Having claude run the browser and then take a screenshot to debug gives similar results. It's why doing so is useless even though it would be so very nice if it worked.
Somewhere in the pipeline, they get lazy or ahead of themselves and just interpret what they want to in the picture they see. They want to interpet something working and complete.
I can imagine it's related the same issue with LLMs pretending tests work when they don't. They're RL trained for a goal state and sometimes pretending they reached the goal works.
It wasn't the wifi - just genAI doing what it does.
For tiny stuff, they are incredible auto-complete tools. But they are basically cover bands. They can do things that have been done to death already. They're good for what they're good for. I wouldn't have bet the farm on them.
We see that the fastest C# version is 6 times slower than the rust/c++ implementation.
But that's super deceiving because those versions use arena allocators. Doing the same (wrote this morning, actually) yielded a ~20% difference vs the fastest rust implementation.
This was with dotnet 9.
I think the model of using GC by default and managing the memory when it's important is the sanest approach. Requiring everything to be manually managed seems like a waste of time. C# is perfect for managing memory when you need it only.
I like rust syntactically. I think C# is too object-oriented. But with a very solid standard lib, practical design, good tools, and speed when you need it, C# remains super underrated.
> The fact that Rico Mariani was able to do a literal translation of the original C++ version into C# and blow the socks off it is a testament to the power and performance of managed code. It took me several days of painful optimization to catch up, including one optimization that introduced a bug, and then Rico simply had to do a little tweaking with one hand tied behind his back to regain the lead. Sure, I eventually won but look at the cost of that victory
I was actually surprised much of the material still exists - though the links don't work. Microsoft performs so much needless self-vandalism and I know some things I care about are gone.
Which just reminded that yeah, all the links I'd made to Raymond Chen's "The poor man's way of identifying memory leaks" no longer work. The Rust implementation is less than four years old, but its link (which worked) now does not. -sigh-
Tempting to go reconstruct that performance improvement "fight" in Rust too. Maybe another day.
It more or less tells you to unlearn all functional and OOP patterns for code that needs to be fast. Just use regular loops, structs and mutable variables.
Great angle to look at the releases of new software. I, too, thought we'd see a huge increase by now.
An alternative theory is that writing code was never the bottleneck of releasing software. The exploration of what it is you're building and getting it on a platform takes time and effort.
On the other hand, yeah, it's really easy to 'hold it wrong' with AI tools. Sometimes I have a great day and think I've figured it out. And then the next day, I realize that I'm still holding it wrong in some other way.
It is philosophically interesting that it is so hard to understand what makes building software products hard. And how to make it more productive. I can build software for 20 years and still feel like I don't really know.
When working on anything, I am asked: what is the smallest "hard" problem that this is solving ? ie, in software, value is added by solving "hard" problems - not by solving easy problems. Another way to put it is: hard problems are those that are not "templated" ie, solved elsewhere and only need to be copied.
LLMs are allowing the easy problems to be solved faster. But the real bottleneck is in solving the hard problems - and hard problems could be "hard" due to technical reasons, or business reasons or customer-adoption reasons. Hard problems are where value lies particularly when everyone has access to this tool, and everyone can equally well create or copy something using it.
In my experience, LLMs have not yet made a dent in solving the hard problems because, they dont really have a theory of how something really works. On the other hand, they have really helped boost productivity for tasks that are templated .
One of the rebuttals at the end of the post addresses this.
> That’s only true when you’re in a large corporation. When you’re by yourself, when you’re the stakeholder as well as the developer, you’re not in meetings. You're telling me that people aren’t shipping anything solo anymore? That people aren’t shipping new GitHub projects that scratch a personal itch? How does software creation not involve code?
So if you’re saying “LLMs do speed up coding, but that was never the bottleneck,” then the author is saying, “it’s sometimes the bottleneck. E.g., personal projects”
Also when vou create a product you can’t speed up the iterative process of seeing how users want it, fixing edge cases that you only realized later etc. these are the things that make a product good and why there’s that article about software taking 10 years to mature: https://www.joelonsoftware.com/2001/07/21/good-software-take...
This is the answer. Programming was never the bottleneck in delivering software, whether free-range, organic, grass-fed human-generated code or AI-assisted.
AI is just a convenient excuse to lay off many rounds of over-hiring while also keeping the door open for potential investors to throw more money into the incinerator since the company is now “AI-first”.
Just like writing Lord of the Rings is actually not just about typing. You have to live a life, go to war, think deeply for years, research languages and cultures and then one day you type all that out
The point of the form is not in the filling. You shouldn't want to fill out a form.
If you could accomplish your task without the busywork, why wouldn’t you?
If you could interact with the world on your terms, rather than in the enshitified way monopoly platforms force on you, why wouldn't you?
And yeah, if you could consume content in the way you want, rather than the way it is presented, why wouldn’t you?
I understand the issue with AI gen slop, but slop content has been around since before AI - it's the incentives that are rotten.
Gen AI could be the greatest manipulator. It could also be our best defense against manipulation. That future is being shaped right now. It could go either way.
Let's push for the future where the individual has control of the way they interact.
you are getting this from the wrong perspective. I agree what you say here, but things you are listing here implies one thing;
"you didnt want to do this before, now with the help of ai, you dont have to. you just live your life as the way you want"
and your assumption is wrong. I still want to watch videos when it is generated by human. I still want to use internet, but when I know it is a human being at the other side. What I don't want is AI to destroy or make dirty the things I care, I enjoy doing. Yes, I want to live in my terms, and AI is not part of it, humans do.
> I understand the issue with AI gen slop, but slop content has been around since before AI - it's the incentives that are rotten.
Everyone says this, and it feels like a wholly unserious way to terminate the thinking and end the conversation.
Is the slop problem meaningfully worse now that we have AI? Yes: I’m coming across much more deceptively framed or fluffed up content than I used to. Is anyone proposing any (actually credible, not hand wavy microtransaction schemes) method of fixing the incentives? No.
So should we do some sort of First Amendment-violating ultramessy AI ban? I don’t want that to happen, but people are mad, and if we don’t come up with a serious and credible way to fix this, then people who care less than us will take it upon themselves to solve it, and the “First Amendment-violating ultramessy AI ban” is what we’re gonna get.
Oh come on, are you 12? Real life doesn’t have narrative arcs like that. This is a real problem. We’re not gonna just sit around and then enjoy a cathartic resolution.
(Maybe skip the mini-insults & make the site nicer for all?)
Anyway I think GP has a point worth considering. I have had a related hope in the context of journalism / chain of trust that was mentioned above: if anyone can produce a Faux News Channel tailored to their own quirks on demand, and can see everyone else doing the same, will it become common knowledge that Stuff Can Be Fake, and motivate people to explicitly decide about trust beyond "Trust Screens"?
A more detailed analogy would be if you owning the robots meant that all food is now packaged for robots instead of humans, increasing the personal labor cost of obtaining and preparing food as well as inflating the cost of dinnerware exponentially, while driving up my power bill to cover the cost of expanding infrastructure to power your robots.
In that case, I certainly am against you owning the robots and view your desire for them as a direct and immediate threat against my well being.
And therein is the problem - if your robots take up so many resources I can't have my dishwasher, is that your right? Is your right to being happy more important than others?
The problem of resource distribution is solved by money already.
If I can't pay for the robots, I am not getting them. And if I buy my robots and you only get a dishwasher then you can afford two nice vacations on top while I don't.
Let's say we have a finite amount of cheap water units between us. After exhausting those units, the price to acquire more goes up. Each our actions use up those units.
If restrictions on water use do not exist, you can quickly use up those units and, if you can easily afford more units, which makes sense as you have enough for robots, you are not concerned with using that cheap water up.
I can't even afford to "toil" with my dishwasher now.
When two entities control essentially the whole "market" for mobile OSes and associated app stores, and use their position to force their app stores on everyone, you no longer have a market. If we just forcibly split Google and Apple into smaller companies with separate app stores then maybe we could see what markets would do.
But the problem is that when everyone has gravitated to the two biggest app stores, there's no scope for market competition. The point is not just to have third-party app stores, it's to not have huge app stores that capture the whole user base.
"There is no market effect"? Why do the market effects disappear if some of the players don't play completely according to the desires of other players? Why couldn't it be that the optimum includes some amount of fee dodging?
You didn't answer my question. Why can't the optimum include some fee dodging? I agree that the app stores provide value to the developers, but it's also true that apps provide value to Google and Apple. If no one developed any apps, no one would use iOS or Android. Therefore it's possible that Google and Apple benefit more from an app that dodges fees but brings in users than from neither having the app nor those users.
They created a market and they are charging to be on the market. If you're saying "I want to be on the market but by my own rules" that's not a free market effect, it's breaking a contract.
If you say "fine, I'll go on another app store" then that is a free market in action - but good luck getting anyone to download your game.
Is there a reason why prompt injections in general are not solvable with task-specific layering?
Why can't the llm break up the tasks into smaller components. The higher level task llm context doesn't need to know what is beneath it in a freeform way - it can sanitize the return. This also has the side effect of limiting the context of the upper-level task management llm instance so they can stay focused.
I realize that the lower task could transmit to the higher task but they don't have to be written that way.
The argument against is that upper level llms not getting free form results could limit the llm but for a lot of tasks where security is important, it seems like it would be fine.
But it can be tricked into delegating incorrectly - for example, to the "allowed to use confidential information" agent instead of the "general purpose" agent
I am also on Wayland (via CachyOS) and recently started using Deskflow for my MacBook KVM-type setup. Linux on one screen, Mac on the other.
Works great except for the copy-and-paste issues and the mod key resets - just as you describe. I've been fiddling with it the last couple of days.
Funny to see someone with such a similar setup and issue.
Esoteric issues aside, the general use of CachyOS has been very smooth.
There seems to be a lot of momentum for desktop Linux, with Microsoft and Apple dropping the ball lately. LLMs also make deep configuration of Linux more manageable. The stars seem to be aligning for it.
>You can't build anything special with someone who keeps forgetting any context. I spent many years cycling between depression and resurrected determination trying. But finally gave up.
Was that an LLM reference or is it the myopia in me?
There's a parallel here, either way. All the documentation in the world will not make a person, or llm session interchangable.
In some sense the new way of coding feels like building a big org with people without memory. If you can document the process perfectly, there is a holy grail out there somewhere.
Having claude run the browser and then take a screenshot to debug gives similar results. It's why doing so is useless even though it would be so very nice if it worked.
Somewhere in the pipeline, they get lazy or ahead of themselves and just interpret what they want to in the picture they see. They want to interpet something working and complete.
I can imagine it's related the same issue with LLMs pretending tests work when they don't. They're RL trained for a goal state and sometimes pretending they reached the goal works.
It wasn't the wifi - just genAI doing what it does.