acureau

Maybe the start of a bigger shift to another platform. I'd wager a large portion of the Discord user-base is underage, and they've got nothing but time.

This essay is dripping with self pity. I do not care that you're a woman, I do not care that you have a PhD, I do not care that you used to be a barista. You can be technical if you apply that label to yourself. You cannot force others to apply that label to you, so why define yourself by their perception?

For what it's worth (very little), there are many of us for which the technical label means strictly "expertise in X domain".

Just did the same thing at the office. Nice!

Thanks for your time Tudor, I didn't intend by any means to diminish your work by associating it with your company. I had read the first two chapters of your open book, and I found it useful to understand how the idea of moldable development came about and what problem it was trying to solve.

I think it has finally clicked in my mind. You're building a way to represent systems that is richer than text, which ideally reduces cognitive load when trying to understand or modify them. Since every system is different, this requires custom code, and that's normally expensive. GT is an attempt at reducing the cost of developing custom meta-programming tools. Building systems with rich representations from the very start is moldable development. Is this accurate?

The jargon makes this project really difficult to understand at a glance. I've gathered that this is an environment for the programmatic inspection of existing systems. It is developed by a company that works heavily with legacy code, and they found that writing code to inspect a system was a faster way to answer questions about it than reading the source. That seems like a really cool approach.

I still do not understand what "moldable development" means. To me this implies a different paradigm for building applications, which does not seem to be what's offered. I don't understand what a "micro tool" is, is it a unit of code? Am I missing something here?

Lua fills its niche of no-fluff embeddable scripting language perfectly, the idea is to bind to your own code. It makes sense to me that they wouldn't focus heavily on the standard library. I think that if there is a large enough community of people who want to use Lua as a general purpose scripting language, they should maintain an unofficial stdlib.

This is exactly what makes this scenario so absurd to me. The authors don't even attempt to describe how any of this could realistically play out. They describe sequence models and RLAIF, then claim this approach "pays off" in 2026. The paper they link to is from 2022. RLAIF also does not expand the information encoded in the model, it is used to align the output with a set of guidelines. How could this lead to meaningful improvement in a model's ability to do bleeding-edge AI research? Why wouldn't that have happened already?

I don't understand how anyone takes this seriously. Speculation like this is not only useless, but disingenuous. Especially when it's sold as "informed by trend extrapolations, wargames, expert feedback, experience at OpenAI, and previous forecasting successes". This is complete fiction which, at best, is "inspired by" the real world. I question the motives of the authors.

Completely false. I went to one of the cheapest accredited colleges around, I received the maximum pell grant, and my federal loans covered about 65% of the tuition. I worked full time through college to cover the remaining tuition and books.

I have a similar setup going on. I'm a heavy user of LLMs, but the only time I use the code they generate is for throwaway scripts. I like to describe the problem I'm working on, paste in my code, and ask about everything wrong with it. Am I missing something? Are there glaring flaws or inefficiencies? Are there better ways to approach this? I never take suggestions unless I fully understand and agree with them. There are lots of poor suggestions, but lots of really good ones too.

Infinite tailored critique and advice. I have found this immensely valuable, and I have learned lots doing it. LLMs are static analyzers on steroids.

Well, it's not as if that promise didn't come true. We do have the world's knowledge at our fingertips. It's more of a curation problem, and a natural result of a decentralized web. Wikipedia could be considered a "megasite", but I'd argue it's more valuable than a million independent sites.

Sounds like you just treat programming more like engineering than art. Some art does have source code, there is plenty of room for creative exploration with code.

Wealth and skill are not necessarily proportional.

Zero technical information in the write-up, vague implications, pretty blatant marketing garbage. Who could have guessed that the Chinese platform would communicate with servers operated by one of the largest Chinese telecom companies? Who could have guessed the CCP would have some amount of influence over this company? Truly groundbreaking reporting here.

Replace DeepSeek with Gemini and China with the United States and the post reads just the same.

Intelligent people do not "hallucinate" in the same sense that an LLM does. Counterarguments you don't like aren't "shortcuts". There are certainly obnoxious anti-LLM people, but you can't use them to dismiss everyone else.

An LLM does nothing more than predict the next token in a sequence. It is functionally auto-complete. It hallucinates because it has no concept of a fact. It has no "concept", period, it cannot reason. It is a statistical model. The "reasoning" you observe in models like o1 is a neat prompting trick that allows it to generate more context for itself.

I use LLMs on a daily basis. I use them at work and at home, and I feel that they have greatly enhanced my life. At the end of the day they are just another tool. The term "AI" is entirely marketing preying on those who can't be bothered to learn how the technology works.

Historians, political scientists, and social scientists come to mind

We live very different lives, so take what I say with a grain of salt. I'm of the opinion that you don't really "control your own destiny". Most things happen by chance. Fighting for control is a losing game. If you feel strongly about something, act on it and see where it goes.

I made no point about the state of most Electron apps. What I mean to say is that rather than pivot to a native toolkit, businesses could prioritize improving their use of Electron. These days most of the UI world is focused on web. Using well-established and widely understood tech brings a lot of benefits.

Not to mention that desktop UI development is just unpleasant. I say this as someone who would really prefer that not to be true, it's just that in my search I've been disappointed. You choose between a number of imperfect "cross-platform" frameworks (and deal with their licensing), or maintain native code for several platforms. Creating custom "widgets" often requires you to start at the drawing API level and have a deep understanding of the framework.

I am keeping an eye on Avalonia, though. The tooling wasn't quite there yet when I tried it.

Electron can be done well when treated as any other Desktop UI framework. VSCode is probably one of the only examples of that. I think the bad reputation comes from lazy wrappers pushed out to tick the "Desktop app" box, which are the unfortunate majority in my experience.

I used to do freelance writing. We did this as a way to break up large blocks of text, it makes the post more digestible.

Checking in from Louisiana, can confirm I hear and say "y'all" on a near daily basis.

Same story here, I've got an old laptop and a wireless keyboard / trackpad combo. I haven't touched the TV interface since I set this up.

I hope we get unlimited access for the reasonable price of a car note

Fairly obvious for those who've spent enough time online, I'd say most people would only recognize that first sentence. The Unabomber Manifesto has become something of a copypasta

I like to ask LLMs to critique my code, they often point out a good number of legitimate improvements and oversights.

As someone who has never lived in a city this is strange to me. I live in a relatively dense part of Louisiana, but around here you couldn't walk at all without jaywalking. There aren't even sidewalks in a lot of places. You just walk along the side of the road.

My platform of choice was YouTube until they introduced shorts and took a more aggressive stance against ad-blockers. I quit initially because I was against the direction the platform was heading, but realized quickly that I didn't miss it. It dawned on me that I had been investing 8 - 12 hours of my life weekly to passively consume content that I couldn't even remember. I noticed that at any slight moment of down-time I'd impulsively reach for YouTube, it's almost like I was conditioned to do so.

At the end of the day that's what Google wants. They want you to spend as much of your time as possible watching as many advertisements as possible. Most social media platforms are adversarial. Once I saw this I could not unsee it. Warning others is pointless. They'll be annoyed or just think you're weird. Not one person I know has taken my advice, so I've stopped giving it.

Tangent aside, for me the benefit of not using social media has been that I can invest the time into what I find fulfilling. The quality of the entertainment I consume has gone way up. The downside is that it's isolating. No I can't follow you on platform X, no I cannot view the link you've sent me to platform Y. Everyone, especially in my age group, considers it strange.

At the end of the day switching from an interpreted language to a natively compiled language will result in a faster program. Of course there are performance gains to be had refactoring with a deeper understanding of the problem. That might be enough in many cases, but if the primary goal is speed the language cannot be ruled out.

Or perhaps we as people just tend to idealize foreign cultures, because from our perspective they're different and interesting. You could view it as patronizing or endearing. You don't have to look far to find the exact opposite happening all over the world.

https://en.wikipedia.org/wiki/Westernization

Yep me too, I'm sure there are many things we've subconsciously adapted to using with our right hands. Actually I've never bought a product purpose made for left-handed people.

That's exactly what the article is about actually.