brianjking

In fairness, nor does Salesforce.

I've been using ChatGPT since release and Codex since before ChatGPT was a thing (with the original Codex model). Am I the only one who didn't realize these limits weren't already merged since the current Codex app was released?

This is sadly not surprising to me as if you listened to/watched the [Latent Space episode on Opus 5.5](https://latent.space/p/how_anthropic_ships_the_1_llm_in_the_...) 25 minutes into the podcast and instead of talking about how they've released one of the best models in the world, they were busy talking about the changes to the API and why normies must not be trusted to perform inference outside of Dario's blessing.

It's frankly disgusting and really makes me not want to use Anthropic, despite previously feeling aligned with a lot of what Anthropic published.

4-6 months? Try between 8-18 days for a foundation lab release.

This is happening at every other AI lab. What are you talking about? Have you not seen the stories from Meta, Anthropic, Deepseek, etc?

Furthermore, it almost surely limits the quality of the output. At least the finterprinting of the text tokens has to, how can you otherwise constrain token output?

This! So much this. After Opus 4.8 I could barely comprehend anything it was attempting to communicate.

.... Minus that Nvidia keeps contributing to the open ecosystem with nemotron, among countless other open contributions.

It's lies all the way down...or up?

What about your life's OKRs?

I've found it so much easier to work without a 5 hour usage limit. I really hope we rapidly get to a world without as much need of one.

Tagging files in Claude desktop in code mode works fine for me, as does CLI. With that said, in Cowork, it's been broken for some time and driving me crazy.

I've shifted most of my usage to Codex/ChatGPT Work. The UX of appshots, browser annotations (now available in claude), and the computer use being so much less intrusive in OpenAI.

Yeah, there's a ton of great improvements in 7-28. I'm personally excited about what you posted about, but also with [tasks](https://blog.modelcontextprotocol.io/posts/2026-07-28-releas...) being officially adopted.

Yes.

Definitely agree on the title. Also could have worked as 'Some minor annoyances that, when fixed, would improve OpenCode. Also, I don't like AI"

I use Aikido safe chain, how does this compare? https://www.npmjs.com/package/@aikidosec/safe-chain

I can understand and appreciate the hesitation. I must admit I've simply adjusted to the idea that anywhere that I am, outside of my house, I assume someone is potentially recording or transcribing at a minimum audio.

I'm frankly far more upset by the lack of privacy due to poor security from credit reporting agencies than I am of these notetakers.

Z.ai is based in China and serves out of Singapore, that's surely why.

We need more geese.

Perhaps very good, perhaps very bad, perhaps not possible. Ultimately, I think it still comes down to whatever PEOPLE decide.

I don't mean one person, I mean people, as a species.

There is a newer BLIP-2, but it's also fairly old. You're better off with many other local models such as Moondream 3 https://huggingface.co/moondream/moondream3-preview.

Moondream is great as it can point, count, perform bounding boxes, descriptions, and visual grounded reasoning.

Great work, thank you for doing this. Just so I understand, this isn't yet available yet, right? Still in SEP stage?

Always happy to see a Hitchhikers guide comment.

This website is awesome.

...Not to mention that they're investors in Anthropic.

I use https://github.com/EveryInc/compound-engineering-plugin/blob... from https://every.to.

Here's a readup on it: https://every.to/guides/compound-engineering

I've noticed some behavior like this, it's a very strange model. Overall I'm into it, but I don't know how into it I'll be once it leaves Max plans on the 22nd.

yes, yes, yes.

I tend to agree, but there's also a lot of tax to build and maintain the different provider abstractions that OpenRouter eliminates.

Everything has a cost of some sort. It's just who you're going to pay and what the currency is.

How are you interacting with GLM 5.1? Via the Claude Code harness? I really wish they'd release a fully multimodal model already.