HNHacker News
TopNewBestAskShowJobs

pwinnski

8,760 karma · joined June 28, 2016

submissionscomments
pwinnski··on Who Blew Up Nord Stream? Investigators Focus on Mysterious Passengers on a Yacht
Hi, I wrote the comment that you seem to be responding to, a comment that includes the phrase "Not saying the US didn't do it," so I'm not sure how that qualifies as "dismissed out of hand."

My point was that Hersh's article very strenuously avoids anything that makes Russia out to be anything other than an innocent victim, lining up exactly with what Russian ministers had already said in public. That is why I mentioned propaganda, it's hard to describe it any other way.

The US is full of propaganda, too, if that helps. We've got 24-hour news channels that pretend to be in opposition to each other but seem to somehow "report" in lock-step on a variety of subjects that exactly mirror official government statements, mostly involving the military.

The US might have done this. Russian might have done this. Someone else might have done this. But Hersh's article is poor reporting, and not something a determination should be based on. Which is too bad! Hersh used to be amazing. Not sure why he so missed the mark here.

pwinnski··on Who Blew Up Nord Stream? Investigators Focus on Mysterious Passengers on a Yacht
There was no big gas leak, the pipeline had already been shut down by the Russians a week or so prior.
pwinnski··on Who Blew Up Nord Stream? Investigators Focus on Mysterious Passengers on a Yacht
It is indeed sad how Hersh started out so strongly and went so bizarrely off the rails in his later years. I want to believe it was a natural progression, that having seen the great evils the US was capable of, he start to see US evil everywhere he looked, and was easily taken in by people with a story that appealed to his priors.

But I don't know, he so studiously avoids so many things that indicate Russian malfeasance that it's hard not to wonder if its more than just hope the pattern will hold up.

The US is as shady as the day is long, and I wouldn't put this past them if it made any sense, but if you know anything outside of Hersh's article, it doesn't. I fully believe the US is doing much shadier stuff elsewhere, but the article doesn't make a good case.

https://newpol.org/the-nord-stream-pipeline-explosions-chall...

pwinnski··on Who Blew Up Nord Stream? Investigators Focus on Mysterious Passengers on a Yacht
There is debate and uncertainty online. I doubt there's any debate or uncertainty among the heads of state being sent the message.
pwinnski··on Who Blew Up Nord Stream? Investigators Focus on Mysterious Passengers on a Yacht
Most things, but mostly by omission. For some examples: https://newpol.org/the-nord-stream-pipeline-explosions-chall...
pwinnski··on Who Blew Up Nord Stream? Investigators Focus on Mysterious Passengers on a Yacht
Nothing about Hersh's story makes any sense at all. Unfortunately, while Hersh developed a strong reputation decades ago, that story reads like Russian propaganda. Embarrassingly so[0].

Not saying the US didn't do it, but that theory makes less sense than any of a few other actors.

0. https://newpol.org/the-nord-stream-pipeline-explosions-chall...

pwinnski··on Who Blew Up Nord Stream? Investigators Focus on Mysterious Passengers on a Yacht
How deep do they need to go? How close do they need to be? Let's not make this more complicated than it needs to be.

Since we're well into speculation mode here, try this: footage of the aftermath[0] reveals an enormous 50-meter hole in the pipe, which was made of pretty thick metal. I suppose it would have to be, to carry gas under pressure from within as well as withstand being kept empty, as it was at the time of the explosion, under pressure from without.

How big does an explosive have to be to cause that much damage? Pretty big, I think. Big and heavy.

How close does an explosive that big have to be? Not very close, I think.

Of course, there's a trade-off! The close it is, the less big it has to be and the farther it is, the smaller it can be. But a 50-meter is much bigger than it needed to be to disable the pipeline. It is almost as if the people dropping the bomb couldn't be sure how close they would get with it, so built a big enough bomb to ensure rupturing it whether it landed directly on top of it or a number of meters away. Maybe they got lucky, and expected a smaller hole. Maybe they got unlucky and expected a bigger one.

I mean, the BBC footage involved zero people getting in the water. They just went to the area and dropped in a camera on a long cable and navigated to the hole.

It's possible that nobody even got in the water for the bombing. They could have used a "fish finder" to make sure they were pretty close to the pipe, and even drop a test object to see how the underwater currents might shift it, but it was a pretty big bomb, I think it probably mostly went straight down. Close enough for a big bomb to make a hole in a big pipe, anyway.

0. https://www.bbc.com/news/world-europe-63297085

pwinnski··on First Republic Bank Executives Sold $12M in Stock in Months Before Crash
I think looking into what their selling pattern was like in 2021 makes sense. If it was similar, then yes, there's nothing to see here. Worth checking, though.
pwinnski··on Epic Games to pay $245M for tricking users into making unwanted charges
I'm sorry, I didn't realize that when I said "all" I should clarify that I meant "all platforms that provide a similar level of services, including subscriptions on behalf of third parties for which payment is made centrally."

RockeyMoney is great, but not only is it irrelevant in the extreme to the original comment at the root of this thread, it's very much an apples-to-guavas comparison with the Google/Apple/Epic/Valve/etc model of operation.

pwinnski··on Midjourney v5 can do hands
Because I'm not sure counting is something that has been explicitly tagged or trained for. Clearly always-accurate counts are not an emergent feature, but if you tagged a large set of images with tags we think are too obvious to tag as humans, like "ten fingers" and so on, if that was an actual goal to be rewarded, I think it could improve results.

I'm overall in the skeptic camp, but it seems like these models can generally deliver what they're trained for. It doesn't appear any have been trained with counting as a primary goal.

pwinnski··on Midjourney v5 can do hands
Weird, I linked directly to the image as a PNG file, and checked it on my computer, it is just the image, nothing else.

Based on your comment, I just tried clicking on my phone instead, and sure enough, imgur intercepts and redirects to imgur.io on mobile. But it still didn't break my back button on iOS, so I'm not sure what's different for you.

pwinnski··on Epic Games to pay $245M for tricking users into making unwanted charges
Anyone can launch a system to manage subscriptions on any platform other than Apple's, and yet somehow all of the systems people have launched on every platform other than Apple's are worse than Apple's. In Epic's case, illegally so.

I responded to a comment from someone who apparently didn't realize that Apple's subscription system makes it exceptionally easy to cancel subscriptions.

They said:

> while Epic uses predatory tactics to trap people into subscriptions Apple and Google do the same

Which is as far from my experience as it is possible to get.

The comments from people stating that Apple exercises platform lock-in seem roughly as relevant as the current price of tea at a Foxconn factory in China.

pwinnski··on Midjourney v5 can do hands
I suspect inability to count is a thing that can be trained for, although whether that will come at the expense of something else, I'm not sure.

I appreciate the humor. :)

pwinnski··on Midjourney v5 can do hands
Neither of them produced a result I could use. One or the other might be more aesthetically pleasing to you depending on what you focus on, but both are very badly broken in fundamental ways.

If anything, I'd say v5 is more confident about its wrongness. It is as if humans have always had four fingers, how could you possible think otherwise? It that sense it seems more like the text-based LLMs: confidently incorrect.

pwinnski··on Midjourney v5 can do hands
Let's try "chinese girl making heart with fingers --v 5", a prompt I had used before under v4. Result: https://i.imgur.com/jPgBqsX.png

In image 1, the subject appears to be missing a digit on each hand.

In image 2, the missing digits issue expands, combined with what appears to be a merging of the thumbs.

Image 3 introduces a supernumerary digit on each hand, and has extra "parts" disappear into and out of other fingers.

Image 4... isn't right in a number of ways, but still seems to have fewer than the expected number of digits.

I don't think these results are any better than the v4 model was, but decide for yourself. This is what I got with the same prompt using -v 4 on March 1. https://i.imgur.com/hQA2K7m.png I ended up just taking a photo of my daughter and blurring the background.

pwinnski··on Midjourney v5 can do hands
I tried the same prompt but with "--v 5" instead, and got this: https://i.imgur.com/eElOkjU.png

I only see "stuff on the face" in the third image, which I guess is an improvement. I'm not sure I'd call hands "fixed" based on this image alone, but they're better.

pwinnski··on Midjourney v5 can do hands
It's easy to prompt for people with closed mouths and hands not visible, but with MidJourney at least, I would consistently get what I can only describe as "stuff on their face" with almost every prompt involving a human. Less often with white people, but even then pretty often.

I mean, I just typed "/imagine asian warrior nun --v 4" into Discord, thinking of Beatrice from the recent Netflix series, and three of the four results show what I'm talking about: https://i.imgur.com/499QCn6.png

pwinnski··on An Uber-like CDN
> With current hardware, it's not a big deal.

You should make this your marketing slogan.

"Try our CDN, with current hardware the gross inefficiencies other CDNs spend a lot of time and effort eliminating are not a big deal."

pwinnski··on Epic Games to pay $245M for tricking users into making unwanted charges
What does that have to do with anything I said? Or is it just a reflex anytime anyone mentions Apple?
pwinnski··on GPT-4
Sometimes it's exponential over very short periods. The fallacy is in thinking that will continue.
pwinnski··on An Uber-like CDN
Three things about "how Uber works":

1. It started out burning huge amounts of investor capital, basically transferring all of that money to drivers and in the form of subsidized rides, to riders.

2. It works today by a combination of the above and underpaying drivers, who generally make less than minimum wage when they manage to cover their completely costs at all. Despite this, prices have gone up dramatically, so that taxis are once again competitive on a price basis.

3. Uber was profitable in 2018, and not before nor since[0]. Deeply, profoundly unprofitable. Why would anyone want to be the Uber of anything?

0. https://www.macrotrends.net/stocks/charts/UBER/uber-technolo...

pwinnski··on Epic Games to pay $245M for tricking users into making unwanted charges
Mostly lawyers.
pwinnski··on Epic Games to pay $245M for tricking users into making unwanted charges
Apple makes it incredibly easy to manage and cancel subscriptions. Would that all companies matched them!

One reason I don't begrudge them their 30% is that I could easily lose that much or more trying to get some vendors to cancel subscriptions if I didn't have the easy options Apple provides. Epic is clearly one of them.

pwinnski··on Certifications create wrong incentives for engineers and recruiters
> The engineer who looks good on paper. This person will simply not list the certification on their profile – they have no reason to!

I don't think I buy this logic. I finished an online course with a good test score, and it prompted me to post the resulting certification on LinkedIn. It took 15 seconds at the moment I was happiest about the accomplishment. If we assume I'm an engineer who looks good on paper, am I going to deliberately eschew that prompt?

Is the argument that as an engineer who looks good on paper, I'm in such heavy demand that I never have a few minutes to kill on LinkedIn, in between declining and ignoring the recruiter come-ons? That logic just doesn't make sense to me.

pwinnski··on Three Million U.S. Households Making over $150k Are Still Renting
Is your argument that landlords are by and large rational actors? Yes, I will concede that landlords are by and large rational actors, although I'm not sure that means as much as you seem to think it does.

A landlord is investing, and usually willing to consider returns over a long term. Landlords I have known are not profitable every year, and almost never from day one on any given property. I have known one landlord--and heard of others--who have seen multiple years of profit on a property get wiped out by a single bad tenant. But that's okay! Most landlords (not the big corporate landlords for which the economics are very different) console themselves in unprofitable times with the knowledge that at least they have the underlying asset, and it should increase in value over time.

> Renters have their rent go up every year. Owners are always ahead after 30+ years.

Look, if the only thing you care about is $X out every month, then sure, fine. Buy a house, don't look back, and whatever you do, don't every tote up every penny you've spent on everything related to your house and chart it against S&P Index returns over the same period of time if you'd rented a smaller apartment instead.

pwinnski··on Three Million U.S. Households Making over $150k Are Still Renting
You only have to move once or twice in 30 years for home ownership to not make a ton of sense. "Constantly" seems like hyperbole.

This also assumes that you aren't facing high maintenance costs. One popular rule of thumb suggests setting side 1% of the total value of the home for maintenance every year, and if the value of the home is rising, that set-aside is also rising. By more or less than rent? Depends on many factors.

pwinnski··on GPT-4
Expecting progress to be linear is a fallacy in thinking.
pwinnski··on Facebook confirms it will drop news sharing in Canada under bill C-18
Lucky Canada!
pwinnski··on Three Million U.S. Households Making over $150k Are Still Renting
Those things are also true here! A "home equity line of credit" or HELOC is very common, especially when interest rates are incredibly low, as they were before last year. Also, most people understand that other than some edge cases like 2008, homes rise in value over time, consistently and dramatically. Most people know of situations where a retired couple bought a house for less than $100k many years ago, and now it's worth five to ten to twenty times that much!

That's about where most people stop thinking. Clearly house values go up, I don't want to end up 30 years from now having paid rent and having nothing to show for it, I'm gonna buy a house. There's even one more factor, which is the US federal tax structure gives you a big break on any interest you pay on a home mortgage, so many people think: I get a tax break on my house payment, but not on rent.

Of course, most people don't put together a spreadsheet to consider how much money that old couple might have had if they had put that $100k into a stock market index fund the same year they bought the house.

Please understand, I'm not saying that buying a house is a bad investment. I'm just saying it's usually not the best investment. I don't think most people over-value buying a home, I think they under-value putting a home-sized chunk of money into an S&P index fund!

pwinnski··on Three Million U.S. Households Making over $150k Are Still Renting
A landlord charges what the market will bear, and a unit sitting empty for a month is expensive, so sometimes they'll charge slightly less. If that doesn't quite cover their own costs, well, they're in this for the long haul, usually, and they'll make it up eventually.

Landlords often don't rent out houses immediately after buying them. They must first either invest in repairs and upgrades to justify a higher rental price, or wait until the housing market itself justifies a higher rental price. For most landlords I've known, their first rental unit involved renting out their own paid-off or almost-paid-off house, and living in a new one they bought. Listing a new unit sometimes means being underwater for the first year or two, but they tell themselves that they're building equity and eventually rents will rise to more than cover their costs--which is eventually true.

← PreviousPage 6 of 34Next →