trowawee

They'll be trash, but after a decade bouncing around startups, that's not exactly a problem unique to LLMs. There's probably going to be more startups with more trash than there used to be, but hey: that's job security.

The interesting thing about art is mostly not its physical existence. The interesting thing about art is that another human being made it to try to express something, in words or colors or film or whatever medium. It's a person trying to show you their interiority, taking something fundamentally unknowable—another living mind—and making it legible to other people. Even when it's art for hire like animation in a commercial, at the end of the day there's some human or humans who put some work in there. LLM-generated art just doesn't have that. There's no interiority to be exposed.

The financials have always been crap, though. A lot of his previous assholery was limited to very online people. Guarantee if you poll 100 people if they've heard of the "Thailand pedo guy thing with Musk", most will have no idea what you're talking about. Now it's on every news channel every night.

Probably worth noting that Pratchett wrote most of the "Adam and the gang" sections of Good Omens, while Gaiman was responsible for the riders and the more "mythological" parts. So I do see a pretty clear throughline from Sourcery to the Pratchett parts of Good Omens.

Hey! I got diagnosed with ADHD last year as an adult and getting diagnosed and medicated has been a massive improvement in my quality of life. If you have the energy to do it, I would strongly recommend it. The medication is great—I have far fewer "stare at the wall, bounce in my seat, get nothing done" days—but I also appreciate having a concrete diagnosis. Even after decades of assuming I had it, it felt good to actually hear it from a doctor for some reason.

I think the author dismissed that too casually. Anything that makes the three less valuable will result in less space around the hoop. That's not necessarily a bad thing if we want to reemphasize the mid-range shot. The only way to tilt the game back that way is to change the math on the current dunks/layups & threes meta.

Yeah, on web with extensive adblocking it's not as miserable (or routing my phone back through my pihole install). But most people aren't adblocking, and the "normal" user experience was jarring.

The Facebook aside was interesting. I've been back on FB periodically this year to try to sell some stuff after years off it and it cannot be overstated how terrible the basic experience of browsing the feed is, at least on mobile. It's a minimum of 3-4 ads per actual post from a human I chose to friend, and the ads are just the stupidest, trashiest stuff imaginable.

All true in the abstract. But doing it the day before an acquisition is always gonna earn you a door knock, even if you're a serial gambler.

"short-dated out-of-the-money call options that cash out a day after a merger/acquisition" is, like, the definition of a suspicious transaction. Somebody's gonna get a knock on the door from the SEC.

Radley Balko has a great overview on the fundamental flaws with ballistics forensics[1]. The tl;dr is that the core claim of ballistics forensic analysis, that it is possible to reliably differentiate between two models of the same gun using just casings and shells fired from those guns, has never been proven to be valid.

[1] https://radleybalko.substack.com/p/devil-in-the-grooves-the-...

Yeah, the “ew, yucky gay people!” asides are both very revealing about the author and fundamentally undermine his argument for the value of libraries.

the above 2 passages cannot be serious at the same time as translations of a literary work

This honestly reads like maybe you don't understand how translation works? I mean, you're clearly coming at this from the standpoint of "I have a culture war axe to grind and I'm gonna grind it", but here goes: languages don't map perfectly into each other. Even languages that exist in the same time, in the same physical locations, where many people are fluent in both languages. There are words in modern Spanish and German that you can't express in a single word in English, and those languages have massive overlap and lots of common roots. There are words in English that mean different things to different people. My 21 year old cousin and my 70 year old father would understand the sentence "I saw Bill Murray on the street today, no cap" in completely different ways.

All of those issues are massively compounded the farther you are from the context of the original text. A 50 year gap between living people in the same country and speaking the same language is enough to make communication confusing; a 3,500 year gap makes it nearly impossible. The article this thread is based on (that you are clearly commenting on without having read) displays 5 different ways that different translators approached the exact same passage in the Iliad over multiple centuries. Lombardo's translation, much like Wilson's, much like any translation, is a work of invention. It has to be, because we aren't Ancient Greeks and we are not operating in the same milieu as the original audience for this work.

You clearly don't like Wilson's work for personal/political reasons, but trying to generalize that to "she's a bad translator" because her translations are different from other translations is a silly complaint that reveals a fundamental shallowness in your understanding of the process of translation.

Right?! This is the dumbest bit from a pretty dumb statement. Politics is just people working out how they're going to interact with each other. There's no way to cordon that off from anything else. You can certainly pretend that those industries don't interact with politics, but that's just another form of politics.

Hard to imagine a sillier complaint about a translation than "it tries to convey an ancient text to the people in the era when it is translated". It's like complaining that cooking a steak changes the nature of the meat. Yes, correct, that's literally the point.

I was just pointing out why your suggestion that hybrid work is a solution was a dumb idea because it's the worst of both worlds, but you appear committed to a willful misreading of my comment, so...have a great day!

Hybrid is the worst of both worlds. You still have to live close enough to the office to get there without an insane commute, so you don't get the "live where you want" benefits of fully remote, but you also have to have enough space in your house to work, so you lose the "home is home and work is work" benefits of working in an office. Plus, unless office and WFH days are rigorously coordinated (which destroys any "flexibility"), you're still going to spend a bunch of time in the office on zoom calls with people who are WFH. It's wild to me that this is being suggested as a compromise. Hybrid just guarantees a baseline level of misery for everyone.

It's quite fast if you've got 240v outlets in your kitchen. Otherwise it's a little slower, altho I do like being able to hit a button and wander away to do something else, which I never feel comfortable doing when something's on the flame.

Even if you throw out SF salaries as a wild outlier, this isn't actually that ridiculous. An average quality mid-career dev (5-10 years exp.) in a second tier market like Chicago/Denver/Austin/Boston can pretty easily make 170-200 in cash. "[A]ll up" is the key here tho; there's a big non-salary component in the US that doesn't exist in Europe. Tack on health insurance and the total cost to the company will easily blow past 250 right there. Plus, you're probably giving them some equity and a yearly bonus.

I'll be honest, it does sometimes blow my mind to see how low salaries are over there when I look at job postings and Who's Hiring and the like. I'm jealous of a lot of things Europeans take for granted, but it's wild to me to see senior positions in major European capitals paying the amount that I made two years out of a bootcamp.

Not really; definitionally an RSS feed is just a big list of everything a site publishes. Sure, the rules-based approach will work, and I assume anyone who's been using RSS for a while has built up a set of filter rules that amount to a personal algorithm for their feed. But it's a big initial investment in time and effort to set those rules up. If you subscribe to a bunch of personal blogs, you're fine. But if you sign up for, say, the NYT, the WSJ, and Bloomberg (not a crazy list) the day you set up an account on Feedbin, you'll have 200+ items by the end of the first day. Are you 100% sure you already knew every rule you want/need to filter those items? If not, are you a fast enough reader or willing to just mark as read and ignore stuff? I think a lot of people aren't, and social media has accustomed people to not thinking about stuff like this.

The firehose issue is definitely a problem with RSS, especially if all you've ever experienced is algorithmic social media feeds. It's solveable, but it does require either a very specific approach (I quickly scan my unreads in Feedbin/Reeder, save whatever I want to read in depth, and then browse saved items later), or a reasonably large suite of rules applied to your feed (takes a long time to build, takes even longer to refine, and you will find yourself perpetually tuning them), or something like Feedly's AI assistant (but then you're still outsourcing your feed to someone else's algorithm, and avoiding that is supposed to be the point of RSS).

On some level, "I want to/am able to wade through the firehose" is a pre-req for RSS. I'm sure that's part of what limited the appeal even in the fabled golden age of Google Reader. Also, like most things, it's better if you pay for it. I've paid for two versions of Reeder now (about $15 total, I believe) and I pay Feedbin $50/year to act as my backend + web interface. The overall experience is better than any of the other modern RSS apps I've tried, and better than Reader was way back when.

Right, the current system is 40% government via bureaucracy, 50% government via lobbyists/industry (including the parts of the bureaucracy that have been fully captured by industry), and 10% "the kind of freaks who run for and win public office" (who I think we should treat with reflexive suspicion on principle, even when I nominally agree with them). It's not at all clear to me that bureaucrats + lottery system would be worse, and it seems likely it'd be better.

You go to a bank and tell it "I would like to buy this company. It has assets worth $N, which could be put up as collateral for a loan of $(.60 x N)." The bank says "Yes, those assets are worth that much. We will sign a contract with you where we will give you $(.60 x N) and in return, once you successfully purchase the company and it is your property, those assets will act as collateral for that loan."

You take the loan (and probably some other money, since by definition the loan is less than the total value of the company's assets), and you go buy the company. You now own the company, therefore you own the assets, therefore you can use them as collateral, and so: the debt that was yours is now on the company's books. You bought out the company with extra leverage provided by the assets you were acquiring: leveraged buyout.

It tends to fail; LBOs tend to target companies that are struggling somehow, and loading an already-struggling company with a shit load of debt frequently results in a bankruptcy. You're probably familiar with companies that dies this way. (My favorite example for fellow millennials is Toys'R'Us.) If you find this troubling: you are correct. But you have forgotten the really important thing: you (or, in most LBOs, a bunch of private equity ghouls) have gotten very wealthy killing a company, and isn't that reward enough?

Yes, that's just for the interest.

You do not seem to know what you're talking about. The 12.7bn that was borrowed is going on Twitter's books. Their debt load prior to this leveraged buyout was 5.5bn; it's now 18.3bn[0], with yearly interest payments of roughly 1bn.

[0] https://www.barrons.com/articles/tesla-stock-twitter-debt-51...

If so, even just moving to on-prem would save them the $1B easily

Except you can't wave a magic wand and shift the entire distributed infrastructure for one of the five most visited sites on earth on-prem. Twitter has two DCs. Running their infra on-prem somewhere would require reworking the entire company from the ground up in a way that would take years and cost substantially more than $1b.

Why does he think anyone would be ashamed of _not_ advertising on Twitter? Twitter ads are worthless. If you're a CMO who manages to kill your entire Twitter ad spend, that should probably go on your resume.

Still doesn't really address the reality that onboarding is dreadful at most startups, I think. Docs tend to be bad/outdated/nonexistent and even "grab an onboarding buddy and talk through it with them" can fail easily, especially if the business is growing quickly. I do think there should be a cliff (maybe six months for seniors/9-12 months mids/juniors) where as the manager, you take a really hard look at how the person has adapted to the job/what they're contributing, while also taking into account their and the company's situations.

Exactly. Of course, none of this is going to come down on the heads of the people in the C-suite who are pulling down the hundreds of millions in comp to prevent, at least in theory, _precisely this scenario_. We've built a system where once you're above a certain point in the hierarchy, it doesn't matter how bad you fuck up; Pichai and Zuckerberg and Hastings et al. can just feed their own employees into the woodchipper and keep making stern noises about how everyone else needs to shape up.

Very good point. His stories about his personal sexual kinks, lightly disguised as news, are probably accurate. Disturbing, but accurate.