I feel that for some time now, the biggest constraint when working with models is not their intelligence, but their speed. It does not matter how smart the model is, it will make mistakes, because the instructions are ambiguous and new facts are found during implementation. The biggest problem I've had working with software developers has always been the lag between seeing the results and steering towards the right direction, not the skills of the developer (with many exceptions of course).
Regardless, working on the wrong things is time wasted. And again, I'm procrastinating here while waiting for Fable to run a benchmark on a few solutions to a problem I have. We can guess what would work, but we only know after the benchmark. A faster model, with fewer capabilities, would've been a much better choice this time... well, "git gud" they said... and live and learn! Faster model = less time for procrastination.
PS. AI models don't live and learn; the discussion about AGI is pretty pointless imo. It's a tool. Does it matter if it is AGI or not if it does what you want it to do? Does the IQ of your colleague matter if he's good at what he's supposed to do? Or bad? Well... I guess it does matter, as many people are up in arms about whether Astro is AGI or not. Personally, I think we're past the point for that debate. These are amazing tools.
I think speed is going to be important for a second reason - ever since I've started using LLM's a lot in my work I enjoy it much less. The main reason is that I ask the LLM something and have to wait because I know it's going to prompt me at random every few minutes. A lot of the day has become staring aimlessly at the screen. The prompts are too random and frequent for me to do something else in the interim. It's pretty soul destroying and this can't be how 'work' looks going forward or people are going to burn out.
This is why I switched to codex --yolo running in a container/vm. Now it does what it needs to do without bugging me and I can do other stuff. When it pings me I know there is something to review.
Yes I know they can escape containers, but that's not what I'm tasking them with.
You can easily setup codex rules to auto approve local stuff but gate external effects like push/jira write/curl - works better for me than full yolo mode - depends if it's a solo project or working with a team.
> It does not matter how smart the model is, it will make mistakes
It does matter, otherwise why are we all using GPT 5.6 rather than GPT 3.5? Because it's way smarter, makes less mistakes and therefore finishes tasks faster.
The smarter the model is, the faster it can complete what you actually wanted.
> Regardless, working on the wrong things is time wasted
Agreed. And "smart" for me, would mean understanding what is the right thing to work on vs the wrong thing, so a smart model would waste less time, thinking like this.
AI models do not live and learn - it's worse. They actually get DUMMER if you don't start with a clean slate. This is important. One has to curate the context carefully.
I use codex daily and a moderately complex plan + a few files is more than enough context to make it blatantly ignore prompts like "stop and ask before adding new layers or data structures"
I'm sorry but no - output quality matters a lot more for me.
Just yesterday I tried to use Google antigravity to do a side project I've had on the back burner for 10 years now. Gemini flash is insanely fast - at first I was amazed at how quickly I was getting responses, and it seemed to hold it's own in technical discussion, although sycophancy is next level. But then when I actually let it do the coding part it was just drivel. I wouldn't even bother improving that code - like cleaning up after a lazy unskilled coworker - throw everything away and start over because the foundation is just leading in bad direction.
I spun up Astra on the same problem and although it was sluggish in comparison, and much more pedantic about irrelevant details - the feedback/pushback was actually meaningful. The implementation PoC also took tweaking but we got on the same page really fast.
Gemini Flash 3.8 was just producing garbage ultra fast, Astra could actually be steered into a direction I want and it provides valuable/insightful feedback.
I don't have infinite reading capacity/mental stamina - I would rather the model take it's time and let me see something high quality rather than get bombarded with garbage. If it can be faster that's great - but I'll always default to smarter model. The only exception is stupid trivial tasks like log analysis and similar.
Good argument going here but here is what internet tells me about 'facts':
Its about the same amount of population who have access to clean water and who have access to internet, and that's maybe about 75% of global population for both. Don't know how they overlap, but probably there is people who have no access to internet but have access to clean water and the other way around.
In absolute terms that is still billions of people. Billions.
with a good enough LLM you can greatly increase the capability to build a functioning well (and other basic infrastructures) in remote and undeveloped areas without great technical knowledge, they are very good as assistants for practical tasks where you can give feedback step by step on what is happening
There are many analytics solutions that dont require cookies. You can do aggregated analytics just fine without. Saving preferences does not require consent either.
My understanding is that any front-end analytics solution will require consent. You're right about explicitly set preferences. I was mixing that up with inferred preferences.
How many requests per second are being served? How many error codes were delivered to clients? How quickly the service responded? Service logs without PII? All perfectly fine to aggregate and analyze without consent.
How long did it take x user to navigate from x screen to y screen is one of the most valuable metrics for any site, and most people consider this to require consent. Or at least it not being worth the risk to not ask.
Acting dense like this isn't productive... And literally this information would be stores as anonymous user 12345, but that still would require consent (probably, or at least arguably).
That can be implemented. Within a session you don't need to know it's the same person tomorrow, so a per-day key derived server-side is enough to measure that someone took 40 seconds from x to y. No cookie, no localStorage, nothing stored on the device, nothing to consent to. Hash ip + user agent + your domain with a secret salt that rotates and is destroyed every 24 hours, and you are on the safe side.
Of course, recognizing users across days requires consent. But is that really necessary?
Or use some JS to put the time-on-page in the next request, right?
Is it a violation to send data that could theoretically be used for more invasive tracking than you actually do? I don't think so, or else you'd need consent just to receive an IP packet.
>How long did it take x user to navigate from x screen to y screen is one of the most valuable metrics for any site
Stats like that are only used to implement dark patterns better and justify user hostile decisions since pretty much the time the idea of telemetry was introduced. Otherwise, we'd live in the world of perfect web ui and we're not.
A lot of people criticizing because it's heavily written with LLM, but I mean, if someone produced this piece pre-LLM, would they criticize it? is the critique due to use of LLM or due to the content being truly hard to follow? I read it and I would say, there are some problems with the writing, but its not a bad piece.
Of course this is a bigger problem, as its now harder to distinguish content that is "AI slop" with "content co-authored with AI that is carefully reviewed" with a quick glimpse, and the "AI smell" is quite off-putting. My initial reaction was also negative, but after glimpsing it through and reading the summaries, I found it decent summary, which also... speaks of this thread, of the content of the blog post and everything about the discussion and the strong feelings people have developed around the use of LLMs.
Anyhow, it would be good to disclose the repo with the code for the statistics & use of LLM in the writing right up front. Which model, and why it was used to do the writing, etc. Its enough to say "I think it writes better than I do" or "I was in a hurry, sorry" or what ever, but it really should be disclosed. It reads more honest.
> content co-authored with AI that is carefully reviewed
The problem I see is that this is indistinguishable to a reader at a glance.
Distancing the writing from the "AI smell" not only improves the quality by dropping the unnecessary ocean of rhetorical devices, it forces the human to have real weight and agency on what's being said.
I think that act of distancing from raw LLM output through refinement is a huge quality leap. Even if you're only doing the refinement with an LLM, it forces the writing to have more voice and ideas from the author.
I can see the work that went into the analysis here but again, as a casual reader, it's impossible to tell that there were any original ideas here expressed by the author.
Thank you for your constructive input, you're one of only a few others here who had any. I'll definitely do that. I didn't think, since the output was templated directly from the numbers generated by a reproducible python script, that people would get so up in arms about the aesthetics, but I guess I forgot to say that.
The most quoted line here is "A simple distributional analysis of every rsync release with bug data. No model. No assumptions. Just placement." Not only is it cringe to read, it's also nonsensical ("placement" means what?)
If OP had said "here's an AI summary of the data" and generated a conscise summary, I think I would fine with it. But default AI writing is really verbose -- the opposite of a compression algorithm, spewing out cliched phrases that don't add information. It's exhausting to read, and it lacks the interesting noise of a human response.
Maximizing profit for the corporation is the goal of any corporation by law, isn't it? Apparently not in the US, but for example the Finnish law explicitly states that the goal of a corporation is to generate profits for the shareholders. If you for example give away company assets for free, it can be considered breaking the law.
This probably is just culturally different understanding of the phrase, because US corporations indeed feel to act greedy, and there is no similar level of protection of the employees.
However, the thing is, in the long term, the business has to make profits to be sustainable. If the company does not make profits, it will die. Its the short term thinking that breaks down companies. You can maximize profits and be ethical at the same time, if the goal is to do it in the long term.
I do understand that the "maximizing profit for the corporation" is a synonym often for short term thinking and vulture capitalism, but for me it meant something else. This is actually quite fascinating now that I think of it, because this phrase means completely different things in different cultural contexts.
So I guess the trigger is that "maximize short term profits over long term sustainability" is the kind of company where I'd never work for.
I don't know about Finnish law, but in most US states, when you form an entity, you can put most anything as a purpose. It is quite common to put it as "to engage in any lawful activity".
Beyond that, publicly traded company directors may have (depending on the state) various duties to shareholders to consider profit, but it is generally not considered to be so extreme as to profit above all else.
IRS regulations distinguish between a trade or business and a hobby by looking at the intent to make a profit (as well as the effort). see https://uslawexplained.com/trade_or_business
But it is generally possible to form a legal business under state laws with no intent to profit, the IRS may just treat it as a hobby for tax purposes. (You can't deduct expenses incurred pursuing a hobby, but you can deduct expenses incurred carrying on a trade or business).
Oh I'm aware of Finnish law on the matter (moikka), and that is my point. I think the LLC (OY), which is the political institution that creates this profit maximizing incentive, is _the_ main driver of the current multi-crisis situation that he world finds itself in.
LLCs and the profit motive will not save us from climate change, they will drive us deeper into it. Sustainable human living and continuous economic growth are not compatible.
I have used past week opencode go with deepseek v4 pro and claude code with opus 4.7 side by side and... they are both good. They are different, both have their good and bad sides... but they do get things done. Especially the OpenCode has been very enjoyable experience. Thank you Anthropic for all the down time, I would have probably not explored alternatives otherwise. I can vouch for the OpenCode Go sub!
I have built a product which uses AI to built Astro sites. LLM builds the sites in steps and make sure that they get 100/100 scores in pagespeed insights. These sites are served with a CDN. You can edit the sites with LLM interface, or use markdown editor to edit sites, or edit texts directly on a dashboard. These sites are static. There is no vendor lock in. If you want to migrate and manage yourself, just go to cloudflare or github pages. These sites cost 0 eur to run, and they always score better in all benchmarks compared to sites that are built on top of a separate CMS server.
I know WordPress is going nowehere and if there is some special backend functionality, that is needed. But 95% of web does not need it.
A static site is always cheaper, and the bottleneck has always been that editing code is indimidating. Therefore, AI actually resolves a big problem here, and this is going to alter the future of platforms like WP.
I’m confused. Static sites are the opposite of the target market for CMS. Even assuming consistent behavior from an LLM that never makes mistakes, how does your system handle access control, or multiple users submitting, reviewing, and publishing content to the site simultaneously?
What do you use to edit texts on a dashboard or cms?
My wife needed a website. I’m not much of a frontend coder but I looked at the price of squarespace for a year and decided to go that route with AI.
Website is Astro. I easily update sophisticated designs with Claude or Gemini cli.
Hosted for free on cloudflare, it’s super fast. Any git update deploys to the website in a minute.
Got an hosted email form with astro action.
The only thing left is integrating a CMS. I was thinking of keystatic but it’s not compatible with Astro 6 yet. That’s the issue with vibe coding a stack you don’t know as much, without realizing it picked a version of astro that was so new that some tools didnt quite integrate with it yet.
For some personal projects in the past I have used google sheets and cloudflare workers, I guess it could be called CMS, but it was quite restricted of course.
This project has a "shared" backend for the content management and its completely self built. Because there are many clients who use it to edit things its feasible to pay for the hosting.
For a solo project, id be quite interested if someone would build a native editor that can be just hooked with git, so it would make it possible to run the site serverless. Maybe one already exists. "IDE" for static websites.
I built the same thing and then just realized that I built a marketing funnel for Cloudflare lol. It's why Cloudflare is trying a bunch of different approaches to the same thing, they're the only ones that actually benefit from it because you can't actually build a business off hosting millions of sites on cf pages, it's a loss leader for them to convert you to a paid product if you end up one day getting a lot of traffic
Hosting a static site isn't free, they just don't charge you for it early on
NASA, White House, and which ever large organizations do not represent the most of the web.
When you have complexity, multiple non techincal users who need to update content, and frequent changes, a CMS is currently a very good solution. But thats just a small fraction of websites.
Most of websites are small, 1-2 person companies websites, non-profts, etc., that are basically business cards. Contact details, possibly a contact form, and few pictures. Thats it. There are likely at least hundred milloin websites like that, which are infrequently updated.
Majority of those sites are powered by WP and various site builders, which is far more complicated than what they need. There has not been good option for non-techincal users that makes it possible to make good looking and functional sites.
Also, please keep it civil. This is not Facebook. People can have different opinions.
Yes, I completely agree. The thing is, this kind of customer just doesn't want to bother themselves with the technical details, and has no frame of reference to understand or even care why Wordpress isn't actually a good fit for hosting their site.
They also usually don't want to self-serve. IMO this became abundantly clear once I saw who was using bolt.new and Lovable and what was being built. You'd think these would be perfect fits for non-technical business owners, but after talking to them more it turns out they just don't have the time or interest to spend hours on building some little marketing site, and want it to be someone else's responsibility. Conversely, I would never build something with Framer and have no interesting in allowing some fly-by-night agency hold my site hostage, but they do a lot better at actually delivering value to end users without making them spend their time on tech stuff they don't care about.
Conversely, the kind of person spending hours building a site on Lovable for some SaaS product nobody will ever use has an abundance of time and doesn't really want to pay for anything. Most of the time they won't even put their own name on the site lol. You just don't want to deal with that kind of person IMO. Cloudflare and Github allow it because there's a small chance that a small portion of that kind of person ends up actually making something valuable, and because they have a different cost structure due to their affiliations with massive infrastructure holders.
I got very, very close to launching a vertical static site hosting product a few months ago but eventually realized this was kind of a market for lemons. Our own site is on a Lovable-like platform we built that uses our own svelte-baesd FOSS static site generator called Statue. But in using it to try to make some visualization on our own site, and vibe-debug stuff like a non-technical customer would (this thing on this page is broken in this way) I realized that this wouldn't actually feel like magic to someone who values their time, or isn't getting paid a salary to be a web developer and doesn't understand/care that it's still quite labor-intensive to do this.
IMO the real money is in actually being willing to take accountability/responsibility for building someone's site, and building real tooling around it that works for non-developers AND developers, which is what we're building towards now. It's historically been treated as a kind of low-prestige/uninteresting/unscalable business doing agency web stuff, but if you can figure out how to make it scalable and give people beautiful websites, and not make people who value their time wade through slop, there's immense opportunity.
What is always astounding to me is that people talk as if caching isn't a thing. It could hardly be easier to cache the html output from wordpress at either the webserver or CDN level, and it will perform just the same as any "static site" (of course, images, css etc will dictate how it performs once the browser receives the cached html)
With a static site you know the output can be cached indefinitely and only invalidated by a new deploy.
With a server rendered site you can only recreate that by tracking ever piece of data a page is dependent on, tracking data changes, and invalidating any page a data change breaks.
No one does that though, so you may go for SWR or a short-ish cache window so changes take some set of minutes to roll out, looking like the delay in waiting for a static site to rebuild.
I'd never pick between static and server rendered based on caching. Factors like the size of the site, frequency of content updates, and technical skill of content authors (I've never found a git-based CMS I'd ask someone totally nontechnical to use).
You could do this with AI for at least the past decade. We saw lots of companies & frameworks spring up that targeted and did the ahrd work pushing this approach. That feels like the big change, with "using AI" to be an incremental gain here.
>Literally talking without knowledge here. There is always something that can be added with a Wordpress plugin and there is somebody who needs that.
So? There's always somebody who needs this or that outlier shit. If all that shit combined is still a small niche, we can just ignore it. And it is.
>There is a reason why NASA, White House, Techcrunch, Reuters et al are all on Wordpress and any of the 'better' cmses out there.
And there reason is not because it has some obscure plugins for features few care about, but about the maturity of the core offering. They're not having any exotic features or have some random niche plugin. And even if they did, they're larger than 99% of websites, so we can ignore their special needs when talking about what MOST need.
I agree. Now we have the ability to have agents reason over our notes, it's more important for them to be in plain text. It was a big part of the reason I developed the AS Notes extension for managing documentation and blogs in VS Code / Markdown (https://www.asnotes.io)
Ditto. Almost all of my internal tools are built in the same fashion and deployed on cloudflare itself.
It is so much liberating to have no overload of CMS and the bloated code they come with.
The limits in the max subscriptions are more generous and power users are generating loss.
I'm rather certain, though cannot prove it, that buying the same tokens would cost at least 10x more if bought from API. Anecdotally, my cursor team usage was getting to around 700$ / month. After switching to claude code max, I have so far only once hit the 3h limit window on the 100$ sub.
What Im thinking is that Anthropic is making loss with users who use it a lot, but there are a lot of users who pay for max, but don't actually use it.
With the recent improvements and increase of popularity in projects like OpenClaw, the number of users that are generating loss has probably massively increased.
I've spent $17.64 on on-demand usage in cursor with an estimated API cost of $350, mostly using Claude Opus 4.5. Some of this is skewed since subagents use a cheaper model, but even with subagents, the costs are 10x off the public API costs. Either the enterprise on-demand usage gets subsidized, API costs are 10x higher, or cursor is only billing their 10% surplus to cover their costs of indexing and such.
edit: My $40/month subscription used $662 worth of API credits.
oh, I figured out the costs for the enterprise plan. It's $0.04 per request, I'm not charged per token at all. The billing is completely different for enterprise users than regular users.
This exactly. I think this is why Anthropic simply don’t want 3rd party businesses to max out the subscription plans by sharing them across their own clients.
A lot of those relevant writings became relevant because of the horrible experiences the author went through forged them into an interesting writer. If we're assuming that we only know retrospectively whether the writing is important then the best course of action would be for people to write as a hobby and make choices that are likely (rather than unlikely) to lead to a comfortable life. Particularly in this current era where we might suspect that writing and publishing a book is getting much easier thanks to technology.
> A lot of those relevant writings became relevant because of the horrible experiences the author went through forged them into an interesting writer.
Sometimes artists suffer, but it's mostly a legend at this point. Plenty of great artists have perfectly fine lives. Look at like, any modern fantasy or sci fi author.
Are you arguing that most good writers from history were poor? This is after all the only "horrible experience" a subsidy would alleviate. I don't think that's actually supported by evidence, most great writers I can think of were relatively extremely sheltered (although they often were sensitive to the horrible experiences of others)
I think the argument is a) most writers have to do a lot of writing to achieve writing consumable/appreciated but sufficient to be considered successful, b) most great writers had to go through some shit in life to incorporate that in their writing to make it interesting in order to be successful.
Regardless, working on the wrong things is time wasted. And again, I'm procrastinating here while waiting for Fable to run a benchmark on a few solutions to a problem I have. We can guess what would work, but we only know after the benchmark. A faster model, with fewer capabilities, would've been a much better choice this time... well, "git gud" they said... and live and learn! Faster model = less time for procrastination.
PS. AI models don't live and learn; the discussion about AGI is pretty pointless imo. It's a tool. Does it matter if it is AGI or not if it does what you want it to do? Does the IQ of your colleague matter if he's good at what he's supposed to do? Or bad? Well... I guess it does matter, as many people are up in arms about whether Astro is AGI or not. Personally, I think we're past the point for that debate. These are amazing tools.
reply