MuffinFlavored

> For less than the cost of this plan, you can get ChatGPT Pro 200, Claude Max 5x, and Kimi Code.

Be a better friend to yourself. Value your time more.

    > Qwen 3.8 27b is the supervisor
    > 
    > Qwen 3.5 4b are the 6-15 minions it controls
    > 
    > Gemma 4 e4b is the validator for the supervisor.

I just use Opus 5.5 and don't think about it?

If the CursorBench 4.0 score diagram is the headline, I read it as "Grok 4.7 xHigh is almost the same as Fable5.1 on low".

Is there a metric for like... time taken when comparing these two? I see score and cost.

If Fable5.1 can knock it out more quickly on low but Grok4.7 might take twice as long to stumble through a problem (and leave behind a bunch of yucky comments or un-needed extra unit tests), are they really comparable?

Or like... the "quality" of the solution? "It works" versus "it's unmaintainable/very messy/hacky".

What is it "inflation adjusted" or is that cyclical because inflation itself measures/is a measure of the cost of diesel (which bleeds into the cost of goods and services)?

Have you ever worked corporate before?

It is better to ask forgiveness than permission.

Downside: Finding a seat at an already crowded gate just got harder.

Upside: loved ones get to hang out with loved ones for longer.

> Produce working product first, validate the idea, stabilize the business, start generating profit,

not everybody is so lucky to be able to go in that order? The first part requires upfront capital/investment?

> you get lots of usage on any plan.

I hit my weekly limit on $200/mo Codex plan in about ~2 days. :/ I'm not doing anything custom/crazy/special. A lot of 5.6 Sol Ultra though, I'll give you that.

I submitted an application for Anthropic's Cyber Verification Program.

I was approved.

3 months later, my approval was degraded into "in review" (revoked). I'm sure my account was flagged based on contents of debugging/researching firmwares/etc.

I opened a support ticket. No response. I opened another support ticket. No response.

1-2 weeks later, I got a response that I will not be re-approved and I need to reapply. No problem.

The page to reapply on does not allow me to re-apply because it my account is stuck in an "in review" status.

https://github.com/anthropics/claude-code/issues/84352

The community thinks it's a bug. I'm 95% sure it's not and a bunch of us who were previously approved had it revoked due to flagged content and will not be reapproved.

I switched to Codex + got TAC approved instantly and have not looked back. It's a shame. That's 100% separate from whatever the heck the quality of Opus 5's outputs are. The way it talks... insane. I would bet a good amount of money their next release will focus "reduced simplified responses" if I had to guess.

$2t company by the way

With the (unknown?) blend of consumer vs enterprise customers, what do "experts" project their margins to be around?

Some notes from me researching trying to answer my own question:

> Wall Street experts and financial research firms project Anthropic’s current blended gross margins to be in the mid-40% to mid-60% range, with internal company forecasts aiming for a software-like 77% gross margin by 2028

> Anthropic’s revenue is heavily dominated by enterprise and developer customers (roughly 75% to 85% of total revenue).

> Premium Token Pricing: Enterprise and API clients generate 3 to 5 times more revenue per token than consumer users.

Net is estimated to be between 10% and 30%

> Yes, Claude Code uses Bun. In fact, Claude Code relies on it as a core dependency and ships as a self-contained Bun executable.

I... somehow did not know that.

Piggybacking onto a second/different "just a suggestion from a user":

The VS Code extension needs love. I'm sure you guys are aware but it feels like it is neglected.

GitHub Issues is a graveyard of 3-10 duplicates of really important issues with no activity getting closed. A few examples:

* Lots of /commands missing in between the CLI harness and the VS Code extension

* No way to monitor subagents/tasks/progress visually

* No status bar/line

I think I've read from a few different sources that the Claude Code $100-$200/mo plans are subsidized so hard that it's basically $2k-$8k/mo in "would-be" equivalent API token usages.

This kind of makes sense in that space while the subsidies (if true) last?

Unrelated, "tokens" feels very like... back-then blockchain to me. All the craze.

The company I work for, the code Opus 4.8 is able to generate, is higher quality than what was left behind by 10+ years of contractors that have come and gone.

> Just stop caring about quality.

I'm not so sure this is true anymore. It may have been years ago but... can you honestly say "the Bun project was fully AI written, therefore the quality is poor"?

Any concrete examples/proof?

I know a ton of people absolutely hate this level of "LLM code + LLM PR description + LLM PR review" but my boss would have an orgasm if I was able to use AI half as well in our org... :/

I'd like the see the author achieve the same setup but with Nix

I must have tried to write the same "perfect" IRC client from scratch in C a dozen times growing up...

Did open source phones win? No, iPhone is pretty dominant.

Did open source operating systems win? No, MacOS/Windows are pretty dominant.

Does open source... cloud hosting, social media, ride sharing apps, you name it win? Not in my experience?

> email service provider (pissmail)

I'm sorry this happened to you.

This is a foothold business living entirely at Apple's discretion.

edit: more research

> Chert is in the Sendblue/Blooio lineage, which runs genuine Apple software on real Macs logged into real Apple IDs. They're almost certainly not doing "Beeper Plus again."

I wonder how many "behind the curve/not super modern" corporations were using Bun or Deno to begin with.

Part of me thinks it's a mild overreaction. It's not like people audit every line of kernel/driver/BIOS/EFI code before running Linux? As long as the tests pass and the performance doesn't regress and it's secure... why are people so mad that it was vibe coded? Is it because it was an irresponsible thing to do? Maybe?

I don't know, I see both sides.

> Well and uber cut the driver pay in half and doubled the price

Devil's advocate:

* inflation caused everything to go up to some degree since then

* if it was "that bad" as you say, they wouldn't be extremely profitable and have so many users

both things can be true? "they cut the driver pay in half and doubled the price" did not lead to the collapse of the business/people to stop using it.

> Frontier AI companies are selling at a loss.

How big/deep of a loss?

I feel like I read this every day for years that Uber did this same "idiotic, losing" strategy (how it was pitched/discussed) and then one day we woke up and... without much fuss, boom, they were profitable seemingly overnight.

Might want to add how this compares to other products in the space.

Some that come to mind that are potentially tangentially related/similar:

https://github.com/evidence-dev/evidence

Have you tried NixOS/flakes? What was your reaction?

I wish they would just rip the bandaid to stop everybody's entitled whining.

"We're sorry, what we were able to give you for $100/mo before now needs to be $200/mo (or more). We miscalculated/we were too generous/gave too much away for too little. It's a new technology, we are seeing a ton of demand, we are trying to run a business, hope you understand. If you don't want it, don't pay for it."

I wish people would pay more attention to:

* Anthropic is in some way trying to run a business (not a charity) and at least (eventually?) make money and not subsidize usage forever

* "What a steal/good deal" the $100-$200/mo plans are compared to if they had to pay for raw API usage

and less on "how dare you reserve the right to tweak the generous usage patterns you open-ended-ly gave us, we are owed something!"

How big of a handicap on performance is the external enclosure for something like an RTX5090?

> Running DeepSeek V3 (685B) requires 8×H100 GPUs which is about $14k/month. Most developers only need 15-25 tok/s.

> deepseek-v3.2-685b, $40/mo/slot for ~20 tok/s, 465 slots total

> 465 users × 20 tok/s = 9,300 tok/s needed

> The node peaks at ~3,000 tok/s total. So at full capacity they can really only serve:

> 3,000 ÷ 20 = 150 concurrent users at 20 tok/s

> That's only 32% of the cohort being active simultaneously.