sesqu
sesqu on Hacker News
Did you just forget about the mobilization drive Russia had in 2022, where they grabbed young men off streets and from their houses?
It was very unpopular, lead to people fleeing the country, and was pushed out of the public eye as soon as they figured out how to forcefully volunteer people instead.
I'm not sure if I'd go down to zero, but he did get fired from OpenAI for lying.
I know the this isn't the thrust, but the "jagged pac-man" is actually supposed to be a puzzle piece, and symbolizes plugins.
The three dots at the right, for some reason, is the menu. It's labelled "Settings and more", and contains the gear icon you were thinking of, but also all of the normal menu items like new window and print. It also contains "Save page", but befuddlingly that one is hidden under another flyout menu called "More tools".
When opening a PDF file in Edge, it loads up some kind of Acrobat embedding for me, and that one has its very own "save file" button, though with identical iconography to Edge's. Thankfully, the button is visible without any further menu diving, but is also placed at the top right, with an extremely faint border delineating it from the Edge chrome.
Your math is wrong.
5W/user * 200e6 user * 4h/24h = 167 MW
11 MWh/a/home = 0.00125 MW/home
167 MW / (0.00125 MW/home) = 133 000 home
A few weeks after I got my first touchscreen phone, I had to google for how to answer it.
I had assumed you'd just tap the "answer" button, but that failed more often than not. It never would have occurred to me to swipe a minimum of 3cm, starting with the answer button, and I must assume this knowledge has spread to users by osmosis rather than discovery.
That's not so much a problem for the people who want to leave a meeting, but it is very much a problem for people who frequently toggle their mic on and off.
Shannon did not use the word intelligence to describe the mouse in this demonstration - instead, he talked about learning. That's why the second run was considered more important than whatever algorithm was used to solve the maze.
To that end, I'm curious about their cache invalidation solution. Are there timestamps, or is it a flag system?
This was explicitly gone over in the announcement. The short of it is that they validate further attacks by promising Trump's allegiance, assistance, and a target.
Indeed, CDC's numbers for March indicate that social isolation is reducing non-ncovid-19 deaths by twice as much as sars-cov-2 is adding them. That's short term, so not what GP was talking about, but very significant numbers nonetheless.
Weight is a weaker condition, since you can construct a polynomial weight sequence that results in linear height.
In general, height is the easiest thing to restrict, but doing so restricts dynamic performance optimizations - you can't use splay trees, for instance
Log-height is a sufficient but not necessary condition. There are a few other ways generalize balanced trees, notably logit-weight.
Fixing the percentile seems akin to the gambler's fallacy: since losses below a certain point don't matter, you just have to bet everything when you fall off, like you would if you thought after a big loss streak you were "owed" a win.
Limiting their size would cut down on the clickjacking annotations, but wouldn't deal with the hated "buy our merch!" annotations that were damaging to the video content and led to people turning annotations off. I think simply removing all links would be clear and effective, without damaging the error-correction use.
The third common use that I've seen is linking to other videos, and that's the one that YouTube's replacements systems tackle.
Another somewhat noteworthy thing is that the annotation system is tied to YouTube's channel logo overlay. Once the annotations go, it's likely that the logo will not be hideable anymore.
My understanding of the argument is that you have to assume you keep playing after you lose, so your stake is dictated by the risk. For the example in the article, the potential upside is 110% profit and the potential downside is 100% loss, so the optimal stake is relatively conservative in order to prevent any loss from impacting your future ability to invest.
Imagine you get the expected result of one win and one loss and had staked 5%: you'll end up with 1.055×0.95=1.002 wealth. If you had picked a 50% stake, you'd be sitting at 1.55×0.5=0.775 wealth instead, since the loss more than erases your winnings.
The unstated contrast is to models where e.g. there are only a limited number of investment opportunities, losses aren't total, or there are capital infusions.
Right. I was making the comparison that mobile games are like horse races, and WoW was like payday loans.
Now, I don't really have a solid base of experience to form these generalizations with, so a lot of it is just being judgmental. But I want to make the point that while the endorphin loop of loot boxes (or worse, time boosts) is rightly condemned, the WoW model of milking your audience for literally billions was - aside from revolutionary - malignant. It wasn't just the good old days, it was a cash cow that consumed Blizzard for a decade until they too switched to the loot box model.
It was definitely less abusive, but I think it was more malevolent - like payday loans instead of horse racing. Both prey on the destitute, but one builds itself into your life (like a need), the other is merely novelty (like opportunity).
WoW had an absurdly high recurring fee that it tried to make you renew again and again and again. Because there was no micropayment infrastructure, it didn't use "pay to throw dice", it used "pay to retain your investment". By contrast, many of the modern games don't do that and are perfectly happy with their high turnover rates.
The article does briefly discuss a slightly more accurate version of that definition, and why it's bad (it's convoluted and relies on the poorly-defined kilogram).
The cloud isn't there to meet processing needs, it's there to meet tech support needs. Everyone gets handed the same keys, so if something's wonky you just get a replacement. No debugging and repair costs, no standards adherence requirements, no getting shoved out of the software layer by a competitor - and the price is just the uncool segment of the early adopters.
That depends on what you use Office for - which is why I qualified my statement. I wouldn't recommend them as general replacements, but they exceed my documentation needs.
And if you're willing to consider LaTeX, I'd definitely recommend LyX.
The claim is that there may be a market for AI-powered technologies, but that market is suppressed by non-AI competitors exploiting information asymmetry.
I paid attention to the part of the article that mentioned that the pizza slice comes packaged with a protein shake. That sounds just like veggie crackers to me.
> as a dev
I've had a number of devs, when receiving feedback, tell me that I should instead seek out their bug tracker, register with their feedback system, open a support ticket, give detailed instructions about how to identify and fix my problem, and then wait for a decision about how I'm wrong.
I've done that maybe twice.
R and Matlab are definitely array-oriented in that sense (though not fully). Here's how you might do it in R:
prod(1:x)
That screen is mounted high in Uber's cars. The driver was looking low.
I'd say I'm an evening person, but I like waking up before the sunrise.
I suppose that makes me also a night person. I'm constantly either a morning person or an evening person. Quiet daytime is nice too, sometimes.
