Have you tested alternatives? I grabbed Open Code and a Minimax m2.1 subscription, even just the 10usd/mo one to test with.
Result? We designed a spec for a slight variation of a tool for which I wrote a spec with Claude - same problem (process supervisor tool), from scratch.
Honestly, it worked great, I have played a little further with generating code (this time golang), again, I am happy.
I've been using GLM 4.7 with Claude Code. best of both worlds. Canceled my Anthropic subscription due to the US politics as well. Already started my "withdrawal" in Jan 2025, Anthropic was one of the few that was left
Are you using an API proxy to route GLM into the Claude Code CLI? Or do you mean side-by-side usage? Not sure if custom endpoints are supported natively yet.
I think the least politically divisive issue within the US is concern about China’s growth as it directly threatens the US’s ability to set the world’s agenda. It may be politically divisive if you are aligned with Chinese interests but I don’t see anything politically divisive for a US audience. I expect Chinese CEOs speak in similar terms to a Chinese audience in terms of making sure they’re decoupled from the now unstable US political machine.
From the perspective of competing against China in terms of AI the argument against open models makes sense to me. It’s a terrible problem to have really. Ideally we should all be able to work together in the sandbox towards a better tomorrow but thats not reality.
I prefer to have more open models. On the other hand China closes up their open models once they start to show a competitive edge.
With a good harness I am getting similar results with GLM 4.7. I am paying for TWO! max accounts and my agents are running 24/7.
I still have a small Claude account to do some code reviews. Opus 4.5 does good reviews but at this point GLM 4.7 usually can do the same code reviews.
If cost is an issue (for me it is, I pay out of pocket) go with GLM 4.7
Your GitHub profile is... disturbing. 1,354 commits and 464 pull requests in January so far.
Regardless of how productive those numbers may seem, that amount of code being published so quickly is concerning, to say the least. It couldn't have possibly been reviewed by a human or properly tested.
If this is the future of software development, society is cooked.
Yes, the idea is to really, fully automate software engineering. I don't know if I am going to be successful but I'm on vacation and having fun!
if Opus 4.5/GLM 4.7 can do so much already, I can only imagine what can be done in two years. Might as well adopt to this reality and learn how leverage this advancement
I've been using GLM 4.7 alongside Opus 4.5 and I can't believe how bad it is. Seriously.
I spent 20 minutes yesterday trying to get GLM 4.7 to understand that a simple modal on a web page (vanilla JS and HTML!) wasn't displaying when a certain button was clicked. I hooked it up to Chrome MCP in Open Code as well.
It constantly told me that it fixed the problem. In frustration, I opened Claude Code and just typed "Why won't the button with ID 'edit' work???!"
It fixed the problem in one shot. This isn't even a hard problem (and I could have just fixed it myself but I guess sunk cost fallacy).
I've used a bunch of the SOTA models (via my work's Windsurf subscription) for HTML/CSS/JS stuff over the past few months. Mind you, I am not a web developer, these are just internal and personal projects.
My experience is that all of the models seem to do a decent job of writing a whole application from scratch, up to a certain point of complexity. But as soon as you ask them for non-trivial modifications and bugfixes, they _usually_ go deep into rationalized rabbit holes into nowhere.
I burned through a lot of credits to try them all and Gemini tended to work the best for the things I was doing. But as always, YMMV.
Amazingly, just yesterday, I had Opus 4.5 crap itself extensively on a fairly simple problem -- it was trying to override a column with an aggregation function while also using it in a group-by without referring to the original column by its full qualified name prefixed with the table -- and in typical Claude fashion it assembled an entire abstraction layer to try and hide the problem under, before finally giving up, deleting the column, and smugly informing me I didn't need it anyway.
That evening, for kicks, I brought the problem to GLM 4.7 Flash (Flash!) and it one-shot the right solution.
It's not apples to apples, because when it comes down to it LLMs are statistical token extruders, and it's a lot easier to extrude the likely tokens from an isolated query than from a whole workspace that's already been messed up somewhat by said LLM. That, and data is not the plural of anecdote. But still, I'm easily amused, and this amused me. (I haven't otherwise pushed GLM 4.7 much and I don't have a strong opinion about about it.)
But seriously, given the consistent pattern of knitting ever larger carpets to sweep errors under that Claude seems to exhibit over and over instead of identifying and addressing root causes, I'm curious what the codebases of people who use it a lot look like.
This has been my consistent experience with every model prior to Opus 4.5, and every single open model I've given a go.
Hopefully we will get there in another 6 months when Opus is distilled into new open models, but I've always been shocked at some of the claims around open models, when I've been entirely unable to replicate them.
Hell, even Opus 4.5 shits the bed with semi-regularity on anything that's not completely greenfield for my usage, once I'm giving it tasks beyond some unseen complexity boundary.
I use Opus 4.5 for planning, when I reach my usage limits fallback to GLM 4.7 only for implementing the plan, it still struggles, even though I configure GLM 4.7 as both smaller model and heavier model in claude code
The Chinese labs distill the SOTA models to boost the performance of theirs. They are a trailer hooked up (with a 3-6 month long chain) to the trucks pushing the technology forwards. I've yet to see a trailer overtake it's truck.
China would need an architectural breakthrough to leap American labs given the huge compute disparity.
Agreed. I do think the metaphor still holds though.
A financial jackknifing of the AI industry seems to be one very plausible outcome as these promises/expectations of the AI companies starts meeting reality.
Care to explain how the volume of AI research papers authored by Chinese researchers[1] has exceeded US-published ones? Time-traveling plagiarism perhaps, since you believe the US is destined to lead always.
1. Chinese researcher in China, to be more specific.
Not a great metric, research in academia doesn't necessarily translate to value. In the US they've poached so many academics because of how much value they directly translate to.
I don't doubt China wouldn't be capable of making SOTA models, however they are very heavily compute constrained. So they are forced to shortcut compute by riding the coattails of compute heavy models.
They need a training-multiplier breakthrough that would allow them to train SOTA models on on a fraction of the compute that the US does. And this would also have to be kept a secret and be well hidden (often multiple researchers from around the world put the pieces together on a problem at around the same time, so the breakthrough would have to be something pretty difficult to discover for the greatest minds in the field) to prevent the US from using it to multiply their model strength with their greater compute.
Perhaps you should pay attention to where the puck is going to be, rather than where it is currently. Lots of original ideas are coming out of Chinese AI research[1], denying this betrays some level of cope.
1. e.g. select any DeepSeek release, and read the accompanying paper
I'll pay attention to where the puck is because that is something I can observe, where it is going to be is anybody's guess. Lots of original ideas are coming out of Chinese AI research but there is also lots of junk. I think in the longer term they will have the advantage but right now that simply isn't the case.
Your 'cope' accusation has no place here, I have no dog in the race and do not need to cope with anything.
I will rephrase my statement and continue to stand by it: "Denying the volume of original AI research being done by China - a falsifiable metric - betrays some level of cope."
You seem to agree on the fact that China has surpassed the US. As for quality, I'll say expertise is a result of execution. At some point in time during off-shoring, the US had qualitatively better machinists that China, despite manufacturing volumes. That is no longer the case today - as they say, cream floats to the top, and that holds true for a pot or an industrial-sized vat.
I could say the same about grok (although given there are better models for my use cases I don't use it). What part of divisive politics are you talking about here?
> Civic institutions - the rule of law, universities, and a free press - are the
backbone of democratic life
It probably was in 1850-1950s, but not in the world I live today.
Press is not free - full of propaganda. I don't know any journalist today I can trust, I need to check their affiliations before reading the content, because they might be pushing the narrative of press owners or lobbies
Rule of law? don't make me laugh, this sounds so funny, look what happened in Venezuela, US couldn't take its oil, so it was heavily sanctioned for so many years, then it still couldn't resist the urge to steal it, and just took the head of the state.
Universities - do not want to say anything bad about universities, but recently they are also not good guys we can trust, remember Varsity Blues scandal? https://en.wikipedia.org/wiki/Varsity_Blues_scandal - is this the backbone of democratic life?
The alternative to all of these institutions is currently social media, which is worse by any metric: accuracy, fairness, curiosity, etc.
I am more optimistic about AI than this post simply because I think it is a better substitute than social media. In some ways, I think AI and institutions are symbiotic
Go on X. Claims are being fact checked and annotated in real time by an algorithm that finds cases where ideologically opposed people still agree on the fact check. People can summon a cutting edge LLM to evaluate claims on demand. There is almost no gatekeeping so discussions show every point of view, which is fair and curious.
Compare to, I dunno, the BBC. The video you see might not even be real. If you're watching a conservative politician maybe 50 minutes were spliced out of the middle of a sentence and the splice was hidden. You hear only what they want you to hear and they gatekeep aggressively. Facts are not checked in real time by a distributed vote, LLMs are not on hand to double check their claims.
AI and social media are working well together. The biggest problem is synthetic video. But TV news has that problem too, it turns out. Just because you hear someone say some words doesn't mean that was what they actually said. So they're doing equally badly in that regard.
Last time I went on X my feed which I curated from ML contributors and a few politicians had multiple white nationalist memes, and engagement slop. Fact checks frequently are added after millions of impressions.
I am sure there are very smart well meaning people working on it but it certainly doesn’t feel better than the BBC to me. At least I know that’s state media of the UK and when something is published I see the same article as other people.
>"but it certainly doesn’t feel better than the BBC to me"
BBC was cutting-edge for creating and fostering methodologies that went on to become most of the "impartial reporting" practices from journalists. So, even if it's not feeling any "better" than BBC, that's still a pretty good step in the right direction!
The parent poster was saying X was better than the BBC, I certainly wouldn't have picked that one, but it's likely because they get their news from conservative outlets outraged by the recutting of Trump's speech on January 6th.
That phrasing sounds like you're not yourself outraged by it. It wouldn't be surprising given the institutional attitudes seen at the BBC (and Channel 4 which got caught doing something even worse) - clearly, leftists have decided that framing politicians and publishing entirely fake news is acceptable if it's to attack right wing people.
Anyone who knows about that event and is still watching the BBC afterwards is saying they don't care about the truth of their own beliefs. Dangerous stuff.
>So, even if it's not feeling any "better" than BBC, that's still a pretty good step in the right direction!
The step in the direction of decentralized filter bubbles isoating society? With no channels to hold info accountable and checked/upfated for accuracy?
GP made a pretty good case for X being a good-faithed attempt at a new distributed structure for mass media that at least TRIES to have conflicting viewpoints or objectivist "fact checks", even if it occasionally misses the mark. I was VERY early on the "hate-Elon" bandwagon and even earlier on not being an active Twitter/X user (search my username).
In a post-Fairness Doctrine world, what else would satisfy you?
I don't think we're in a post fairness doctrine world, for one. So no, I haven't given up on the idea of he 4th estate. Your solution to bias is, as always, to not take any one source for granted. Take time to actually read articles from multiple angles that fall in line with the Fariness Doctrine. Then from there, use your own lived experiences to form your own viewpoint.
Outsourcing that to soundbites from randos on twitter with middle school lieracy is insanity. But let me use a charitable lens here.
Any notion of X being a good faith attempt at being a community-lead fact checker got broken with the introduction of Grok. Then those hopes were shattered to pieces when Grok was shown to be massively compromised by yet another central figure. One who, yes, has the literacy of a middle schooler. We somehow ended up with the worst of both worlds having centralization of a bad knowledge hub and stupidity.
>what else would satisfy you?
if using our brains is out of the equation and lack of censorship is truly the most important metric of "free discussion": let's just bring back 4chan. no names or personalities, 99% free-for-all, it technically has threading support to engage in conversations. There is centralization, but compared to the rest of the internet the moderators and admins stay very quiet.
There's a lot I hate about modern social media, but surprisingly 4chan only has like 2 things I strongly dislike. Big step up from the 20+ reasons I can throw at nearly every other site.
As I said, AI is better than social media. AI is trained on and references original sources, which makes it better than reading and believing random posts.
AI is also trained on "random posts". Google made an 11 figure deal with Reddit for this sake.
The biggest factor of social media is being able to curate personalities you go to for whatever reason. If you care about reason you will find the reasonable writers. This also enables disinformation, but people looking for anything to fit what they want to hear wouldn' fo towards the reasonable writers anyway.
1. It censors some topics. Just for fun, try to write something about Israel-Gaza, or try to praise Russia and compare the likes/views with your other posts and over the next week observe how these topics is impacting your overall reach even in other topics.
2. X amplifies your interests, which is not objectively true, so if you are interested in conspiracy or Middle East, it pushes you those topics, but others see different things. Although its showing you something you are interested in, in reality its isolating you in your bubble.
1. Are those topics being censored? You don't seem to know that is true, you're just making assumptions about what reach should be. They open sourced the ranking algorithm and just refreshed it - can you find any code that'd suppress these topics?
2. The media also amplifies people's interests which is why it focuses on bad news and celebrity gossip. How is this unique to social media? Why is it even bad? I wouldn't want to consume any form of media that deliberately showed me boring and irrelevant things.
BBC for all its faults is definitely better then Musks X when it comes to truth.
> Claims are being fact checked and annotated in real time by an algorithm that finds cases where ideologically opposed people still agree on the fact check. People can summon a cutting edge LLM to evaluate claims on demand. There is almost no gatekeeping so discussions show every point of view, which is fair and curious.
>I am more optimistic about AI than this post simply because I think it is a better substitute than social media. In some ways, I think AI and institutions are symbiotic
If I made a pitch for a cyberpunk dystopia where knowledge is centralized by a for-profote corporate trillionaire, I'd get a resounding yawn for originality. Yet here we are vouching for that in real time.
Social media has a lot of noise, but people understand not to take poisonshadow_42 as a central hub of general knowledge. That is sadly not the case with Grok, despite its obvious, blatant abuse of such a title.
Paradoxically, these institutions are probably the best they've ever been. We trusted them more 100 years ago because we didn't know better, but we're now letting perfect be the the enemy of good. Wise men once said:
"In prison, I learned that everything in this world, including money, operates not on reality..."
"But the perception of reality..."
Our distrust of institutions is a prison of our own making.
I can't speak for the other institutions but I'd be shocked if the press, as an institution, is the best it's ever been. I know a lot of people who left that industry because of the way that the Internet and social media eroded the profitability of reporting while pushing on virality, articles were tuned to declining attention spans, outlets leaned more on centralized newswire services, and local reporting collapsed nearly to zero.
I think the press, as an institution, was at its peak post-Watergate, and pre- ... something. I don't know when exactly the press began to decay; possibly with the rise of 24-hour cable news in the 1990s; maybe the elimination of the Fairness Doctrine in 1987, maybe the Telecommunications Act of 1996. The media landscape was certainly severely decayed by 2003, and has not gotten any better.
pre-24 hour news cycle. When you need to generate news around the clock at any and all times, you quickly run out of interesting topics. So the best solution is to report on every tiny morsel of information you can grab on whatever story gets the most eyeballs.
>do not want to say anything bad about universities
I'll make a slightly warm take: Co-opting our higher education institutions to be used as an extended job pipelne was a huge mistake. Your primary goal for attending college should not be to prepare for a job unless you are aiming for a highly specialized position.
Hotter take: jobs above a certain size should require a 3 month onboarding pipeline that is demonstrably used if they want to make the argument of hiring H1-B's. If you can learn the job in that period, it's clear that there is domestic talent.
The press has always been full of propaganda, it's just that in the time period 1850-1950 there weren't any dissenting media outlets so it was impossible for anyone to recognize that there was anything different from the propaganda
Every society is going to have problems. Democracy's benefit is that it allows those problems to be freely discussed and resolved
My (non-authoritative) understanding was that after Vietnam there was a more recognised need to control what the media published, resulting in Operation Mockingbird and such. However, given how centralised the media has always been, I could see it being influenced before this.
I really shouldn't be so gobsmacked by people's ignorance of history, but it is astounding to me the number of replies here that seem to believe that the press really was well-behaved in this time period. When learning about the Spanish-American War, pretty much the most important bullet point covered in history class is the role of the press in essentially inventing the cause of the war, as exemplified by the infamous quote from a newspaper baron: "You furnish the pictures and I'll furnish the war."
The general term to look up is "yellow journalism."
I don't think, but I feel like situation was slightly better for some reasons:
* there were no internet, so local communities strived to inform things happening around more objectively. Later on, there were no need for local newspapers
* capitalism was on the rise and on its infancy, but families with a single person working could afford some of the things (e.g. house, car) hence there were no urgent need to selling out all your principles
* people relied on books to consume information, since books were difficult to publish and not easy to revert (like removing a blog post), people gave an attention to what they're producing in the form of books, hence consumers of those books were also slightly demanding in what to expect from other sources
* less power of lobby groups
* not too many super-rich / billionaires, who can just buy anything they want anytime, or ruin the careers of people going against them, hence people probably acted more freely.
But again, can't tell exactly what happened at that time, but in my time press is not free. That's why I said "probably"
> * not too many super-rich / billionaires, who can just buy anything they want anytime, or ruin the careers of people going against them, hence people probably acted more freely.
The provided timespan encompasses the 'gilded age' era, which saw some ridiculous wealth accumulation. Like J.P. Morgan personally bailed out as the US Treasury at one point.
Much of antitrust law was implemented to prevent those sorts of robber baron business practices (explicitly targeting Rockefeller's Standard Oil), fairly successfully too. Until we more or less stopped enforcing them and now we're largely back where we started.
I would disagree about capitalism being on the rise. Marx and his views grew after the 1850s and communist / socialist revolutions spread throughout Europe. There may have been more discussion of "capitalism" and an increase in industrialization, but "capital" had existed and operated for centuries before that. What changed was who owned the capital and how it was managed, specifically there has been a vast increase in central / government control.
I think this centralization of authority over capital is what has allowed for the power of lobbying, etc. A billionaire could previously only control his farms, tenant farmers, etc. Now their reach is international, and they can influence the taxing / spending the occurs across the entire economy.
Similarly, local communities were probably equally (likely far more) mislead by propaganda / lies. However, that influence tended to be more local and aligned with their own interests. The town paper may be full of lies, but the company that owned the town and the workers that lived there both wanted the town to succeed.
He predicted capitalisms fall, (which happened in the 1930s) but didn't predict that instead of the workers uniting and rising against the bourgeoisie that the bourgeoisie would just rebuild it and continue oppressing the masses
Capital continued to function just fine through the 1930s. Crops still grew on land. Dams produced electricity. Factories produced cars. What exactly failed?
Capitalism is subject to periodic crises; the Great Depression of the 30s beginning with the stock market crash of 1929 was the largest of those at the time it happened.
Yes, at the very least there wasn't strong polarization, so the return on propaganda content is lower. Now a newspaper risk losing their consumer more if they publish anything contrarian.
Publishing something to the contrary of popular belief is not being contrarian. It is not a virtue to be contrarian and forcing a dichotomy for the sake of arguing with people.
They are (part of) the backbone of democratic life. But democratic life hasn't been doing well in the US in the last decades. The broken backbone is both cause and symptom of this in a vicious cycle
> Venezuela, US couldn't take its oil, so it was heavily sanctioned for so many years, then it still couldn't resist the urge to steal it, and just took the head of the state.
Could you provide supporting evidence for your statement?
The press has never been believable. How many innocent people were beat, framed and shot and the press just took the word of the police? Rappers in the 80s were talking about police brutality. But no one believed them until the Rodney King video in 1992. Now many don’t instinctively trust the police because everyone has a camera in their pocket and publish video on social media.
On the other side of the coin, the press and both parties ignored what was going on in rural America until the rise of Trump
If anyone has already read this book, can you share your thoughts on how does it compare to software engineering, do you see parallels, are they applicable and so on
One thing bothers me a lot is, if government organizations can be influenced by lobby groups, which primarily owned by corporations, then why do we need government?
Corporations and governments should be considered as balancing forces, one works to increase its profits by any means, other works to protect humans living in that area by any means.
You might say, corporations benefit its employees, true, but it is a small subset of people living in the country. If you allow everything to corporations, they will set up a slavery system from the birth of a baby
Lobbying having undue influence is entirely a government problem that it needs to fix. It will never be fixed while people have the attitude that lobbyists are the problem. (I'm not saying you're saying this; just making a statement.)
This shows why Machado didn't deserve the prize in the first place.
She was asking for military intervention to own nation, in other words, she wanted people get killed because she wanted to rule the government and she was ready for every possible brutal solution
Even worth, Nobel committee lost its credibility, subsequently diminishing the importance of Nobel prize itself
would be nice to see sample conversation shared publicly to understand what it exactly generates. It could be even good for your SEO if you share couple of them
Realistically, PMs incentives are often aligned elsewhere.
But even if a PM cares about UX, they are often not in a good position to spot problems with designs and flows they are closely involved in and intimately familiar with.
Having someone else with a special perspective can be very useful, even if their job provides other beneficial functions, too. Using this "resource" is the job of the PM.
How can a PM do their job if they don't *care* about UX?
I mean... I know exactly happens because I've seen it more than once: the product slowly goes to shit. You get a bunch of PMs at various levels of seniority all pursuing separate goals, not collaborating, not actually working together to compose a coherent product; their production teams are actively encouraged to be siloed; features collide and overlap, or worse conflict; every component redefines what a button looks like; bundles bloat; you have three different rendering tools (ok, I've not seen that in practice but it seems to be encouraged by many "best practices") etc etc
Oh, I agree completely with you, sorry if that wasn't clear. The PM should, must, care about UX. Still, they don't always do, or at least end up not caring eventually, for various reasons.
I'm just responding to this:
> what were your Product Managers doing in the first place if tech writer is finding out about usability problems
They might very well be doing their job of caring about UX, by using the available expertise to find problems.
It's a bit like saying (forgive the imperfect analogy): what are the developers doing talking about corner cases in the business logic, isn't the PM doing their job?
Yes, they are. They are using the combined expertise in the team.
Let's allow the PMs to rely on the knowledge and insights of other people, shall we? Their job already isn't easy, even (or especially) if they care.
Yes, product managers and product owners should also be looking for usability problems. That said, the docs people are often going through procedures step by step, double-checking things, and they will often hit something that the others missed.
I take your point, but a good PM will have been inside the decision-making process and carry embedded assumptions about how things should work, so they'll miss things. An outside eye - whether it's QA, user-testing, (as here) the technical writer, or even asking someone from a different team to take an informal look - is an essential part of designing anything to be used by humans.
I will share my experience, hopefully it answers some questions to tech writers.
I was terrible writer, but we had to write good docs and make it easy for our customers to integrate with our products. So, I prepared the context to our tech writers and they have created nice documentation pages.
The cycle was (reasonably takes 1 week, depending on tech writer workload):
1. prepare context
2. create ticket to tech writers, wait until they respond
3. discuss messaging over the call
4. couple days later I get first draft
5. iterate on draft, then finally publish it
Today its different:
1. I prepare all the context and style guide, then feed them into LLM.
1.1. context is extracted directly from code by coding agents
2. I proofread it and 97% of cases accept it, because it follows the style guide and mostly transforms my context correctly into customer consumable content
3. Done. less than 20 minutes
Tech writers were doing amazing job of course, but I can get 90-95% quality in 1% of the time spend for that work.
Your docs are probably read many more times than they are written. It might be cheaper and quicker to produce them at 90% quality, but surely the important metric is how much time it saves or costs your readers?
How I perceived it, Epictetus wants to say: things happen and you are on a spectrum of emotions based on the context (in case of death, how close you were to the person), try to minimize the length of the spectrum.
I agree in part. You could read Epictetus as saying "just try not caring about people," which I think is the incorrect reading. Instead, I think he's saying something like "take a step back and realize that your deep personal attachments don't look so important when you step outside your perspective. You can use this realization to help get past the deep emotional pain that is normal for people to feel."
However, the line about other's indifference I think can only be read as dark funny to a modern reader:
> has someone else's child or wife died? There's no one who wouldn't say 'So it goes.'
The world being indifferent to your pain is not helping if you're in acute pain. Step outside your perspective, sure. I guarantee you this will not work if you have real issues like physical pain due to terminal cancer.
You should read "A Man's Search for Meaning" sometime. While not exactly Stoicism many of the ideas are similar/related. How a person views and responds to their situation, has a huge impact on them. No one is saying any of this will remove all of someone's acute pain, but as crazy as it sounds, accepting that suffering can lead maybe the pain not being quite so bad.
Although I like the model, I don't like the leadership of that company and how close it is, how divisive they're in terms of politics.
reply