cloverich

I take it you have - i guess can you provide some examples and how they would materially differ from what Anthropic would do then? It feels very similar on the surface.

They don't have anything competitive with Opus 5.5 yet. There is a supposed model in the works, "Bel" last time I paid attention to gossip, that is thought to be better, later this year. Allegedly Anthropic caught OpenAI off guard and / or their different training approaches ended up working better than anticipated.

My expectation is Fable 5.5 is next overall improvement, and my bet is OpenAI's "Bel" will be the next competitive one (or, their lack of safety culture will bite them/us in the process).

Wouldn't a better analogy be Google, which would actually contradict the point you are trying to make?

To clarify, i use both subscription pricing (personal use) and api pricing (at work). My spend going down was referring to api pricing, while using it more (for "everything").

Personal efficiency is noted as never hitting my limit despite it doing huge amounts of work at a time (ex: rewrite my notes app in swift for ios and install on my phone, 2 hours later its opened on my phone, looks nice, is working (base set of features). No follow up prompts, around 2 hours of work. 45% of my 5-hour limit.

Work efficiency is spend, task complexity, code quality (how often i have to steer - rare now), and wiring it up to more stuff (ex having it setup metrics and dashboards etc etc).

> increasingly commoditized

Is it increasingly though? As of 5.5 I stopped using other models all together (and I used to regularly swap, indifferently). My $20 sub no longer runs out, even though I do _much_ more than ever. I use it for every single task at work, even trivial ones, and my spend has gone down. I was given _free_ Codex and realized... it still wasn't worth it. My rational brain says Anthropic has no moat and this is temporary. The progress of all the other models is amazing. I certainly wouldn't bet on them.

But if I look at the actions of myself and the most competent coders around me, they are all doing this same thing - using 5.5 and not really looking around any longer. And we know its not even their best model. OpenAI may respond, Chinese labs may catch up. But for the moment, Anthropic seem to have added separation, not lost it, based on current vs prior _behavior_ of coders I know at home and work.

Also 2 years back the teachers strike in November for funds that... weren't there... that was pretty damaging to my argument to get my kids out of private and into public (we ended up moving to WA in the end).

Oregon is also extremely rural - Portland is the most populous city in the state by far, and its e.g. far smaller than all cities in the Houston/Austin/Dallas/SanAntonio metroplex. There's important cultural factors that impact its education too, like historic underfunding, poor education spending priorities, and out of touch unions (many days off, striking mid year for funds that aren't there). But I always felt the rural-ness impacting wider education trends was very under represented in discourse.

Build tools to help you understand. Build metrics that are proxies for the things you'd care about. Add constraints to the agent's work related to performance, error monitoring, language and architecture choices, review bots (that you tune and eventually, trust). If you don't have good ideas for tools, ask Fable what the best coders and industry experts would look at to establish trust, then feed that into models to build it. Don't sleep on how useful something like a (data-backed) claude artifact can be - i setup one today to track dependency and known vulnerabilties with a burndown chart and (agent) run log, one prompt, it was usable 5 minutes later. A colleague of mine did the same, but as a datadog notebook that also wires in relevant API metrics depending on what code has gone out in the last week, and suggests new longer term metrics to monitor as a result. Etc etc. its a crazy world but this has helped me cope so far.

> Doesn’t matter how smart the model is

Is this true? I'd heard loosely that the higher intelligence models are less reliant on the harness; I'm still wading into the details of those claims so maybe someone more well versed can chime in, but my expectation was the current phase of harness engineering was likely only needed temporarily. I suppose turning Opus 5.5 loose in a variety of harnesses could be a decent if anecdotal test of that?

A mix of interest in the various ways images can be represented in alternative formats and then applied artistically. I can't speak to why, but every time I see one of these techniques I'm fascinated by it, and want to play with it[1]. There's a regular stream of related posts that come up on dithering, half tones, etc, and tools people build off of them. I find everything from the history/why, the details of the techniques, and the artistic uses of them (in modern UI/prints that no longer _need_ them) all fun and interesting.

[1]: https://cloverich.github.io/artifacts/horizon-plate/dist/hor...

> has some input

The president proposes the entire budget, and can veto the entire budget regardless of which party is in power. Of course Congress has immense influence and you also can't pass a budget they don't agree with. That doesn't change that not only does the president matter (the claim I'm refuting), they are the most important single person that matters, by design.

The president effectively does because of Veto. They inly cant control the most bipartisan aspects which is very limited today.

> Then we can spend however long it takes trying to understand how it works.

...unfortunately it took 57 years to unravel, but the cryptic exploit resulting in our telomeres suddenly expiring took a mere 42.

Until Opus 5.5 I agreed with this. But since then... I'm now able to build just about anything I want, quickly. Including... all the reasons I'm on a mac (or presumably, one on windows) in the first place. IDK if I believe in my heart just yet, but as a long time mac user, if you gave me the choice between mac + commodity model or a near naked Linux laptop + Opus 5.5, I would easily choose Opus 5.5 and build what I want. If this is real, I think Google is still in a good spot, but are Apple and Microsoft? They have no models themselves. They do have hardware but the value of their OS (in Apples case) and software (in MS case) becomes extremely devalued, relatively speaking.

They already are. I asked opus 5.5 to plan out and begin building a swift version of my electron notes app, using a similar architecture / feature set. I hut enter and honestly forgot about it. 2 hours later im reading slack while making my next round of coffee, and my session is interrupted by a native app opening on my phone (the model installed it for me to test). It.... has 2/3 of the features already and looks awesome, holy crap. It skipped my suggestion to use the Lexical (web) editor and is.... using. platform native one.

time will tell if it can make it the way i like but.... it can definitely get far. Used 40% of my session limits on my $20/mo sub.

They are in the tabs vs spaces camp. Choose your preference and walk away, silently. Unless you consider yourself a Real Programmer, in that case, fight to the death about the matter.

(...I also find them annoying to look at, idk why)

It makes sense but in a post AI world - what is of value? Not the software itself, but data and network effects. MS needs to move from a world where their proprietary software makes money, to one where they enable a captive network and have data; Github provides both.

So maybe them providing service at literally all costs to themselves is a smart strategy long term.

Need to take into account how belief and human brains work. Once human brains become locked on to something ideologically, similar to religion, its not the quality of the evidence or argument that changes minds. To an extent we're designed this way to keep the world moving, as the natural order of things is relatively indifferent to suffering. You can indeed flood segments of the world with disinformation, and by the time you move to refute it, its roots are set and actions in motion.

I don't really know what the best approach is to it, but I don't think a simple hands off, engage with the argument as such, is a winning strategy. I don't really like the alternatives any better.

Population growth is driven largely by immigration though.

Supply is rising from builders and also from lack of buying, iiuc the flattening and recent drops does line up with immigration and layoff changes, and its often quoted by the multiple local real estate agents weve talked to.

Obviously supply matters most but that doesnt mean demand isnt still a significant factor.

I'm very curious about the viability of this. Like many, I've built something along the same lines, twice (for personal, and work use); I staple on the features as I need them. On the one hand, it doesn't feel unique enough to attempt to polish and monetize. Then on the other, I expected thousands of these and yet when I look at my colleagues workflows, they aren't doing this. So I'm curious if it'll work out that its simply more efficient to use a paid service like this which you can drop into place, than rolling your own.

Two other quick thoughts. One is I think if people are struggling with costs, they really should explore a worfklow like OPs. Merely breaking up long running tasks by producing summaries / next steps and spinning up new agents off of them greatly reduces runtime costs. I'm regularly running out of time / mental energy these days than my $20 claude subscription, but I realize if I don't note share early and often, I can burn up the quota in 20 minutes. Its amazing the difference. Second is I feel encryption plays an important role. I feel like it is more important than merely keeping your own ideas private, and somehow impacts the generation and evolution of unique / diverse ideas _in general_.

Anyways, good luck OP, the use case is definitely there and important!

Brings back memories; so much nostalgia. I still remember clearly the first real graphics card (I think it was voodoo) I got and when I fired up TFC (which I'd played sans graphics card up to then), just staring at the water for a bit and being amazed by it. The funny thing too is when 3d graphics were first taking over, the games themselves didn't look as nice as the former 2d ones; I remember being a little confused at first that everything looked worse. But of course it enabled all kinds of new and more immersive experiences. It was funny to experience... how it felt. To look both crappy and amazing at the same time; to be its own "thing" really.

I also enjoyed introducing people to it, and watching them look at it the first time. You could tell who would get hooked. Some people would play the game a minute or two and be like wow yeah this is pretty neat, and then go about their life. Others would kind of just stop and almost watch themselves, and almost invariably just start laughing. It was so pleasing to watch it melt people's minds the way it had mine.

Anyways, a special time. Really fun seeing these projects.

The reason the code is the specification is because people don't take specification seriously. Usually for good reason, but sometimes not. Real, long-lived RFC's can exist, and can have directional and corrective impacts on your LLM generated code. The trick is the RFC needs to be human written and maintained. The moment the org allows the LLM themselves to modify the spec... then yes the spec is no longer useful; or rather, "why" comments in code + a sliver of high level directional / summary content is probably all that is valuable.

A fun example can be having the LLM implement a well known spec (which it cannot edit), and then to use the spec in review to find mistakes and corrections. That's helped it really click for me. All these tools that let LLMs generate spec, even with iterative planning... I haven't found it useful for very long. It's great for building something complex in the very short term (e.g. days). But i haven't found keeping them afterwards to provide any benefit.

Their whole algo is good. I had it work on me recently on facebook marketplace, the only social media outside hn im still using. And its shockingly quick at finding niche things im interested in (somy cameras in a certain range, 30L osprey backpacks, im expecting it to infer my shoe size soon). and its fun. And also revealing in how it can make you want things you dont really need, if youre paying attention. Clearly bad for you. But definitely a strong (sub) product on their otherwise dismal facebook app.

Notably, I read this on both my iphone and mac (LG Ultrafine 5k). It is was enjoyable to read in a fun way on my phone, it did not really bug me as a fun blog font. I was more excited about the theme/design than the content. But on my main screen, which is hi res, it _did_ bug my eyes quite a bit. Interesting.

There is such a thing as difficult concepts that code can express elegantly and words cannot. Its far beyond the simple preferences you allude to. Recursion is the simplest example i guess. Perhaps maths is even more elegant, but im among many a non mathmetician for whom code is the only way ive experienced it thus far.

Especially when you combine and mix those concepts in expressive, simple, intuitive ways. Im just not sure what else is like it that our minds can percieve.

Narration is an obvious one. Theres also a huge risk to less qualified folks and jr employees. ie in most software shops, youll get more mileage giving your best engineers bigger opus budgets and a raise than jr engineers to train. Which means we meed maybe half or less as many, and at least half of that savings is floated up to sr folks and ai companies. That seems rather disruptive?

because you are now getting coded products written in large by people who do not have technical foundations, so the way they interact with and even prompt the model is different. We all know how to fix this scenario; be more specific, or diagnose the abstraction mix ups and straighten those out.

Ask for change A and get unwanted change B happens all the time with bad programmers and tradgedy of the commons (ie poorly architected, no restraint) codebases.

Most normies dont know about this stuff.

Shutting down the isp doesnt let them read what you've sent. There wouldn't be intensive government efforts to eliminate private encryption if it wasn't effective.

notifications specifically played, and still play, a most prominent role.