It's something that you'd pause and think about if you were trying to carefully craft a sensible argument. But the online attention economy now runs on AI op-eds. In this world of overabundance, the winning strategy is to produce outrageous takes and do it as quickly as possible.
I respect the author, but he succumbed to an... unnecessary temptation. There will be a dozen other "here's what AI means for humanity" posts on HN before the end of the week. For the human-written ones, I doubt that any of the authors will be proud of their analyses five years from now.
Read the article, that's not what this is about. Title is clickbait, they fired contractors for using AI to label data when the whole point of labeling data is to distill human knowledge into weights, not distill weights into weights.
You're replying to a joke, and yes, that's all it's about. For all the tales of rapid self-improvement, all labs are real careful not to taint their precious training sets with anything that comes out of their systems. Website reputation modeling, fingerprints in generated text, firing contractors that rely on AI.
I've been told on HN about a year ago that the era of scraping is over and that it's all AI-training-AI now. My web server logs and stories like that disagree.
There's a difference between shoving LLM output back into the input and RSI more broadly. Using synthetic training data didn't work super well, that would have been one path to RSI, but it doesn't work. The main path towards RSI people are talking about today is using the models to do research and experimentation for training new models. Machine learning is a very empirical field, you need to run tons of experiments, tune hyperparameters, and try different architectures. It's something agents are extremely good at doing, just because one path to RSI doesn't work doesn't mean it won't work at all.
I've been told on HN that 1+1=3, so, yes, skepticism is required on broad claims.
There are more subtle truths on this. Things that can be proven algorithmically are much more apt to be in recursive AI loops now. Hence things like programming and hacking keep improving steadily over time with much less human training data being added.
Think of labeling as establishing a ground truth. It's more important than training, but it's far less complex than training at an individual level. A kid can tell you what an ice cream cone looks like, but they cannot tell you the best algorithm to use to get the best model with the least power usage.
Labeling is more like working a checkout at Walmart. Just about anyone can do it with the smallest amount of training, but you have to ensure your labelers are not just scanning one item multiple times and bagging up the rest as your dataset can skew from reality since AI cannot just capture this data fully reliably at this point (well in many fields it can or can do even better than humans, but it's still lumpy as to where and why).
yeah but most of that is 1) covered by bogleheads, and 2) doesn't require serious education or advisors.
take emergency funds -- how likely is it you can lean on friends or fam for money? do your budgeting and then plan for 6-12 months of budget. if you can get a load from the bank of mom and dad then maybe 3 months.
In my case, I am double-taxed (both Japan and US side) on capital gains. Tax treaties reduce, but not eliminate, the extent of double-taxation.
Many US-based brokers do not allow Americans abroad to purchase mutual funds, so VFIAX is not a choice for me.
Maybe I can go with eMAXIS Slim All Country... Oh, but that is a PFIC under IRS rules and I'd be taxed on unrealized capital gains. So I guess no Japan-equivalents of VT for me. That's fine, I guess I'll just buy VT in my US-based brokerage account; but now I'm in a suboptimal spot with respect to monthly contributions, calculating JPY-denominated income tax on dividends, etc.
I even made an implicit assumption when I said "calculating JPY-denominated income tax on dividends". That assumes your tax status is permanent resident. If your tax status is non-permanent resident, then a decent financial advisor will recognize that only the extent of income remitted to Japan gets taxed, so VT distributing at all isn't an issue (until 5 years later). What should you do before the 5 year threshold is hit? etc. etc.
But yes, if you're born in America and plan to stay within the same state for the rest of your life, then a 100% automated setup that simply deposits $1,000/mo into VFIAX is probably fine. (But keep in mind, to most non-Americans, VOO is not really diversified compared to funds like VT).
Otherwise, there is value in consulting someone (or something) that regularly handles taxes and financial planning.
Full agreement. I'm double-taxed in Germany and the US. I don't have a residency in the US (so I cannot vote but still get taxed, nice!) so I'm completely unable to buy VFIAX or any other fund from the US. I cannot but them over an EU brokerage, either, because US ETF and its ilk violate European transparency laws. I can buy EU-domiciled funds, but those are treated as PFIC in the US with prohibitive taxation. I relegated myself to an old-school dividend growth investment scheme and it's financially working out much better than I anticipated but still, I'd like to have a core of a stupid ACWI tracker.
> I relegated myself to an old-school dividend growth investment scheme and it's financially working out much better than I anticipated but still, I'd like to have a core of a stupid ACWI tracker.
roughly same approach in Canada, for the same reasons.
if you look up most of the dividend aristocrat funds they'll give a breakdown of holdings... so duplicate those in roughly the same ratio (to the best you can) and call it a day
Some people do have complex financial situations. It's not as simple as that.
For example in the UK (and maybe US?) you get tax relief for money you put into your pensions, but there's a limit of £60k/year. Unless you earn a lot (which I do, yeay) when that limit is tapered. Except that you can also use up to 3 years of previously unused allowance. But you have to use this year's first.
Also interest is taxed, but you can put up to £20k/year into an ISA which isn't. And if you still want to avoid some tax you have kids ISA's and even pensions!
Then there are also startup investment schemes that save you some tax. Those seem to be not worth it, but you get the idea - it can be complicated. Especially if you are near one of the many tax/benefit thresholds.
The marginal tax rate in the UK bounces all over the place - it's even technically possible for it to be over 100%!
Though you don’t directly say it, your comment strongly implies that you can’t buy fractional shares of $VFIAX. (You can, same as $FXAIX, $VTSAX, etc.)
It depends though - I'm not in the US and in my country ETFs get taxed continually whereas mutual funds don't get taxed until you sell.
In general, invest in low-cost index funds is pretty solid advice everywhere. In different countries you might use slightly different instruments due to tax advantages (like the ISA in the UK etc.)
I don't have the time to review the underlying research and decide which one is more correct. My personal biases make me want to believe the current one. Your personal biases may be pulling you in the other direction. How do we make the conversation more intelligent than that?
Strongly disagree on that last part. Unlike software engineering, where training is nearly non-existent and we expect everyone to show to their first job interview knowing about the big-O complexity of sorting algorithms, traditional manufacturing has a robust tradition of apprenticeship and training on the job.
It also has much better and more predictable hours than, say, the ever-popular gaming industry.
You're right that low pay is an issue, but pay is crap in part because of outsourcing. Any machine shop in the US is competing with China to an extent that, oddly, we don't actually see in the software space. The West is mostly using Western-made software, and outside of that, only India is invited to the club (and even that only as a subcontractor).
But a lot of it is also just culture / parenting. Every parent wants their kid to become a software engineer, a lawyer, or a doctor. There's basically zero pride in being a machinist.
>But a lot of it is also just culture / parenting. Every parent wants their kid to become a software engineer, a lawyer, or a doctor. There's basically zero pride in being a machinist.
I disagree.
A lot of people, nerds (and me) included, would see no problem in taking an operational role inside manufacturing if it paid a livable wage and offered a career.
But somehow people still act shocked when nobody wants to work a job that pays retail wages.
>> traditional manufacturing has a robust tradition of apprenticeship and training on the job.
We also have such tradition. Senior software engineers are teaching juniors.
Would tradition manufacturing employ an apprentice without knowledge about the tools and materials? I doubt that. Everyone now wants to get a senior employee with 20y of experience and pay junior rates.
Probably not. Plumbing pays pretty well and it's almost never a parent's dream. It's probably below being an office assistant. We just attached a stigma to everything blue collar (= "uneducated").
Plumbing does not pay well for the average plumber. Not doctor-well at least, which is what you are responding to.
Can some plumbers make good money? Absolutely. But you need to do dirty, body-breaking work for years.
There’s nothing wrong with being a plumber and it’s a profession we surely need young people to go into. But the pay and trade offs vs a white color job will lean heavily in direction of the latter.
Being a plumber pays acceptably, not well. Being a plumbing business owner can pay well, but would have less catchet than a doctor making the same amount, sure, but it's also far less common.
There is a lot more demand for software engineers and doctors. (I'm not sure about lawyers - it used to be good, but last I heard it wasn't as good) The world doesn't need many machinists anymore since CNC can do such a good job. The world needs people to design the parts the CNC makes though - often in the past you would take that to a machinist to design on the machine, now you need to know CAD.
> Every parent wants their kid to become a software engineer, a lawyer, or a doctor.
Uh, what? Both sides of my family is nine aunties/uncles, and my parents. None of them pressured their children to select a specific career or even major in university. I think my experience is fairly common.
> There's basically zero pride in being a machinist.
I cringe when I see the term "pride" used this way. Would you say that to a career machinist? I wouldn't. When I was growing up, I can still remember my elementary school bus driver named Barbara. She really took pride in driving our bus everyday, but was probably paid peanuts.
If you don't attend university, being a machinist is skilled labour that pays well. Plus, you can continuously self-educate to climb into more and more skilled roles that pay more. The salary is solidly middle class, and the work is probably in areas with a low cost of living.
What are you talking about? Show me one survey that shows DIY crafts as a whole becoming more popular in the past several decades. It's just not a thing. People increasingly don't have room for it, computers and cell phones have won as a way to spend the bulk of our free time, and even if you want to create, the primary outlet is the digital world.
> 3D printing at home is more popular than ever
Sure, but that's a given because the technology didn't exist until very recently. I can also say that Tesla Cybertruck is more popular than ever, but that doesn't mean much, right.
> Even families I know who don’t have a computer at home, only their phones, now have 3D printers and use them.
A vanishingly tiny proportion of families in the US or in the EU own 3D printers and use them.
A vanishing small number of people ever did those things. Many people owned tablesaws but they rarely used them to make something for fun. They owned the dream and subscribed to the magazines but they only rarely made anything.
Yeah, unfortunately you are correct. Life's too busy nowadays and there's not enough time to do a quarter of the things one wishes especially projects that require planning and months to implement.
Moreover, the reduction in or loss of manual trades (woodworking, machining, toolmaking, etc.) whether at home or in industry, amounts to an overall deskilling of the citizenry which has strategic implications.
I don't see how that holds. Detroit and other cities it we're once filled with skilled machinists, pattern makers, carpenters, pipefitters, etc. Now they're not. None of this has anything to do with what those people did for fun in their spare time (and you're right: that was rarely "the same thing I do at work but in my garage"). We are certainly deskilled in these areas compared to times in decades past.
Sherline made tools for hobbyist. While they were high quality such that a professional could use them, professionals typically would buy larger more expensive machines (which can still to the small jobs when that is they got, but could also take on larger work the Sherline couldn't handle).
The the hobbyist in the garage needs to be the topic here, not professionals. The hobbyist enjoying things in the garage still exists, but is still rare. Those that do the work are skilled. If you want to join them there are plenty of books on the subject, it just takes some practice to develop the skills.
> Sure, but that's a given because the technology didn't exist until very recently.
I built my first 3D printer over 10 years ago. The latest printer I bought is much more refined, but it's still stepper motors, belts, a gear driven extruder, and a nozzle like my original printer. The technology has been around for a long time. The popularity explosion is recent.
> A vanishingly tiny proportion of families in the US or in the EU own 3D printers and use them.
We're talking about hobbies and that's how hobbies work, yes.
The hobby is growing significantly, though. Arguing that most households don't have a 3D printer is just a strawman.
"Most households", no, but a statistically significant number? Yes.
Two of my co-workers have one each, and another is planning on buying one for her kids for Christmas --- DJI revenge-investing in Elegoo to spite the folks who left their company to start Bambu Labs pretty much knocked the bottom out of the market and halved (again) the price with the Centauri Carbon, resulting in a glut of ~$300 CoreXY printers, many of which are enclosed and hence have a higher perceived value.
Most estimates are that 2024 had ~4.5 million shipped, and 2025 looked to be even more, and no signs of the upward trajectory decreasing for 2026. For one small snapshot see the recent Snapmaker campaign for the U1, >$20M raised from >20K backers. With most sales to the U.S., for ~135M households, while not at 10%, it is almost certainly past 5%, so approximately 1-in-20 households.
> You could Google this and find a thousand articles about it.
Such articles are frequently by the people who would benefit the most from them and usually lack specifics. My favorite version of this is to do a web search for "fastest growing hobby in the world". Lots of hobbies are the fastest growing hobby in the world, it turns out.
I'm not sure if it's that simple. Many space-intensive DIY hobbies are on the retreat simply because more people are moving into cities or very dense suburbs where you have no practical opportunity to, say, build a woodworking shop. Metalworking was even more severely impacted, for obvious reasons.
Also, long before mail-order offshore manufacturing, people have been lured away from most of this by the promise of 3D printing. Every other techie I know owns a 3D printer. Most have not made anything interesting with it because they soon learn that the real difficulty is materials science and 3D design. But as a nerd, it's the thing you're supposed to own. In contrast, I know maybe four people in the SF Bay Area who are into woodworking, metalsmithing, CNC, or stuff like that.
I don't see mail-order prototyping as a conspiracy... it's more of a culmination of the DIY community chasing the mirage they were sold on in the early 2010s. I still remember the HN claims: home manufacturing revolution! It's the beginning of a brand new economy, why would anyone buy anything anymore when you can just click "print"... with some amusing parallels to how some folks now talk about building your own software with AI.
The mail-order stuff isn't evil insofar that it may allow some people to make things they couldn't make before. On the flip side, it's a very inefficient way to learn, so I think it will reduce hardware literacy among those who could afford a garage workshop.
FWIW, I think the same goes for AI-assisted hardware design. On one hand, I see people who otherwise wouldn't be DIYing getting into DIY: "I asked Claude and it one-shotted this rectangular enclosure for me, it's amazing!", "Claude made this RP2350 breakout board for me, how cool is that". But I also see established DIYers getting a degree of brain rot.
we have had a home manufacturing revolution, though, and making every kind of hardware is crazy accessible, look at the explosion of niche products in every single hobby space over the past ten years
I miss absolutely nothing about etching my own single layer PCBs at home
First, PCBs: neither do I, I'm not sure why you're assuming that's the position I want to take. This has nothing to do with 3D printing or mail-order metal parts.
Second, I don't think there's a revolution for the general population. Your archetypal middle-aged dad / curious teenager are DIYing less than 20 years ago, which in turn was less than 40 years ago. The stats here are clear, most of "garage" occupations are on the decline. Less space, higher costs, more emphasis on safety. Some hobbies, such as backyard chemistry, are effectively gone. Shop classes in most schools are gone. Making your own bookshelf or building a radio from a kit is no longer a rite of passage.
Third, the "explosion of niche products" - maybe, but I think some of it is just visibility. Kickstarter, HN, Hack-a-Day, and so on. Before that, we had hobby journals for woodworking, electronics, etc, and they were always full of ads for niche gear. You just weren't seeing the totality of it in a doom-scroll loop. Visibility has benefits, but my point is that the qualitative change isn't as profound.
I actually believe it's mostly about cost. Things you'd DIY in the past are now crazily accessible: DIY is usually more expensive than getting a commodity product.
But still, I would love to have a way to build PCBs at home and I'm keeping an eye on people (mostly the Opulo/LumenPnp guy) experimenting milling and more recently engraving with fiber & UV lasers.
I've reached the point where I know I'll die before I get a faction of my todo list done. Building PCBs is on of those required steps for several parts of my todo list, and so I'll be glad to outsource if that is affordable. (if is key here, if it turns out expensive I might get a setup to do my own)
PCB manufacturing is reasonably cheap. YMMV, but JLCPCB was ~$20 for a set of 5 boards (minimal order quantity) last time I ordered.
What you lose is not monetary, it's the ability to iterate quickly. It can be mitigated with next day manufacturing+delivery from local manufacturers... at a cost (IIRC, ~$80).
But frankly, it would be nicer to pluck a blank board in a laser, press burn, and have a good enough board in 10 minutes.
Fiber lasers are kind of capable of doing that, but they are still a bit expensive (~$2000), the result is imperfect (scorched epoxy board), and you can do at most 2 layers... if you are willing to manually tap rivet vias to connect the layers.
It's a two-way street. When someone tells you "don't do that", they usually mean they don't want to be on the receiving end of what often amounts to LLM-powered literary diarrhea.
Yes, in theory, the quality can be the same or better than human output. In practice, the tool creates two problems: it's too easy to produce walls of text for no compelling reason, placing asymmetric burden on the recipients; and it's too tempting to click "send" without reading because you're "90% sure" the email / doc is correct, wasting everyone's time multiple times every week when it isn't.
I think it's hard to generalize because people have different approaches to writing, but to me, there's a marked difference between iteratively polishing my own text and editing for someone else (and that includes LLM output produced by my own prompts).
In the second case, these sentences were never a "voice in your head", and you never invested any effort into shaping them, so it feels more like janitorial work / drive-by nitpicking than craftsmanship.
