My most recent example of this is mentoring young, ambitious, but inexperienced interns.
Not only did they produce about the same amount of code in a day that they used to produce in a week (or two), several other things made my work harder than before:
- During review, they hadn't thought as deeply about their code so my comments seemed to often go over their heads. Instead of a discussion I'd get something like "good catch, I'll fix that" (also reminiscent of an LLM).
- The time spent on trivial issues went down a lot, almost zero, the remaining issues were much more subtle and time-consuming to find and describe.
- Many bugs were of a new kind (to me), the code would look like it does the right thing but actually not work at all, or just be much more broken than code with that level of "polish" would normally be. This breakdown of pattern-matching compared to "organic" code made the overhead much higher. Spending decades reviewing code and answering Stack Overflow questions often makes it possible to pinpoint not just a bug but how the author got there in the first place and how to help them avoid similar things in the future.
- A simple, but bad (inefficient, wrong, illegal, ugly, ...) solution is a nice thing to discuss, but the LLM-assisted junior dev often cooks up something much more complex, which can be bad in many ways at once. The culture of slowly growing a PR from a little bit broken, thinking about design and other considerations, until its high quality and ready for a final review doesn't work the same way.
- Instead of fixing the things in the original PR, I'd often get a completely different approach as the response to my first review. Again, often broken in new and subtle ways.
This lead to a kind of effort inversion, where senior devs spent much more time on these PRs than the junior authors themselves. The junior dev would feel (I assume) much more productive and competent, but the response to their work would eventually lack most of the usual enthusiasm or encouragement from senior devs.
How do people work with these issues? One thing that worked well for me initially was to always require a lot of (passing) tests but eventually these tests would suffer from many of the same problems
Am I spending too much time on HN or is every post/comment section filled with this same narrative? Basically, LLMs are exciting but they produce messy code for which the dev feels no ownership. Managing a codebase written by an LLM is difficult because you have not cognitively loaded the entire thing into your head as you do with code written yourself. They're okay for one-off scripts or projects you do not intend to maintain.
This is blog post/comment section summary encountered many times per day.
The other side of it is people who seem to have 'gotten it' and can dispatch multiple agents to plan/execute/merge changes across a project and want to tell you how awesome their workflow is without actually showing any code.
> This is actually not a million dollars in singles. It is over $1,000,000. The box was created with the wrong dimensions by the contractor, but they still decided to fill it, display it, and claim it is $1,000,000.
>
> Source: Tour Guide at the Chicago Fed
> 4.5.4 Push Notifications must not be required for the app to function, and should not be used to send sensitive personal or confidential information. Push Notifications should not be used for promotions or direct marketing purposes unless customers have explicitly opted in to receive them via consent language displayed in your app’s UI, and you provide a method in your app for a user to opt out from receiving such messages. Abuse of these services may result in revocation of your privileges.
> You might be asking: why did you rewrite tmux in Rust? And yeah, I don’t really have a good reason. It’s a hobby project. Like gardening, but with more segfaults.
I love this attitude. We don’t necessarily need a reason to build new things. Who knows what will come out of a hobby project. Thanks to the author for the great write up!
Also, my gardening is full of segfaults, coding a new project is definitely safer to my yard.
Many employers want employees to act like cult members. But then when going gets tough, those are often the first laid off, and the least prepared for it.
Employers, you can't have it both ways. As an employee don't get fooled.
> Companies that monetize user data in exchange for “free” services that abuse your privacy aren’t affected by this [the app store tax], as they don’t process payments through the App Store. However, privacy-first companies that monetize through subscriptions are disproportionately hit by this fee, putting a major barrier toward the adoption of privacy-first business models.
Huh. I’ve never seen it framed this way and it might be the most compelling argument I’ve heard to date. It’s not simply a debate about whether a company should be allowed to be vertically integrated in isolation, but whether that vertical integration allows them to exert unfair distorting pressure on the free markets we are trying to protect.
> Subject to applicable law, Comcast may disclose information generated by your WiFi Motion to third parties without further notice to you in connection with any law enforcement investigation or proceeding, any dispute to which Comcast is a party, or pursuant to a court order or subpoena.
Sounds like, at least in some limited circumstances (using the provided WiFi AP, having this feature turned on, etc), ISPs are going to be able to tell law enforcement/courts whether anyone was home at a certain time or not.
It's an incredible success story, and the engineering they did upfront (primarily led by co-founder Evan Wallace) that set the stage for their success is the stuff of legends. https://madebyevan.com/figma/ has links to numerous blog posts breaking it down, but here are some choice quotes:
> [Evan] developed the hybrid C++/JavaScript architecture for Figma's editor that made it possible to build a best-in-class design tool in the browser. The document representation and canvas area is in C++ while the UI around the canvas is in JavaScript (the team eventually settled on TypeScript + React for this). This let us heavily optimize the document representation to reduce memory usage and improve editing speed while still using modern UI technologies for fast iteration on our UI. C++ development was done using Xcode (not in the browser) to provide a better debugging environment.
> Even though the contents of Figma documents are similar to what HTML can display, Figma actually does all of its own document rendering for cross-browser consistency and performance. Figma uses WebGL for rendering which bypasses most of the browser's HTML rendering pipeline and lets the app work closely with the graphics card. The rendering engine handles curve rendering, images, blurs, masking, blending, and opacity groups, and optimizes for high visual fidelity.
> [Evan] developed Figma's multiplayer syncing protocol, worked on the initial version of the multiplayer live collaboration service (a kind of specialized real-time database), and added multiplayer syncing support to Figma's existing editing application. The initial version was written in TypeScript but [he] later ported it to Rust for improved performance and stability.
It's a great reminder that it's not premature optimization if your UI's fluidity is your distinctive feature and your calling card! And the business acumen to turn this into such a wildly successful product in the context of competitors with kitchen-sink feature lists can't be understated, either. I have an incredible amount of respect for this team, and they should inspire all of us to tackle ambitious projects.
Apple without Ive and Jobs increasingly has a taste problem. Everything from their ads to things like this are just in really poor taste, and aren’t something that they would have done 15 years ago because they would have thought it was beneath their brand.
I like Apple, so I’m really hoping they bring on someone to solve this. Otherwise they’re on track to be the same as every other tasteless tech company.
I don't see what's so bad about wanting to avoid an area where there's police activity going on. It has nothing to do with whether or not you're doing anything wrong, it's as simple as not wanting to get hassled at a DUI checkpoint or get stuck in traffic because they need 8 squad cars taking up a lane to k-9 search someone. As a more tan law-abiding US citizen, the possibility of some agent asking me for papers and then asking probing questions to "prove myself" anywhere that's not an airport is enough for me to want a heads up not to be in area where that might happen.
There's barely any point examining the app on its merits.
The mere existence of the app shows resistance to the government's attempts at establishing something approaching a police state. They are against the app for that reason. They don't really care about what it does or does not do. It could be an app where you press a button and the phone says "boo ICE" and they'd still happily claim it endangers officers lives.
(the fact that they're also able to attack independent media at the same time just makes it all the more alluring target)
Mildly reminds me how being online on AIM or ICQ was an actual invitation to chat. I had so many interesting conversations with people I barely knew.
There's no source of that signal that someone is open to chitchat these days, and it's in my opinion kind of killed what was once great about online communication.
> To download W3C's editor/browser Amaya, _click here_.
Is extraordinarily clear. I'll click the link and it will either download directly, or it will be a download page.
In contrast:
> Get _Amaya_!
That suggests a link to the Amaya website, not a download page. That's not effective for a download.
Similarly:
> Tell me more about _Amaya_: W3C's free editor/browser that lets you create _HTML_, _SVG_, and _MathML_ documents.
This is terrible. It's not about downloading, and "tell me more" is the command, but not linked! For all I know, the "Amaya" link goes to a corporate landing page, not the "tell me more" information I actually need.
The conventional uses on the web are totally fine:
> To download W3C's editor/browser Amaya, _click here_.
> _Download Amaya_, the W3C's editor/browser.
The idea that links shouldn't be verbs seems very silly to me. Links should absolutely be verbs, when they involve an action like downloading or finding out more. Obviously that's different from "reference" links like in Wikipedia, where you're finding more about a topic.
And "click here" makes it exceptionally clear that a link isn't merely a reference link, but an action link. When I see:
> Get _Amaya_!
That... doesn't tell me how to get Amaya. That tells me "Amaya" is a reference link, not a download link.
The problem is these meetings are so low information density even an AI summary is not worth my time. And it’s not some elitist mindset. It’s like the entire reason there are these regular meetings is to make some mid level person feel better. They like giving directions vocally because that authority is harder to question than if they wrote up a memo and all the receivers can poke holes in it. I’m convinced most meetings are to make up for poor writing skills.
It's surprising how hostile youtube is to multilingual users. Probably all in some attempt to show off their translation capability or to improve the experience for users who may want to access content in a language they don't speak? Or it's just as dumb as this was on some product managers "designed and implemented" line to get promoted?
But, surely someone sane there has to realize there is a large number of users out there who speak more than one language, and don't need Google do "help" them or "guess" for what language they like more.
We really need better ways of measuring economic health. I could lose my six-figure job, turn around, and get hired on as a server at Applebee's for minimum wage, and the "unemployment" rate would stay the same. Not to mention that it doesn't include those not actively looking for work.
Either way, "full employment" doesn't mean much unless you take into account whether people are actually able to live a stable lifestyle or are burning the candle at both ends just to put food on the table. One of these enables folks to buy nonessentials and fund all those sectors of the economy, the other doesn't.
It says something that he still believes he has "missionaries" after betraying all the core principles that OpenAI was founded on. What exactly is their mission now other than generating big $?
I just feel extremely sad about the mass quantity of events like this happening right now because they are all aggregate to huge negative effects but the average person knows nothing of it. It feels so unfixable.
> With over 90% of the PC market running on NVIDIA tech, they’re the clear winner of the GPU race. The losers are every single one of us.
I have been rocking AMD GPU ever since the drivers were upstreamed into the linux kernel. No regrets.
I have also realized that there is a lot out there in the world besides video games, and getting all in a huff about it isn’t worth my time or energy. But consumer gotta consoooooom and then cry and outrage when they are exploited instead of just walking away and doing something else.
Same with magic the gathering, the game went to shit and so many people got outraged and in a big huff but they still spend thousands on the hobby. I just stopped playing mtg.
Last year PlasticList discovered that 86% of food products they tested contain plastic chemicals—including 100% of baby food tested. The EU just lowered their "safe" BPA limit by 20,000x. Meanwhile, the FDA allows levels 100x higher than what Europe considers safe.
This seemed like a solvable problem.
Laboratory.love lets you crowdfund independent testing of specific products you actually buy. Think Consumer Reports meets Kickstarter, but focused on detecting endocrine disruptors in your yogurt, your kid's snacks, whatever you're curious about.
Here's how it works: Find a product (or suggest one), contribute to its testing fund, get detailed lab results when testing completes. If a product doesn't reach its funding goal within 365 days, automatic refund. All results are published openly. Laboratory.love uses the same methodology as PlasticList.org, which found plastic chemicals in everything from prenatal vitamins to ice cream. But instead of researchers choosing what to test, you do.
The bigger picture: Companies respond to market pressure. Transparency creates that pressure. When consumers have data, supply chains get cleaner.
Technical details: Laboratory.love works with ISO 17025-accredited labs, test three samples from different production lots, detect chemicals down to parts per billion. The testing protocol is public.
You can browse products, add your own, or just follow specific items you're curious about: https://laboratory.love
This one is coming in fast, it has an eccentricity of over 6 with the current fits. For point of reference, 1I and 2I have eccentricities of 1.2 and 3.3.
Right now it is mostly just a point on the sky, it is difficult to tell if it is active (like a comet) yet. If it is not active, IE: asteroid like, then the current observations put it somewhere between 8-22km in diameter (this depends on the albedo of the surface). From what we know, we would expect it to likely be made up of darker material meaning given that range of diameters it is more likely to be on the larger end. However if it is active, then the dust coming off can make it appear much larger than it is. As it comes in closer to the sun and starts to warm up it may become active (or more active if its already doing stuff).
It will not pass particularly close to any planet. It will be closest to the sun just before Halloween this year at 1.35 au, moving at 68 km/s (earth orbits at 29-30 km/s). It is also retrograde (IE, it is moving in the opposite direction of planetary motion), for an interstellar object this is basically random chance that this is the case.
The next couple of weeks will be interesting for a bunch of people I know.
Source: Working on my PhD in orbital dynamics and formerly wrote the asteroid simulation code used on several NASA missions: https://github.com/dahlend/kete
We have good metrics. The problem is the media seems to only ever look at one of them at a time but we need to look at several at once to get a more complete picture.
Your scenario would be called out by median household income, or better median disposable household income. Even the good old GDP per capita covers your case.
Workforce participation also can be valuable instead of or in addition to unemployment numbers, since you fall out of the count once unemployment benefits expire. However, we need to look at it by age bracket. Lower workforce participation between 20 and 60 is probably bad whereas higher workforce participation over 60 might also be bad.
IMO the problem isn't that the metrics aren't there but that the public discourse either lacks motivation, understanding or incentive to take a proper look. That every discussion of these numbers on social media has a substantial portion of people not understand the difference between median and mean certainly doesn't give me confidence this will ever improve.
It was falling behind. The dodgy stores were getting more creative and Fakespot needed to play catch up.
You've got stores that would include a $5-$20 coupon/gift card in the item in exchange for a positive review. Sure, this didn't 1:1 translate but if a user did it would look like a legitimate review.
You've got a plethora of LLMs out there just itching to GENERATE.
Then an expensive option I was suprised happened - I bought a Dyson clone vacuum cleaner off of Amazon. A few weeks later, the company emailed me and said 'We have a new model. Buy that one, leave a review, we'll refund the purchase'. So I did it. This happened about 10 more times in 2024. My outdoor shed is entirely stick vacuums.
Feel a bit dirty doing it but that's ok I've got 12 vacuums that can clean my conscience.
I think Fakespot would have difficulty with all 3 of these scenarios.
Big picture, I'll always believe we dodged a huge bullet in that "AI" got big in a nearly fully "open-source," maybe even "post open-source" world. The fact that Meta is, for now, one of the good guys in this space (purely strategically and unintentionally) is fortunate and almost funny.
This is something I've always wanted to write about, and I imagine that someday I'll end up with a long article, but basically, it's the idea that the internet used to be offline by default, and now it's online by default.
People used to be offline by default. You had to “connect to the internet.” Open MSN, go into forums and check the latest unread messages, come back from a concert and manually upload the photos to your Fotolog or wherever. Now it's the opposite. We are online by default. The expectation is that we're always connected and respond quickly. Going to a sports event or a concert? You have to post a story to Instagram from that very place, not when you get home. Someone sends you an email or a WhatsApp message? You’re expected to reply as soon as possible.
That’s what I miss most about the internet—the idea and the feeling that I would go online when I wanted to, not that I lived inside the internet 24/7.
What John Carmack is exploring is pretty revealing.
Train models to play 2D video games to a superhuman level, then ask them to play a level they have not seen before or another 2D video game they have not seen before. The transfer function is negative. So, in my definition, no intelligence has been developed, only expertise in a narrow set of tasks.
It’s apparently much easier to scare the masses with visions of ASI, than to build a general intelligence that can pick up a new 2D video game faster than a human being.
Not only did they produce about the same amount of code in a day that they used to produce in a week (or two), several other things made my work harder than before:
- During review, they hadn't thought as deeply about their code so my comments seemed to often go over their heads. Instead of a discussion I'd get something like "good catch, I'll fix that" (also reminiscent of an LLM).
- The time spent on trivial issues went down a lot, almost zero, the remaining issues were much more subtle and time-consuming to find and describe.
- Many bugs were of a new kind (to me), the code would look like it does the right thing but actually not work at all, or just be much more broken than code with that level of "polish" would normally be. This breakdown of pattern-matching compared to "organic" code made the overhead much higher. Spending decades reviewing code and answering Stack Overflow questions often makes it possible to pinpoint not just a bug but how the author got there in the first place and how to help them avoid similar things in the future.
- A simple, but bad (inefficient, wrong, illegal, ugly, ...) solution is a nice thing to discuss, but the LLM-assisted junior dev often cooks up something much more complex, which can be bad in many ways at once. The culture of slowly growing a PR from a little bit broken, thinking about design and other considerations, until its high quality and ready for a final review doesn't work the same way.
- Instead of fixing the things in the original PR, I'd often get a completely different approach as the response to my first review. Again, often broken in new and subtle ways.
This lead to a kind of effort inversion, where senior devs spent much more time on these PRs than the junior authors themselves. The junior dev would feel (I assume) much more productive and competent, but the response to their work would eventually lack most of the usual enthusiasm or encouragement from senior devs.
How do people work with these issues? One thing that worked well for me initially was to always require a lot of (passing) tests but eventually these tests would suffer from many of the same problems