> the ability to quickly and safely change significant parts of the code and product.
Hum, this reminds me something... "O: open to extension, closed to modifications". Old things are new again.
From context efficiency with approaches such ad DDD and clean architectures, all the way to items such as this one, AI is not creating new tradeoff, it just acts as a multiplier, multiplying productivity for teams doing things right, and multiplying debt for teams having a low quality bar as far as design and architecture are concerned.
Exactly, it's funny how most Americans have no self-awareness on this topic.
And beyond VCs, which are like massive subsidies funded by printed dollars to which no other country has access, even in industries like electric vehicles, Chinese total direct subsidies to their EV companies are like $5bn per year, while the the ones provided by the US to their auto manufacturers are in the range of $50bn per year.
I don't think the US are cheaters or are doing something bad. But i do think that this propaganda about China flooding the market through "overcapacity" and subsidies is very dishonest and needs to stop.
Your hostile tone is unfortunate, especially since my post was actually friendly. I was just trying to point why it is very likely the OP won't give you what you're asking so you're not left confused if he ends up ghosting you.
Many people use the term harness to refer to the agent coding software (eg. Opencode, Claude Code...), i use this term more broadly to refer to the environment (set of skills, system prompts, constraints, memory, hooks etc...). What the OP is referring to is not just one giant skill. It's usually a comprehensive ecosystem of skills, bespoke tools to make certain agent tasks deterministic (eg localization), and so on.
I've seen someone post Github repos in this thread, these can be very useful especially if you use the same tech stack, but you won't reach the level of productivity reported by successful teams unless you invest substantial time to build your own harness. But the way to do so is to do it progressively : start with something simple to address the need you have on day 1 . And then, turn recurring prompts into skills, turn recurring coding patterns and coding style recommendations into guidelines, turn repetivive tasks for which the LLM tends to build a python script that it occasionally gets wrong into a deterministic tool documented in a skill etc...
And after a couple of days, weeks, and months, you'll have a very dependable harness giving you optimal productivity, without needing to invest weeks of work upfront or take the fun out of agent-assisted coding.
That's a big ask. This kind of harness usually contains plenty of proprietary insights about their business. And also, nowadays, a good harness is a major competitive advantage.
As someone who uses quite a couple of different AI providers (codex, glm, deepseek, claude premium among others), i've noticed that claude tends to move too fast and execute commands without asking for permission.
For example, if i ask a question regarding an implementation decision while it is implementing a plan, it answers (or not) and immediately proceeds to make changes it assumes i want. Other models switch to chat mode, or ask for the best course of action.
Once this is said, i am not blaming Anthropic
For that one, because IMHO the OP has taken a lot of risks and failed to design a proper backup and recovery strategy. I wish them to recover from this though, this must be a very stressful situation for them.
Sometimes we STEM folks are way too rigid, I obviously meant "IN MY OPINION, GLM models are at this point superior to...".
I do not think that anyone who read my comment understood it differently. But I grant you this point, this is just my opinion based on my personal experience not the result of a scientific study.
Once this is said, i wasn't submitting a scientific paper for preprint, just posting my opinion on an internet forum.
Not sure why you are making such a big deal out of it, especially for something for which people can decide within minutes if it works for them or not. And I haven't seen you nitpick on other people saying that all Chinese models are garbage incapable of doing even the most basic task, without quoting any study. This kind of scrutiny tends to be one-sided.
Edit: and regarding what the z.ai team is saying about their models, just check their Discord and the articles they link there. They themselves say that their latest models have leading performance on a number of aspects. It is misleading to suggest that the authors of the model are not proudly saying that their models have best in class performance.
I know my use case and my personal experience :) i am not trying to pretend that it is the best in benchmarks, just sharing my experience so people know that some folks are having a very good experience with GLM models, compared to the competition.
My only goal is to encourage people to try it out so they can see if it moves the needle for them, because there are fair chances that it will. I am not trying to start a flamewar or something.
I appreciate that it's not working for your use case but it's unfortunate that you dismiss the experience of others. And i am not chinese, I am European. Thanks for your feedback anyway.
I'e been using their models pretty much daily for the past 2 months to work on the codebase of a very complex B2B2C platform written in an unusual functional language (F#) with an angular frontend.
I also use Claude premium daily for another client, and i use Codex. and i can tell you that GLM5 is at this point much more capable than Claude and Codex for complex backend end work, complex feature planning, and long horizon tasks. One thing i've noticed is that it is particularly good at following instructions and guidelines, even deep into the execution of a plan.
To me the only problem is that z.ai have had trouble with inference : the performance of their API has been pretty poor at times. It looks like this is an hardware issue related to the Huawei chips they use rather than an issue with the model itself. The situation has been substantially improving over the past few weeks.
GLM5.1, GLM5-Turbo and GLM5v are at this point better than Opus, Codex, Gemini and other claude source models. We have reached a major turning point. To me, the only closed source model still in the game is codex as it is much faster at executing simple tasks and implementing already created plans.
Try GLM5v for your PDF work, it's their last generation vision model that has been released a couple of days ago.
Except it does not work that way in the US, you can freely incorporate in any state without worrying about this kind of tax drama. The EU really needs to improve the integration of their single market, as this is precisely the kind of barrier preventing people from exploring what other EU states have to offer.
I think you are making a big deal out of nothing: today adoption is tiny but there isn't any coordination problem. Anyone wanting to buy or sell bitcoins, even for very large amount has no problem doing so.
As adoption increases, liquidity (which may be what you allude to when you talk about coordination) will be even better.
I'd also note that facilitating liquity can indeed be a complex problem, but there are people specialized in making money out of it and who therefore have an incentive to continually keep the market liquid, read about "market makers".
It does not need to be all or nothing though. It simply needs to be true for a large enough number of people, to create an extremely valuable market.
We are over 8 billion on earth, even if as little as 1% of the population rely on it directly or indirectly due to some form of oppression real or perceived due to political opinions, a long investment thesis on bitcoin can make sense.
I am not a bitcoin holder once this is said. Arrived at the party too late IMHO, and i have better investment options for now from a risk-adjusted perspective.
However if i had more cash to invest than i know what to do with, i'd definitely take a long position on bitcoin.
I was referring to the need to be careful to not modify/update packages also used by the base system. Since all packages are treated the same on Linux, you often can't tell which package can put you in trouble if you update it along with the dependencies it drags with it.
This kind of problem happens frequently when users add repositories such as Packman on Linux providing dependencies versions different from the ones used by the base system of the distro.
Experienced people know how to avoid these mistakes, but this whole class of problem does not exist on FreeBSD.
The key thing is that on freebsd you do not risk bricking your system by installing a port. Even though this guarantee has become less true with PkgBase
Done the same since 2018 circa, never looked back.
For a while even used it on the desktop, but was too much trouble due to specific tools we need that weren't supported properly. so we're using Linux on the desktop.
FreeBSD is stable, lightweight, gets out of the way, and without drama.
Yes you are right, it does not properly support NativeAOT yet.
But it isn't a need for most use cases, unless you want to do mobile development and meet app store policies. But even then, mature F# frameworks like Fable transpile your F# code to React & Cie.
What you are looking for is called F#. You get native interop with C# and access to all .NET/C# libraries as a bonus. We use it as a daily driver for a complex B2B2C cloud platform.
Since storage is a critical component, I closely watched it and engaged with the project for about 2 years circa as i contemplated adding it to our project, but the project is still immature from a reliability perspective in my opinion.
No test suite, plenty of regressions, and data loss bugs on core code paths that should have been battled tested after so many years. There are many moving parts, which is both its strength and its weakness as anything can break - and does break. Even Erasure Coding/Decoding has had problems, but a guy from Proton has contributed a lot of fixes in this area lately.
One of the big positive in my opinion, is the maintainer. He is an extremely friendly and responsive gentleman. Seaweedfs is also the most lightweight storage system you can find, and it is extremely easy to set up, and can run on servers with very little hardware resources.
Many people are happy with it, but you'd better be ready to understand their file format to fix corruption issues by hand. As far as i am concerned, i realized that after watching all these bugs, the idea of using seaweedfs was causing me more anxiety than peace of mind. Since we didn't need to store billions of files yet, not even millions, we went with creating a file storage API in ASP.NET Core in 1 or 2 hours, hosted on a VPS, that we can replicate using rsync without problem. Since i made this decision, i have peace of mind and no longer think about my storage system. Simplicity is often better, and OSes have long been optimized to cache and serve files natively.
If you are not interested in contributing fixes and digging into the file format when a problem occurs, and if your data is important to you, unless you operate at the billions of files scalability tier Seaweedfs shines at, i'd suggest rolling your own boring storage system.
Hum, this reminds me something... "O: open to extension, closed to modifications". Old things are new again.
From context efficiency with approaches such ad DDD and clean architectures, all the way to items such as this one, AI is not creating new tradeoff, it just acts as a multiplier, multiplying productivity for teams doing things right, and multiplying debt for teams having a low quality bar as far as design and architecture are concerned.