There's a big difference between utility classes which are shorthand for inline styles (random example from Tailwind's site where every single class is a one-to-one mapping with a single style override.):
A human consumes about 100 watts when not doing any physical exertion (round number, rule of thumb). So unless you can show an LLM running on 100w compute with capabilities similar to a human, they’re less efficient.
100 W is only the start. Let's say that I consume 100 W all along the day. I use an LLM for coding assistance in the old way of asking questions and copy pasting code. It's much faster than me at writing that code. I don't think it ever works 1 hour for me per day. It's probably 10 cumulative minutes, probably much less. Round it up to 12 minutes to make it 1/5 of a hour or round it down to 6 minutes for a 1/10. So instead of 24 it's 0.1 hours, 240 times less. Those 100 W could be 24000 W and the total power per day would be the same. Is that LLM consuming 24 kW when working for me? No idea but I hope it's less than that.
Of course I could do all of my coding alone again, but I would be slower. It's like walking to the mall several times per week, several hours per time, instead of once or twice per week with a car, three cumulative hours. I trade a higher energy consumption for more time to do other things and the ability to live far away from shops.
Right, but you do stuff other than work 24 hours a day, right? You have fun, relax, etc.
Counting the 100w for 24 hours for a human doesn’t match up with counting the power usage from “AI” for only the 10 minutes it’s doing a task.
Also - units issue: 100 watts for a day is 2400 watt-hours. It’s a moot point anyway because the power draw for the frontier models is an order of magnitude off that the division by 24 is basically meaningless.
Wouldn't attention to getenv() calls yield more benefit? Such calls are where input typically isn't parsed--because parsing is "hard"--becoming targets for exploit.
The present fix is to sanitize user input. Does it cover all cases?
Publisher is the equivalent of InDesign. It was meant for brochures and so on. If you want to write a long technical manual today most people use Word. In that respect we are using less powerful software today than our grandparents.
Note: Adobe bought FrameMaker and continues to sell FrameMaker. But Word has captured the market not because of its technical merit but because of bundling.
I have never written any technical manuals, but I'm surprised that Word is the choice of tool. How does one embed e.g. code easily in the document? I feel there must be a better way to do it, maybe some kind of markdown syntax? Latex?
> How does one embed e.g. code easily in the document?
You don't. For APIs and such, documentation is published online, and you don't need Word for that. Word is used in some industries, where printed manual is needed.
What about the printed manuals? I think they still have some of those not too long ago (e.g. Intel manuals). What was the tool chosen? Very curious to know.
Or, maybe a legacy example -- how were the printed manuals of Microsoft C 6.0 written? That was in the early 90s I think.
Yeah I agreed. Kinda missed the old days with thick manuals. I bought one for gdb a couple of years ago and love it -- despite it is just the paper version of the online one.
>Which banks do you use? I’m looking to switch away from Chase (which does this).
Do you mean SMS codes or a Chase Bank App?
I have to deal with the former because I auto-delete cookies when I close tabs and use Multi-account containers on Firefox.
I've never been required to install any application (Chase branded or otherwise) on my phone in order to use the Chase website. I'll note that I've been a Chase customer since they acquired Chemical Bank in 1996.
Am I missing something important here? If so, I'd love to hear about it.
They're all going to move that way - it's sort of fundamental to PassKey. It can be done with just a laptop and their built in hardware but I suspect that since everybody has a mobile phone the UX will be built around that more often than not.
I quite like it though. At one of my banks I don't even use a password. My browser has the right material (from a prior authn) and then it pushes a validation request to my phone and with FaceID I'm in.
Within the EU, there is a law that mandates accessibility without a smartphone. The banks will sell you some proprietary dotcode scanners then which are all manufactured by the same crappy UK company (as a sidenote).
But the upside is: they work offline, and makes your 2FA app unhackable because it's not an app and instead a physically separate device.
If you're as serious about your opsec as I am, I heavily recommend to not use apps on smartphones for banking.
It absolutely did not. The RCS of an F-14 v/s an Airbus A300 is an order of magnitude different (probably 2 or 3 orders).
> There really was an F-14, just on the ground at an Iranian airbase
There was, but that’s a red herring for the root cause. Each ship’s radar independently and correctly identified and tracked the Airbus separate from the Mode 2 targets, but when communicating the track information between ships, the tracks were mixed up.
> There was a combat camera team aboard the Vincennes, and the footage depicts considerable confusion and even ill-discipline amongst the crew (cheering, shouting, football game atmosphere) that contributed to one of the most tragic events in U.S. Navy history
The point is this is a fog of war situation. Mixing up who is who in combat is a very real issue where we have gotten better over time but have never truly solved.
I put the primary blame on Iran because they cleared a civilian plane to overfly combat they initiated. They set the situation up, a mistake happened.
Fog of war isn't like in a video game where it's just whether you can see something--in the real world by far the biggest factor is identifying what you see rather than simply seeing it.
The URL you linked to results in a 503 error (Service unavailable) and the Wayback Machine returns "Error code: 403 Forbidden" with "Looks like there’s a problem with this site", for all timestamps I tried, in 2025 or 2024.
I'm outside the US so that's probably the cause. Is such information available elsewhere?
Not saying it’s good/bad, but it feels like that’s the use case
reply