will come to newer apartments, and then older apartments.


In my metro there are a lot of apartments in the suburbs that are just as car dependent as any single family house. Some have bus service, but it is the service for those who after 5 DWIs can't get someone else to drive them, not the fast frequent service that makes people willing to take the bus if they have any other option.


Sure, same situation here. Lots of apartments in mostly car dependent areas.

But there's still a lot of single-family homes, yes? Like, the majority of people living in those areas are in single-family homes? So probably the majority of car owners?


30% in my suburb (10 years ago there were farms between it and the city so this is new) live in apartments of some sort. Majority is single family, but still a significant enough number in something else.


So 70% live in probably new build single family construction, likely with 200A electrical service in their garage. Those people can probably easily swap to an EV, outside other complications.

Easily 50%+ of households around you could probably have an EV. And yet people act like it's only some tiny miniscule slice of the population which could easily change to an EV.


Maybe people otherwise like me, but I've long ago realized for most people an EV with decent range (the leaf is not quite there) would work for most people. However such do not exist, if you buy new there are few options, if you are drive used cars you are stuck with what you can find.


So 70% - and this is a very atypical suburb.


It doesn't sound incredibly atypical. It actually sounds pretty typical to the areas I know, which are several of the largest metro areas in the US.


The atypical part is only that this is an all new one - suburbs have existed for more than 100 years now, so there are a lot of older houses in the suburbs, many of them don't have 200 amp service (though the owners are upgrading them as they as things like AC that need more power and were not considered in the original design)


Sure, there are suburbs that have existed for a long time and are still going strong. And a lot of 100 year old suburbs that became financially unviable and are now empty or torn down.

But large chunks of some of the largest metro areas are made up of massive suburbs largely built from the 2000s onwards. I'm talking Houston, Phoenix, DFW, Austin, Atlanta, Denver, San Antonio, Kansas City, Oklahoma City, Tucson, Knoxville, and similar areas. And that's just areas I've visited.

Sure, there are 100+ year old neighborhoods in Dallas. But around me, some homes were made in the 50s, a big spurt were made in the late 70s and early 80s, then a massive amount were built in the 00's to literally still being constructed. Massive neighborhoods, apartment complexes, shopping centers, and more exist where there was farmland just five years ago. This makes up a large percentage of the growth of this city. It's not like they're building 100 year old houses for the 30% of growth over the last 25 years.

Go take a look at satellite images of Prosper TX. Of Princeton TX. McKinney TX. Allen TX. Rockwall TX. Dickinson TX.

Where I grew up in Houston in the late 80s/early 90s, you practically couldn't find a house that was built before 1975 other than the derelict remains of old farm houses. The house I grew up in was built in '93 and was kind of out there at the time (had to take a dirt road to get to the neighborhood!) And there's been a ton of construction since then, going further and further into the swamps. Woods where I grew up playing in are now full of single family homes.

So no, the sprawl of suburbs continuously marching more and more into the country side isn't a new thing and it isn't something rare in the slightest. It's happening practically everywhere that's growing in the US and has been for 30+ years.


Technically not two-phase, it is just both sides of the single phase.

https://www.youtube.com/watch?v=jMmUoZh3Hq4


> Many other countries, including EU members and China, already apply tariffs and non-tariff barriers far in excess of what the US is applying now.

Source?


For people who want to have a professional social presence (FB/linkedin) as well as an anonymous one (Reddit etc), it’ll be super useful to see if the accounts are truly unlinkable. Moreover if you are opening a new anonymous account, maybe a good idea to search the new username using this tool to make sure it’s not “taken”


Until some ML process is learned to give a probability that accounts are the same based on writing styles

Staying anonymous is very difficult


Stylometry tools may be useful if you already have a small candidate pool of suspected aliases. They produce too many false positives to be useful for blind cross-linking of accounts. Once or twice somebody has done stylometric analysis of HN accounts and I've looked at the results for my accounts. Even though I don't try to obscure style across accounts, stylometry didn't match my actual accounts with each other. My top matches were for accounts controlled by other people.


I specifically write with different perspectives, tones, and opinions on different sites in a probably vain attempt to mitigate this.

For example, on YouTube I use twitch slang, and on Reddit I use TikTok slang, and on TikTok I use reddit slang. On hackernews a use a slightly whimsical pedantically-infused undergrad tone.


You really care about this and use most privacy invasive platforms at same time? Sounds like interesting acrobatics to me.


Using stats this is called stylometry and I agree this will probably be easier at scale now. You can also match posting windows, pull additional features from database dumps/hacks.

Fun post applying it to HN, not sure if the site is still live: https://news.ycombinator.com/item?id=33755016


Then people will start using browser extensions that automatically "fuzz" your writing style randomly. That is, if chasing anonymity is someone's true goal.


> Hard fact is that Trump presidency delivered actual wage growth for people in many years

Do you have a source for this claim? Covid did end up causing salary increase I know, but we can hardly attribute that to the president


BLS data is public and can be crunched from here https://www.bls.gov, but posing the basic question to ChatGPT:

Wage Growth:

Obama: Modest but steady wage growth, particularly from 2015 onwards. Real wage growth averaged 0.5%–1.0%.

Trump: Stronger wage growth pre-pandemic, with real wage growth averaging 1.0%–1.5%, but the pandemic and high inflation in 2021 dampened these gains.

Inflation:

Obama: Inflation was relatively low and stable, averaging around 1.3%–1.5%.

Trump: Inflation was low until 2021, when it surged to 5.4%, outpacing wage growth.

Real Wage Growth vs. Inflation:

Obama: In the later years of his presidency, wage growth generally outpaced inflation (2015–2016).

Trump: Wage growth outpaced inflation until 2021, when inflation surged and surpassed wage increases.


I don’t think ChatGPT should be cited as source


I cited the actual source. You can do the analysis yourself to verify. I'm sorry I don't have time to manually distill it further for you.

That said, I also don't think my specific claim is commonly disputed by the other side.


The BLS data seem to disagree with ChatGPT; It states that average wage growth for both Obama and Trump (pre pandemic) were similar at around 3% and inflation for Obama was around 1-2% and Trump (pre pandemic) 2-3%


There are plenty of south/East Asian counties that have just as much if not more restrictions in textbooks. So I don’t think that alone accounts for the discrepancy


Join us for AI Startup School this June 16-17 in San Francisco!

Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: