I have brown eyes, still hurts to drive at night. Canadian situation is just as bad as our neighbours down south. Now I long for the days of being blinded by yellow headlamps again.
Remember, "don't support JS" is just a shorthand for a broader variety of situations:
- Old browsers without the modern JS features you're trying to use
- Different browsers, with different features
- JS didn't load due to network errors
This is so much worse now that crossovers are the default car in the US. I seem to be the last person in America driving a car that isn't a tank and have never understood why everyone loves those massive machines so much.
Arch Linux added php-legacy and php-fpm-legacy, which tracks (I think) the oldest supported PHP release, a couple of years ago. After I switched to that it's been pretty smooth sailing.
It doesn't help your case when you state inflammatory remarks like "gender-affirming hood heights". This isn't reddit.
The data even points to the fact that, by total vehicles vs vehicles that cause pedestrian deaths, regular passenger cars cause 19.9 pedestrian deaths per 1MM registered vehicles while trucks, as and entire category, cause 19.2 pedestrians deaths per 1MM registered vehicles.
"nothing to regulate" is also an exaggeration. Many states to regulate aftermarket lifts. 6" lifts are typically the maximum legally allowed limit for trucks like the F150. You only see them higher because there is no enforcement of the rule.
It's ridiculous to consider MITM attacks out of scope for taking over your computer. Also, there are probably ways to exploit this without a true MITM like DNS cache poisoning. But it's best to just assume the whole internet is MITMed.
It doesn't make sense to include the capex cost to train a model in this kind of discussion, because that cost is fixed.
Consider a model that costs $100m to train.
If the vendor then prices it such that each inference token has a margin of 10% over the variable costs to serve (power + server costs), whether or not they cover their costs is based entirely on how many tokens they can sell.
If they sell less than $1bn of tokens, they lose money - the break even point is 10x100m = $1bn.
If they sell $10bn of tokens they make a ton of money.
This also means you can't credibly calculate how much of the fixed training expense is covered by your token spend, because until the model is retired and you can account for how much inference it ran you don't know what percentage of the training cost each sold token was responsible for.
This is not exactly what you’re asking about, but I started learning Japanese when I was in the middle of playing Cyberpunk 2077 (for unrelated reasons); and I gotta tell you; realizing that 98% of the Japanese text everywhere in the game was just “hotel” or “karaoke” definitely took away some charm from it.
> Why? In the end it's a patch's quality that counts.
LLM patches tend to be significantly harder to review. Mostly because LLMs let people who don't know what they are doing get much further.
It might be an unfair heurestic as there are plenty of competent people who use it to good effect, but the vast majority of negative value patches use LLMs and it can be a bit exhausting. Lowering the technical barriers of entry just means more pressure on the human ones.
It can be a one way ratchet, which is my argument here. Regulations, medicare rates, and pricing power from drug companies set a floor for prices. Insurance companies already run pretty lean and have so-so margins. The 80-20 rule incentivizes a slow ratchet up on prices and the other factors prevent competition on price. The system is designed to have slow predictable price increases.
Small successful companies are great, but the hardest thing for me psychologically has been when I'm at a small company that is struggling to convince anyone to use its product. Being a small cog in a giant machine serving lots of users is more satisfying (to me) than building things that nobody is using.
The other issue is the few icons that existed before Tahoe generally highlighted import or more frequently used actions. And acted as a visual anchor point so if you frequently used another item you know, "it's 2 items below the save which has the save icon"
When everything has an icon it all just becomes clutter and no one is going to waste their time trying to parse tiny glyphs that are inconsistent between apps anyway. You could try to figure out which of the 4 icons with a plus button is the "New X" option you're looking for, but you'll probably just be reading the titles anyway so the icon does nothing.
Pretty neat that you can just install it and start using it (at a Sonnet 4.6-level model) without needing to sign in or pay.
Typically, Chinese websites are a big pain to log in or sign up because they require a +86 phone number due to legal reasons. Being able to use it without having to make an account is amazing for friction reduction. I could probably even just install it onto new machines to help with set up.
I wonder how they are gonna detect and block abuse though?
As stated on the project page, by design, this is not meant to be a full-featured X11 server.
As such it will have it's own use cases, so I wanted to mention that there is a project that is a fork of X11 called XLibre.
It is, by design, meant to be a drop in replacement that will perpetuate X11 as Wayland starts to become the default.
The current release of Ubuntu 26.04 is the last version that will use X11. The next major release will use wayland.
Personally, I think it's a mistake to just write of X11 in favor of Wayland because they are not compatible in fundamental way, so I'm always glad to see projects like this and XLibre.
There are (or have been) lots of languages using the JVM as a compilation target, whether it is well-suited for this or not. Wikipedia has a partial list: https://en.wikipedia.org/wiki/List_of_JVM_languages
anecdotal but a good number of people have clearly stopped working as much or quiet quit. some even left the company or the industry. CEOs are also telling everyone to 5x their output, AI will replace them.
Whatever productivity gains models are giving us is being eaten away by other factors.
Yeah, I think this will really be a thing. I have been using `exe.dev` lately, and I was at an AirBnB with my family where we wanted to play a game, but it needed pencil and paper for each person, but we didn't have those. I thought it would be nice to have a little web app where you could write your input and then vote on other people's responses directly from your phone. I just spun up a VM on exe.dev, asked the AI to write the app, and then made the VM public-facing. It was a little buggy, but it worked well enough and we had a fun time.
I heard in Germany when a vehicle is being inspected (yearly?) the headlights' angles are checked to not be beaming into oncoming traffic. Feels like useful regulation to include in every country.
