- Quoted phrase in the opening sentence
- Opening quote: “Software engineers are the most expensive employees on Earth”
They are not. CEOs are.
"It built a machine that turns valuable things into free things, and then pointed it at the businesses it needed to stay alive. That isn’t a revolution."
It is a revolution. Free models are out there. The "toll booths" (software companies) might not survive, but if that's the case, that is the revolution: Because something that you had to pay a massive "toll" for is now essentially free. That might not help the AI companies but it will be a revolution for everyone else that had to pay the tolls, likely making everything cheaper and everyone better off.
(Yes, everyone. Even a very poor person in a highly developed European country lives better than a mildly rich person before the industrial revolution, and I'd say the average person lives at the levels even royalty could barely dream of.)
Even if this bites the AI labs and they don't survive, building software is not going to become significantly more expensive than it currently is with free models that are being sold at cost (barring something major like a global nuclear war), mainly because you can always continue to run the existing models on existing hardware even if progress were to magically stop.
This is already present for the last decade. It's the foundation which makes "sports DSLRs" possible. AI here is not what we know as today's "AI". These are specialized models baked into DSPs to run at inhuman speeds. In Sony's case it's phase data, image itself, information from the lens, and sometimes the accelerometers on camera. This is not making something up from thin air.
AI assisted autofocus is smarter phase detect autofocus. Subject tracking is an educated guess for where the focus shall move, based on movements in the image.
My ~20 year old Nikon D70s has a 1004 pixel "scene detection and metering" sensor "trained" on thousands of images to make educated guesses in full-frame metering mode to exposure correctly. This is in no way helping or altering the image in any way. Same for faster autofocus and object tracking.
Exposure, white balance, noise reduction, color correction, HDR et al can be all achieved with deterministic and traditional algorithms as well. Some of these processes have history in film photography and carried over to digital from analog. Some photographers are famous for their darkroom treatment of their shots during development.
Processing an image with deterministic algorithms using the data in the raw photo is fine. This is how darkrooms work as well. However, guessing things, altering images by adding something which was not there and claiming otherwise is not acceptable. This can be a new element inside the scene or pixel details which was not resolvable by the lens/sensor/film to begin with. Both are "photomanipulation", not photography.
Edit: an earlier statement from Nikon said that he was complying with the investigation and "At this time, we don't see that any rules were violated" https://www.bbc.com/news/articles/ck4gjn1yzprno
- Quoted phrase in the opening sentence
- Opening quote: “cmd”
- Quoted phrase in the opening sentence
- Opening quote: “giving away”
- Opening sentence uses a quote to challenge or distance a term
- Opening sentence challenges the term “giving away”
It clearly isnt happening.
- Quoted phrase in the opening sentence
- Opening quote: “AI automates exploits at scale”
There are dozens (or hundreds) of functional exploits already. The problem with php2shell exploits is that (as of now) there's no "one exploit pwns all", and researchers need to spend nontrivial amount of time on every (potentially) exploitable target.
So I don't think I agree - it's a fun challenge for the itsec people around here, and some people will definitely try. There's no real opportunity cost here. The only problem is that (as of now) exploiting PHP requires some knowledge about the server configuration, and we don't have that here.
- Quoted phrase in the opening sentence
- Opening quote: “AI can trivially automate exploits, be afraid”
My understanding has always been that the goal of open source software is to make software more free and to be able to make higher-quality, more ambitious software. With LLMs making it easier than ever to modify, study and read software, and independent developers able to work on more ambitious projects, I would expect open source advocates to view LLMs as strongly congruent with the free software movement. Instead, this post stresses the incongruency of open source software with AI-assisted coding on grounds of human-centeredness and the "cheapness" of tokens. What does human-centeredness have to do with open source software? Isn't the cheapness of tokens directly aligned with the goals of open source software of making it easier and cheaper to create software? I just don't get it.
- Quoted phrase in the opening sentence
- Opening quote: “carefully checked and corrected”
- Opening sentence uses a quote to challenge or distance a term
- Opening sentence challenges the term “carefully checked and corrected”
I quite like Daft Punk but that ™ makes them uncool very quickly. Immediately makes me assume it's just a brand name now, run by an agency, music work outsourced to freelance ghost composers, and so on. Like Philips TVs and "Jamie Oliver" cookware.
- Quoted phrase in the opening sentence
- Opening quote: “decomps”
Their lazy cloud-chasing is normal but my high-effort clout-chasing is lame and makes me a fraud?
We're social animals. We like doing cool stuff and getting social status for it. We like breaking new ground. We like going where nobody has gone before.
Is it irrational and emotional? Maybe. Being human is being irrational and emotional sometimes. It's fine.
I believe though it is an old phenomenom that if (given the default context) you ask a random number, it will more often than not give numbers from a very small set.
I just used a "Temporary chat" with ChatGPT twice to "Give a random number" and it gave 42 both times (this shuold be rather understandable why it ended up with this). Then I asked "Give a random number in 1-10" twice and it responded with 7 for both (and this too is an old "magic number").
It's one thing to discard it as "numerology and gifs" but it's another to willfully disrupt legal elections in a soverign country.
Or is a decade a really long time and we've all forgotten about Cambridge Analytica?
We're in bubbles everywhere now. Enforced by mods and payment processors and social networks. I'm now much more careful about what I post online. We don't have an actual public square to discuss things.
When you are mostly just shown content that you’ve agreed to see, there’s less of an issue when you see something that offends your particular sensibilities, values, etc. because you agreed to see it, or you went looking for it.
But when the potential content you’ll see can include literally anything, there is a real incentive for the platform to make that content as neutral as possible. Because if you’re shown something that you really don’t like, you’ll stop scrolling and the engagement stats will go down.
- You need to have a mental model to grow an idea
- When a pattern is known you can use a library!
- When a pattern is not known AI will make stuff up
No matter how "good" coding with an AI will become there is no way you get around learning, having someone understand what's going on and take responsibility.
I'm fine with formal checking and fuzzing or getting hints where a bug might come from - but these are iterations on tools we used before like a debugger, strict types, functional programming and well fuzzing.
I tried to have an AI iterate on a step i don't fully understand yet and i always regretted it. Highly interconnected projects will grind to a halt until you do learn.
You can check with people doing highly specced software in the medical domain that even with very good specs there is ambiguity in natural language thus it can never be the source of reliable and manageable software.
However - i read about some smart guys in the 40s or so encountering the need for a formal language to tell a computer what to do with 0 ambiguity and i think they called it a "programming language"
With the tutorial, the summary also makes more sense. I have not read the other parts yet, but so far I like what I see.
All is in pretty good taste, and I like what I've seen so far. I do not care if the author himself came up with the idea, or if AI came up with it and the author approved.
I do not understand why you "have to" know this to make use of it. You can judge ideas independently of where they came from (and you should).
Of course Humans can make errors when translating that to English. But if you’ve read any amount of Claudish it should be obvious that interpolating the code to documentation with an LLM typically produces a significantly worse result in almost every way. Except for superficially “sounding” correct.
Even Rust's "enum" feature, which can store different sized things all in the same space, does that at the cost of the smallest variant being just as big as the biggest.
Goose can do all of that not only inline (no pointers), but with variable size, meaning small items take less space. It's like a serialized data representation, but iterable. It makes for big space savings vs these languages.
The purpose of software is to help people with their problems. I think it's easy to tell ourselves that other people want to use our tools which we really made to help ourselves (especially if them using it also helps ourselves even more!). And also, this is all so new and there are real opportunities that it is good to learn and try new things and be open to actually making something valuable.
But! Other people extensively using agents are literally the last people who would pay for a project like this! And unless it's a fulltime job like it is for a 5 or so well resourced tech companies, it's hard to make something as good and useful as them. I say this as someone who spent a long time on this problem as my fulltime job.
We should still have fun with our work. But literally, "software for other software people, that solves the problem of us needing to write more software for writing software, faster", who is that helping? There are no constraints or end users besides ourselves, and half of that is not so much it benefiting us through productivity, but learning new things and play - again, important. But there's a reason so many people are doing this and I think it's because they don't have to go outside and solve someone else's problem, or do something unusual, or involving unfun non-software aspects - agents are fun! But how many random people's agent have you installed?
IMO: if the technology does not help a non-technology person live a better life, and especially if it involves "using an agent to build a framework for agents" we should do some more, deeper introspection about what problem we are actually solving and whether or not we're being ambitious. And I say that not to pick on you but myself.
A lot of the AI discussion seems to be focused on what these tools mean for an individual developer, but in a team setting any gains are at risk due to coordination tax and ideological friction. In my experience the ideological friction is most of it right now.
Figuring out a way to accelerate the process in light of the code writing bottleneck being removed is essential. I've observed a development team regress back to fully manual development because their daily PR/code review process is too slow. It doesn't matter how fast you write code if you can't actually merge it and ship it to the customer faster too. PR stacks can only take you so far before the conflict monster gets too big to fight.
I don't think AI can fix the process, ideology or people issues. I've seen how long an average PR review actually takes. If these employees were simply doing their jobs at all we'd be 10x better off regardless of AI. A truly skilled developer who makes a genuine attempt should be expected to solve 99% of their tasks within the same business day as of October 2026.
Burning down the backlog to zero doesn't mean the development team will be subsequently laid off. Quite the contrary. It will create a vacuum that the business will fill with new work that may incidentally increase the developers' compensation packages. No one is laying off the development team that just chewed through 50 items per day for 3 straight weeks. Who would support all the potential bugs they just created? That's right. They might hire even more developers. This is called a "win-win" scenario. Not every game must be zero sum, but I understand rejection of this possibility to be part of the aforementioned ideological issues.
Any prompt jockey can make a faster ripgrep. The spec is "faster ripgrep".
What is the one thing that only you can specify, where your competitive advantage is maximized? Find that thing and pursue it with all your energy!
- Quoted phrase in the opening sentence
- Opening quote: “a ripgrep fork that ... is faster for my codex ripgrep workloads”
Perhaps we should be more ambitious in the types of problems we try to solve, and the types of value we try to provide with the software we build.
- Quoted phrase in the opening sentence
- Opening quote: “slow software that the creator is claiming is fast”
"funnily enough, I have some faith in some of the projects that people are the most skeptical of, e.g., every time I see pgrust somewhere, there are a lot of skeptical comments. But, without having looked into the details of what he's optimizing, I would trust that they're not doing something shady with their benchmarks because Michael Malis started the project (and is still involved)"
This aged like milk. pgrust was caught cheating on all benchmarks this week.
Claude in its current Opus 5.5 form has made some wild reaches in response to reasonable-seeming prompts. For instance: "this error is getting swallowed, make sure the real termination message exposes the root cause" -> discovering that the issue is actually in a third party library that isn't exposing the original error except so through logs, so doing some hackery to inject a log-capturer that gets the stdout, buffers it, and looks for the message in order to re-throw based on that...
And it can demonstrate that working, but it's added some fragile and/or unintended consequences like (a) string matching on a particular WARN log message of a particular version of the lib, and (b) buffering those logs such that there were now cases where they would never make it to the actual output stream before the program terminating.
It's certainly capable of being used smarter than that, but it's also really easy to end up with messy fragile stuff if you get lazy about reviewing the output.
- Quoted phrase in the opening sentence
- Opening quote: “complaining”
> I have a similar thought about "there's no reason for software to be slow" BTW; in many cases, if you use AI to speed up some flow, you will add a bunch of messy code that will be overfit for the use case you tested on and will neither generalize, perf-wise, for all the use cases you should care about, nor will be maintainable upon future changes.
LLMs have hardened hundreds of apps where I work.
Maybe it's a language-specific thing. Most of the time, in Java, LLMs take into consideration tradeoffs (especially if they are part of the initial prompt) and actively explains them. It generally won't make messy workarounds, without a lot of extra prompting forcing it to do so.
Ofc, if you have:
- Functional tests - you'll be able to see it in the execution times of specific APIs
- Stress testing - performance profiling is the most important output artifact
I don't mind too much, I just miss overall quality and don't care who tf is "represented" or not.
Software is a very clear example on this front. Valheim [1] is a game that was made mostly by one guy, at first at least. Not that long ago, this would have been a multi million dollar game produced by a wide array of people. It's just that tools, markets, and more have made it possible for even single developers to have an absurd degree of efficiency. And LLMs are probably going to bump this up exponentially.
This is why a metaphor I'm currently drawn to is one of people living in a very agrarian society just prior to the Industrial Revolution. What would they have envisioned being created? Super farms of course. And those super farms were indeed created, but the overall impact was such that those super farms became a practically inconsequential part of the overall economy. The future we're moving into isn't going to be the same but more efficient for those on top. That efficiency is going to revolutionize the economy in ways that are probably impossible to see from a distant, much how somebody living in an agrarian society couldn't have even begun to imagine the 'real' impact of the Industrial Revolution.
- Quoted phrase in the opening sentence
- Opening quote: “save money”
Economic hubs work when there is more efficiency to be found in industrially-codependent resources being located close to each other, than there is in cheaper labor costs outside the hub. Hollywood's heyday was when there was enormous value to putting actors, agents, producers, cameras, sets, editors, VFX, writers all within close driving distance of each other, so that you could organize new movies and shows at the drop of a hat. In a day and age when teenagers are able to use cellphone cameras and pocketmoney to produce stories that are compelling enough to command attention and advertising dollars, the value of that geographic proximity has declined tremendously.
AI didn't build the coffin, but it's been putting nails in it.
- Quoted phrase in the opening sentence
- Opening quote: “real”
- Ordinary term used more than once in the same comment
- “efficiency” appears 3 times
To me, the entertainment industry suffers from a lot of the same problems in most other industries that chase "efficiency", in terms of resources used and profit made. Economies and industries and labor are interdependent and more "alive" than balance sheets and bank accounts.
For the film industry, when you start moving things away from LA to "save money", it has knock on effects that happen over long periods of time. Multiple Quarterly Earnings Reports long, if you will. Local businesses that depend on the film industry close due to less work, so then there is less expertise available. Less expertise means it can command high prices. Suddenly it's too expensive to film in LA altogether. So the rest of the jobs leave too.
This pattern seems super common, not just in the film industry.
It is a narrative that interweaves with AI, too. AI is supposedly an efficiency boom. In the macroeconomy, who is the "efficiency" we are chasing for, who benefits? Who is getting the money that was "saved"? Regular people do not have more more money or time. The "efficiency" is for people who are already very rich, to get richer. Is that worth the price we're paying? I know my answer. The article gets close to it by mentioning the consolidation in Hollywood, but I do not see how we fix this without enforcing our antitrust laws. It is embarrassing what we let rich people do in the name of stock price go up.
- Quoted phrase in the opening sentence
- Opening quote: “You're going to begin to see demand destruction based on the pricing environment.”
I came back to simpler life some years ago. Being close to 50yo punk rocker who every time when asked to reply on Whatsapp answers "I don't have Whatsapp, if you want to call me you should afford to call me without making Zuck richer at your own cost". and people.. are.. bamboozled in most cases and just carry on doom scrolling themselves.
I don't. I finally have time again for movies, books, games and deeper communication with fewer and just occasionally pop up in more... I hope, moderated places like even here (although Hacker News that mixes tech, politics and food links is not exactly Hacker to me anymore and that's where actually some harder moderation would be beneficial here instead of just letting cry outs down voting just because we may have different opinions).
World is screwed as this guy described in his phlog. Hard to believe river can be turned back with a stick as we say in my country... but skipping as much BS as possible and "don't feed the troll" should be rephrased as don't support social media barons and political adversaries who manipulate us via non-moderated pseudo journalism.. not just punishing simple people who just want simple lives.
I personally miss times when netiquette (the one that emerged from early UNIX and USENET communities) was still the thing in the net. At full extend.
- Quoted phrase in the opening sentence
- Opening quote: “ancients”
What they seemingly failed to grasp until more recent times was the slow creep of incremental mild exposure:
However, they were so fond of its diverse uses that they minimized the hazards it posed. Romans of yesteryear, like Americans of today, equated limited exposure to lead with limited risk. What they did not realize was that their everyday low-level exposure to the metal rendered them vulnerable to chronic lead poisoning, even while it spared them the full horrors of acute lead poisoning.
~ Lead Poisoning: A Historical Perspective (1985) https://www.epa.gov/archive/epa/aboutepa/lead-poisoning-hist...- Quoted phrase in the opening sentence
- Opening quote: “GPT OSS 120b, Qwen 3.8 27B”
Openrouter on nitro gets me there. Although the TPS is 5X less still.
I have tried GPT OSS 120b on cerebras and it flies. But it's mostly wrong.
> Brazil’s government announced Friday that annual deforestation in the Amazon fell to its lowest level since records began in 1988 [. . .] During Jair Bolsonaro’s presidency, from 2019 to 2022, Amazon deforestation surged by about 60% [0]
This highlights the greatest difference between these two candidates (imo), one wants to preserve the 'lungs of the Earth', the other wants to rip them out for profit and give the world a middle finger on their way out.
[0] https://apnews.com/article/brazil-amazon-deforestation-lula-...
- Quoted phrase in the opening sentence
- Opening quote: “removes restriction on maximum number of simultaneous NVENC video encoding sessions”
https://github.com/keylase/nvidia-patch
So yeah, buy an Nvidia card then download and run all the software needed to make it work to its full potential so Jensen can buy his 500 millionth leather jacket.
- get the largest PSU you can find (1500W at a minimum, more is better)
- get the ADT PLX Quad PCIe backplane. Run 4 thin (single slot) GPUs inside your main enclosure, stick the ADT interface board in between two of the thin boards (don't stack them close to each other or they won't be able to breathe).
- ensure you have as much or more RAM as you do VRAM
- Ensure your CPU has enough PCIe lanes to run at least five boards in x16 (four for the GPUs in the main enclosure, another for the PLX PCIe switch)
- In the BIOS set BAR1 to 'resizable' and > 4G decoding should be enabled.
- ditch the cables that come with the ADT board, they may work but there are much better ones out there and they're not expensive, this will save you days of troubleshooting
- Build your own enclosure rather than trying to adapt a standard one, it won't work anyway (the board pitch is wrong, as is the placement of the mounting holes).
- Put another four double or triple boards on the backplane.
- Add another 3x 140 mm fan to the front of the case so you get good airflow without the feeling of being next to a Harrier at take-off
- I used 3D printed parts to mount the PSU and the board as well as the fans (see below).
- Make sure the board is properly grounded to the enclosure
- If you're running 3090's or other cards that are P2P capable that were hobbled by NV then look at this https://github.com/aikitoria/open-gpu-kernel-modules driver module.
- Power on the ADT board, wait three seconds then power on the main machine. Enjoy your VRAM :)
If you have more questions please email me and I'll be happy to help out.
> Its limited four-theater release in New York and Los Angeles on October 9 had a respectable opening weekend gross—each screen earning an average of $17,225 for a $68,900 total. Expanding to more theaters in more markets was the logical next step, which its distributor Bleeker Street Films did last Friday, raising the number of screens to 472. But Musk’s wider-debut was an unmitigated flop, netting a paltry $614 per screen, for an abysmal second weekend total of $290,000. Box Office Mojo has the film’s to-date total up to $571,902, as of this article’s publication.
https://gizmodo.com/it-seems-like-the-critical-musk-doc-just...
The article also questions this showbiz411 article
> Showbiz411’s post claims the doc is “gone from almost every theater in America,” save “one or two odd places that didn’t get the memo.” It attempts to substantiate its case by saying the list of theaters on Bleeker Street’s website for Musk showtimes “shows that all 472 were listed as playing the movie today. But no more.” There’s no elaboration as to what that means. The only sort of receipt we’re offered is a blurry, unzoomable screenshot below of… the list of theaters showing the doc prior to the purported change.
> The blog also has an unsourced line that “online, people started commenting earlier this week that shows they’d bought tickets for were canceled without warning.” I was able to find a Reddit post backing that up. But no other outlet of good or ill repute is saying Musk’s been pulled.
I can almost guarantee that someone has built a custom btree for their specific domain which is faster than this; and that they probably have some or all of the techniques used in this library that are the main contributors to performance, as well. Prefetching is probably the main one they'll claim, and I know personally of some proprietary implementations which have it.
I would like to see some more humility, but I suppose that when you've already resigned your own potential to being below an LLM, then what's the point.
It's not possible or useful for them to grow anymore. Some new frontier model next week will outdo whatever ephemerally useful project this person writes, right? That's the logic?
1. If AI would bring great productivity but all that extra value is captured by closed-source model providers, then everyone would switch to open-weights models + bare metal servers, which would non-tech companies to capture most of the value. Your argument that there is great value in the use of AI, but it's all captured by OpenAI and Antrhopic falls flat on it's face here. The market would adapt quickly.
2. OpenAI and Anthropic are still losing money. Your argument is that 100$ of extra value is produced, but the closed-source model providers capture 99.9$ of that because they are "rent extractors". If you're right, and they capture 99.9 of the value created, that's a huge issue because they spend 150$ to create that value. If even after eating the whole (value creation) pie they're still unprofitable, that's the worst news for the AI industry.
For sake of comparison, consider e-commerce. As a proxy for "impact on the economy", we can use Amazon's revenue. They went public in 1997 and that year had revenue of $148M (just 0.14% of Walmart's revenue that year). By 2000, revenue was $2.76B (1.65% of Walmart). Bricks and mortar didn't start feeling anything until probably 2011, when Amazon crossed 11% of Walmart's revenue. And it wasn't until 2025 that Amazon finally surpassed Walmart.
But, if you worked in tech in the e-commerce business, things were already feeling pretty hot by 2007. And at that point in time, the e-commerce sector was still miniscule compared to traditional retail (~4%). Software is very exciting these days and the most fantastic things are happening. They will eventually percolate out to the rest of the economy. But it will take many years before everyone notices the difference.
- Quoted phrase in the opening sentence
- Opening quote: “You can see the computer age everywhere but in the productivity statistics.”
As a non-physicist, QM only clicked for me when I learned the linear algebra of one- and two-qubit systems and observables on those systems [1]. The math is accessible at a pre-calculus level, and the most "counter-intuitive" aspects of QM (interference, the uncertainty principle, entanglement) make perfect sense in this framework. I think it's the best starting point for understanding what QM really says.
[0] That's not to say they aren't important historically or for a complete understanding of QM. I'm talking about conceptual understanding for someone, especially an amateur, encountering QM for the first time.
[1] Chapter 2 of Nielsen and Chuang is what did it for me. It's heavygoing if you've never seen linear algebra before, but I don't know of a clearer explanation of the postulates of quantum mechanics.
When the positrons are at the top, they attract negative charges into the top end-cap. This is called an "image charge". When the positrons are close to the bottom end-cap, the image charge into bottom end-cap. This forms a current, the "image current". The more positrons you have, the larger the image current.
Since the image current (on the order of femtoamperes) has to flow through the large resistor, it is turned into a voltage that can be amplified and measured by cryogenic amplifiers. But it has a second effect: The voltage "back-acts" on the positrons, damping their motion. If you have more positrons in the trap, this damping gets stronger, because each positron experiences the back-action of all other positrons.
In the paper, you can see pictures of the driven axial motion, where you take an RF excitation and sweep the frequency over the resonance, recording the positrons' amplitude as a response. The more positrons you have, the "wider" this resonance, because of the damping effect.
As you reduce the number of positrons by doing a "dip" (making the trap really shallow, hoping that some, but not all, positrons escape), you can see the width getting smaller and smaller in discrete steps. That's how you know you are dealing with individual particles.
I haven't worked on Gabrielse's setup, but I have worked on other, very similar (better, ha!) experiments 10 years ago. I don't see what's so special about this trap, but putting antimatter into it always makes for a nice headline.
Apparently, nanograms / micrograms is enough to create a "antimatter catalyzed" fission/fusion propulsion system with enormous specific impulse / efficacy.
PDF: https://www.niac.usra.edu/files/studies/final_report/1071Bic...
[0] An example would be something like:
"The user presses a button, and the form submits."
There are valid intransitive uses of "submit," but none that make sense in this context. In this case "submit" either requires an object (like "the form submits data"--plus maybe an indirect object, like "the form submits data to the server") or should be swapped for "the form is submitted" if you absolutely don't want to supply an object.
- Ordinary term used more than once in the same comment
- “incorrect” appears 2 times
I'm certain there are folks out there who have a hard time with written English that could benefit from an AI editor, but I would almost always rather read casual English written by a human, even if idiosyncratic and "incorrect" sometimes.
And, to be clear: Hard to read English can never be correct, no matter how carefully it follows grammatical rules, and AI often writes hard to read English.
- Quoted phrase in the opening sentence
- Opening quote: “mind”
LLM-generated text is hard to understand because while it's perfectly good language that follows valid cognitive and informational patterns, it does not reflect the intent or mental model of another mind.
- Ordinary term used more than once in the same comment
- “Gmail” appears 2 times
"Google shipped their" -> "Google shipped its" (I use 'their' to make it clear I'm talking about the people who build features at Google, and not merely about a faceless corporation)
"feature or whatever" -> "feature" ('or whatever' is chosen to tell everyone that I'm not actually sure what the name of the feature is, and I can't really be bothered to figure it out)
"gmail" -> "Gmail" (meh)
"I wrote, it had useful tips" -> "I wrote received useful tips" (if you remove the comma, the pacing changes, the reader doesn't linger on the word 'wrote', and you invite some ambiguity around the writing and the receiving of useful tips. I feel like this suggestion is particularly brain damaged, I simply can't understand my own writing when I make this change)
(and so on. I now write emails in notepad, and copy-paste them into Gmail when I'm done, that's how much the suggestions annoy me).
And pretty much universally, they were awful. It was like somebody had blindly taken an English writing textbook and applied every little helper rule to every clause of every sentence, paying no attention whatsoever to deeper concepts like pacing, drawing attention to key words, and moderating emotion.
And I think that's basically what's going on in all of the frontier models as well. When they are being fine tuned, some reward score is telling them to be concise, not to use filler words, to avoid run-on sentences, etc. And there's no counter-balance in the fine-tuning that looks at the writing and flow as a whole.
My guess is years of RLHF optimized for rater's first impressions of how well-written something was. It uses lists! It has headings! It uses fun phrases like 'smoking gun'! It's long and informative! These things are relatively fast and easy to identify and select as a preference.
But none of this really covers whether the response is _useful for the audience_. If you're not the audience, it can be hard to tell what is and isn't useful.
- Quoted phrase in the opening sentence
- Opening quote: “Skyking”
https://opencollective.com/phpfoundation/transactions?search... and https://opencollective.com/php-foundation-eur/transactions?s...
(seems like "work" selects most of that)
And it's not cheap to pay that many people, especially not full time (which is partly why they don't have more full-time contractors). According to https://opencollective.com/phpfoundation#category-BUDGET they only have a balance $350k currently, sinking.
It'd be really awesome, if more companies would sponsor the PHP foundation (especially with thousands or tens of thousands of dollars per year. A lot of small contributions is nice, but not enough.). The work of a few core maintainers is really crucial to the continued development of PHP.
But I think the hardest thing to replace is Active Directory, Outlook, all the integration with Windows, all the admins certified on Microsoft technology. It's not about the actual application "MS Office", it's about all of IT (that line was not AI).
- Quoted phrase in the opening sentence
- Opening quote: “Move the elephant next to the ball”
- Quoted phrase in the opening sentence
- Opening quote: “regular”
How so? I used both at work and they're functionally the same.
- Quoted phrase in the opening sentence
- Opening quote: “special military operation”
- Opening sentence uses a quote to challenge or distance a term
- Opening sentence challenges the term “special military operation”
- Ordinary term used more than once in the same comment
- “enemy” appears 2 times
It's already happening.
> They are politically and societally paralyzed from building even basic things like datacenters which are essentially just industrial compute factories and the foundation for anything pertaining to data sovereignty and technology sovereignty.
No, they are not entirely blind to the consequences of this buildout and are trying to find a balance between environmental concerns and technological progress.
I agree that we should probably do more but to use words like 'societally paralyzed' is rather uninformed about what actual democracies look like.
edit: never mind, I just checked your comment history, this seems to be some thing that you are repeating over and over and getting the same kind of answers but you just keep doing it.
- Quoted phrase in the opening sentence
- Opening quote: “Europe could never build this!”
Europe would first have to invest "tens if not hundreds of billions" themselves in order to develop any viable replacements and that is very unlikely to ever happen.
They are politically and societally paralyzed from building even basic things like datacenters which are essentially just industrial compute factories and the foundation for anything pertaining to data sovereignty and technology sovereignty.
I find that banking apps deserve that extra bit of anger pointed towards them. (It’s also increasingly hard to find an ATM in many countries. What are banks up to?)
- Quoted phrase in the opening sentence
- Opening quote: “official builds of GrapheneOS”
None. Explicitly banned. And, though I feel bad for saying it, I can't blame them. If some idiot roots their phone and downloads malware, the bank may be responsible for refunding them. I can see why they don't want that headache.
Some. App works, but some functionality is unavailable. I've got a credit card app which works, but refuses to set up virtual cards or use "remember me". That's good enough for my use. Similarly, I had a merchant account which let me see my balance, but I couldn't use my phone as a credit card reader.
Works, but only in main profile. I'd rather keep financial things in a locked profile. But lots of banks disallow that.
Full. Everything works (although you may need to toggle some exploit protection off).
Some banks are reasonably good at letting you pass on technical queries to their developers - mostly the smaller ones. The nice thing about the UK banking sector is that it is trivial to move banks. So I switched accounts to providers which worked with my phone and let the old ones know why I was moving.
Ironic for apps that are often so paranoid about "security" and "hardware attestation"
The substack popup to join and follow the author made me giggle: Why would I join that substack to read that "author" when I already have multiple free AIs that can give me the same take, but without all the waffle?
In the past, I might have signed up to read the author (I have signed up to a few authors before), but if all the author has to offer is basically already available to me for free, why would I join up?