te_chris

A thankfully American reference

GCP is miles better. Their IAM is at least understandable, for a start.

“EC2 or Lightsail”. And this right here is why I use GCP. Google got VMs right.

Why am I back on the terminal and 40k lines deep in my own ai-workflow tui? Two reasons: 1) AI writes my dotfiles. It fucking rules. 2) charm bracelet tui ecosystem in go is ace and ai writes it well.

What? In what world should the tool be given due credit.

Claude code and codex do this all the time too. Fucking annoying.

Cool that they don't let others do it. I tried to make a work fb account to access the hellscape that is business manager and they blocked it. Pricks.

the country isn't blocking his site. his self-importance is. OFCOM doesn't give a shit about this blog.

I’ve built this into a cli TUI. Passes the whole diff to Claude code with a schema and gets a structured narrative back out. Works really well for understanding.

Reconstituting messy things is exactly where LLMs can help.

I try codex, but i hate 5.4's personality as a partner. It's a demon debugger though. but working closely with it, it's so smug and annoying.

I only pay adobe because, as an amateur, I can dump in raw files and they then make them go away from my hd. It's wild that no one else offers this, but that seems to be where we are.

How does BM cloud work in this regard? Can we dump a card straight in, have it sync, edit, export etc and never think about the files again?

Yes. Obviously it does.

Honestly: jog on.

You can see the transition happening. Right now.

Without wanting to sound rude: I think the mistake people make with AI prototypes is keeping the code at all.

The AI’s are more than capable of producing a mountain of docs from which to rebuild, sanely. They’re really not that capable - without a lot of human pain - of making a shit codebase good.

So naive. Private only works this way in Britain because it doesn’t have to be responsible for anything. It’s a luxury good and works accordingly.

We have insurance, it’s amazing! But it’s fake. If you want to know how a whole system of this would work, look at the US

And you’ll never guess who wrote it…

What are you talking about? £206pm is the low range of the car leasing market, and the price per km is under petrol. Millions of people lease cars in the uk and can get favourable tax deals through their employers. So yes, still bollocks.

Honestly the car is a luxury for us and a minor tax dodge. I was a consultant with some retained profits in a ltd co so we decided to take advantage of the lease deductions. Got a great deal on a top end VW - £206pm ex VAT.

The key is to remember what they are: bags of weights which you’re throwing some data into.

In that sense they can’t offer advice because the “know” nothing.

But they can reframe, they can reflect. They can take one idea and reflect it into another intellectual framework.

I’ve used them a lot like this to help get perspective on life decisions. But not for advice.

Try something like: I have to do x and y, give me multiple psych perspectives on this problem from different schools. I find this takes something abstract (your problem) and grounds it in things it actually knows (the sum of ingestible human knowledge).

I’m just providing a data point.

I don’t see why it would be different in another city, and in a less dense one you can home charge which is dirt cheap.

0.49p, gets around 4 mi/kWh.

ChatGPT reckons this gives 7.6p/km when petrol would be 9-12.

Obviously it could be cheaper, but it’s still pretty good for us. We’re not using it for commuting or anything though.

Bollocks. We live in a flat in London, we have an EV. Charging isn’t cheap like if I was charging on my home tariff, but still cheaper than gas and there’s lots of options around me. I get a discount as a local resident too.

Your points on home solar and heat pumps are accurate, but people are working on splitting solar and battery across flats, e.g. Negativity isn’t the answer; especially not when there are actual solutions available.

This looks nice, but on mac you can virtualise really easily into microvms now with https://github.com/apple/container.

I've built my own cli that runs the agent + docker compose (for the app stack) inside container for dev and it's working great. I love --dangerously-skip-permissions. There's 0 benefit to us whitelisting the agent while it's in flight.

Anthropic's new auto mode looks like an untrustworthy solution in search of a problem - as an aside. Not sure who thought security == ml classification layer but such is 2026.

If you're on linux and have kvm, there's Lima and Colima too.

Wild that you're getting downvoted for saying something so obvious. Weird.

There’s turbopuffer

Film gets better results with less effort - but more money.

I reviewed the LiteLLM source a while back. Without wanting to be mean, it was a mess. Steered well clear.

London is pretty hot right now. But lots of the jobs are Founding Engineer roles at startups which I’m deeply skeptical will get out of seed.

“Rather poor reds”. Um what? Germany and Austria produce a lot of excellent wine, red and white.

Austrian reds are some of my favourites.

In London, Georgian orange wines are very popular and that seems to be what most people who are cursorily into wine would know them for here.

Nationalists are all the same and all hate the country as it is vs how they imagine it to be - see the uk brexiters ignoring science and the creative industries.

Most of all they hate intelligent people as they see their schemes for what they are.