scuderiaseb

I think you’re missing the point. Back then we weren’t overstimulated overall due to constant internet access and smartphones. So reading in the bathroom wasn’t as bad. Back then you used to get a lot of opportunities to have breaks from stimulation, now the ’taking a shit’ break might be a good place to just not have any external inputs and just ponder.

As they say, piracy is the only true ownership.

I have been the lead of a very stressful project that was the highest priority for the company I work for and it was crucial to get this working as it was a hard deadline. I started using AI a lot, but I still read the code and sometimes it was so garbage. I had two developers with me, but they were both kind of new in the domain and one didn’t know the programming language. The project finished and the part my coworkers developed was fundamentally flawed, it worked for ~85% of the cases but not all and caused an incident.

Now I get to work with two of my other friends/colleagues and they are amazing. They don’t use AI at all and so neither do I and it is glorious. It is so nice to build things like this and honestly not meaningfully slower either. It took us some time to figure out all the details, edge cases and all. Writing tests first, I love it!

I like the idea and even though I would like to have something like this, I still find need for a smartphone for certain things. In Sweden we are basically required to have an identification app (Bank ID) and that app alone can’t even be used unless you have NFC, which for instance an iPad doesn’t so that you could offload to that. Small things like this I run into constantly.

Fortunately I have discipline enough that my smartphone is a glorified dumb phone with a great camera and some other useful apps. Everything else I try to offload to iPad since I am much less often on that device. It also takes a bit of time and effort but focus modes are great too.

So this is how they’ll remove access from Claude Pro to the biggest models. You would need at least a Claude Max subscription for the bigger than Opus models I bet.

I use ghostty and tmux, that makes it really nice to be in th terminal and have access to good TUIs. I use neovim and k9s mostly and those are such amazinng tools, AI stuff sure but that’s because I don’t really have any other choice. I also use lazygit lately but that’s mostly to get familiar and to have a nicer view of commit history than ’git log —graph’. Also I hate electron apps so well made TUIs are a step up compared to running a web browser for one app.

> I do not know how you can do it on a Pro plan with Claude Opus 4.7 which is 7.5x more in terms of limit consumption and any small to medium size codebase would easily consume your limits in just the planning phase up to 50% in a single prompt

I also don’t understand because all I ever hear is people saying $100 Max plan is the minimum for serious work. I made 3-4 plans today, I’m familiar with the codebase and pointed the LLM in the direction where it needed to go. I described the functionality I wanted which wasn’t a huge rewrite, it touched like 4 files of which one was just a module of pydantic models. But one plan was 30% of usage and I had this over two sessions because I got a reset. I did read and understand everything line of code so that part takes me some time to do.

I must be doing something very different from everyone else, but I write what I want and how I want it and Opus 4.7 plans it for me, then I carefully review. Often times I need to validate and check things, sometimes I’ve revised the plan multiple times. Then implementation which I still use Opus for because I get a warning that my current model holds the cache so Sonnet shouldn’t implement. And honestly, I’m mostly within my Pro subscription, granted I also have ChatGPT Plus but I’ve mostly only used that as the chat/quick reference model. But yeah takes some time to read and understand everything, a lot of the time I make manual edits too.

Opus 4.7 was released a week ago, at that point all limits were reset, so this was very beneficial to them because basically everyones weekly limit Was anyway about to be reset.

I have found it very useful in pointing me in the right direction, exploring codebases and writing up boiler plate or other scaffolding. However I do need to review and test, and at the end of the day it’s me who’s responsible for the code. So I only make it write parts of code at a time, not oneshotting a new Github company.

The “retreat” framing is too US-centric. I’m in Sweden and European OEMs aren’t retreating — BMW Neue Klasse just launched, Renault 5 is a hit, Skoda Enyaq/Elroq are everywhere. These are competitive new platforms shipping right now.

And the Iran/Hormuz situation actually strengthens the case for EVs, not weakens it. Swedish electricity is almost entirely hydro, nuclear and wind. When oil spikes, petrol drivers feel it immediately. EV drivers barely notice. Pulling back from electrification right now is doing exactly the wrong thing at the wrong time.

Worth noting something the PR comments are glossing over is that the auth plugin wasn’t actually deleted in this PR, just removed from the default bundle. A third-party plugin doing the same thing is still technically on the table. OpenCode just can’t be the one shipping it anymore. Whether that holds up legally is anyone’s guess, but it’s not a hard wall.