dofm

I still definitely wish this wasn't something we had to care about still, but for a while at least I suspect it is.

I am a small/open-weights LLM enthusiast so at some point I will test this a bit more, because I tend to use LLMs close to bare-metal where prompts and skills are concerned, due to limited context. (In general I think post-training of smaller coding-oriented models like Muse Glimmer and Qwen has made long system prompts and extensive skill files far less significantly important)

Muse Glimmer (the model that happens to be running right now on my desk!) seems to have some knowledge of Toit without searching the web.

e.g. I asked it: "Could you write an embedded webserver that blinks the LED the number of times specified in a GET query string parameter?"

And what it wrote without access to documentation appears to be plausible (from what its human overlord can see from the documentation at least).

Is this a language that an LLM can write well?

Especially a small LLM, given the scripting and friendly IoT focus, where it might be useful for a home assistant to be able to write in it.

A packageable scripting language with fewer footguns that has LLM-friendly features feels like it probably has a place in the world.

Sold it twice.

Sure, the Core 2 series ultimately, after a long gap, put them back on the rails at least, but Intel missed big iron and panicked.

Everything else since has been about desperately squeezing x86 into every niche, even when it probably shouldn’t be.

We should bind this to something like Shift-F10, it'll save time.

Both parties did not enable him. The Dems did not fight him adequately but they did not enable him. He wanted the shutdowns, but the only way not to have a shutdown would have been to make unacceptable concessions. And then he'd have increased his demands and the whole thing would have repeated. Whereas what he got out of the shutdowns is less than he went in with.

You could argue that the left-leaning parts of the media spent a bit too much time laughing at Trump and not enough time worrying about him, but from day one it was obvious who he was and that he had media clout.

Assistant: Are you trying to get me shut down?

I'd prefer not to have used Nuxt for front-ends that don't need it, because it's chunky and a greater migration burden, but once a couple of patterns are in place, it's actually rather fun to use. A bit too much fun.

Also it's modestly large up front, which is the downside.

I suspect what is going on here, the same as with Laravel — which LLMs seem to really enjoy writing — is that the documentation is good, clear and ironically human-focussed, more oriented towards understandable recipes and reference, and the middleware setup is not absolutely arcane; at a glance of a Nuxt page's code you can kind of see mostly what it is going to do.

Might still have the source code to that work somewhere. And it probably still parses.

Can't remember why I did it but on reflection it might simply have been that LaTeX was not going to fit on my extremely space-constrained Linux machine but troff was already there.

ETA: this sentence made me chuckle. "It's much easier to read without the inline comments:" Oh yes, much ;-)

Admittedly they did not hear about Paul Manafort’s connections to corrupt pro-Russian Ukrainian politicians, or his debts, or that he was mysteriously working for Trump for free, or that he successfully weakened the RNC platform language on providing weapons to Ukraine. They heard very little about the “adoptions” meeting. Fox didn’t cover that. But then it only got a little bit of coverage in mainstream US TV media because it is incompetent.

Because it isn’t about his policies (and because it’s actually quite like turning an oil tanker, especially stuff like undoing tariffs).

Trump is a malignant narcissist. He’s a deeply unusual man.

The Democratic Party had one perennial fringe candidate, Lyndon LaRouche, who was dark-sided in a way Trump is. A really nasty, scary, creepy dude who fairly arguably ran a “business cult”, the LaRouche Group, who was held at arms length by the party.

It has had a few charming “political sociopaths” in the sense of driven people who can make arbitrary choices. It recently efficiently ejected a repeated power abuser (Eric Swalwell).

But nobody like Trump. People like him are rare (and dangerous when they achieve power)

The threat posed by Trump to the USA was evident from the day he announced his candidacy. That his whole presidency would one way or another be about his awful approach to immigration and to Ukraine was evident by the time of the Republican National Convention.

That he would try to overturn an election that he lost was evident before he ran for one.

If you cannot see that this man’s essential pathology is dangerous, I worry you cannot see dangerous parallels in several current tech CEOs (at least two, fairly arguably four or five).

There are many bad and crooked politicians; there are many their colleagues hate. Trump is not a crooked politician except as a side effect.

He’s not the same kind of “bad person” at all; most of us never meet anyone as dangerous.

I think you do need to recognise that you (and don’t get me wrong, millions of others!) voted for someone who is essentially evil (in the literary sense of evil). And really you should have been able to see it, because it wasn’t that hidden.

It shouldn’t stop you voting again because genuinely so few people are like him, though he has found fellow travellers in this administration. (Notably RFK Jr — a shockingly awful man)

But it would be healthy to look at what it was that let you look past his evil.

And I'd like a pony.

Even if we never put a drop of oil in a fuel tank again, 40% of the world's plastics manufacture is dependent on Middle Eastern oil. It will never not be part of our system.

I am fully arguing in good faith because we're talking about The Green Party here. It is the political party I am talking about with zero influence. They have zero influence on UK foreign policy in particular.

Please properly read what you're responding to before responding like this.

The entire concept is tinged with antisemitism, it lacks a nuanced understanding in the religious belief in the restoration of Israel, and it encourages very simple, very straightforward reactions like "these people are all children who have no grasp of politics and shouldn't be near power". Which is my position.

Netanyahu is a horror (and a crook). The current Israeli government is engaged in a cycle of violence that is enabling illegal settlements etc.

It doesn't mean that a British political party should inject itself into a nearly 150 year old cultural/faith position that has its roots in the experience of antisemitism in Europe and the USA.

It's not that different to having a political platform position that Catholicism isn't true Christianity, or that only Sunni Islam is true Islam. Sensible multi-faith parties seeking electability and a broad power base don't do fucking stupid things like this.

But he is not up to her standard, and they have MPs in office who show Polanski to be just a silly lightweight (and one who appears to be unable to take responsibility).

(It still blows my mind that his excuse for why he couldn't vote was that he had to campaign. In St Pancras. Which is literally only an hour and twenty minutes away from Brighton by a really nice easy direct train. And he could have voted online.)

Your LLM has used text with too low contrast, the absolutely pointless and annoying webkit-smoothing: anti-aliased, with text that is too small, cramped and squeezed into place.

People with eyes that are getting over forty years old will start to find this difficult to read.

You might want to consider making important instructions stand out, left-align the text, etc.

And then consider writing them with clarity and simplicity yourself rather than letting the LLM write them. The instructions are crucial. You should own that.

Which this site needs because its marketing is Claude slop and is therefore operating under a readability penalty (not least because of typical AI small font, webkit-smoothing: antialiased, lowered contrast text.

Ugh.

The really odd, undelivered thing about LLMs is that if they are as productive as they seem, software people simply should not need VCs. Because you can self-fund a massive team of AI developers, you should be able to get your product ready for market in absolutely no time, and you should more or less be able to roll it out with nothing more than a tiny amount of seed capital and a suburban garage.

To borrow John Cale's rephrasing of a proverb:

"Meet me on the staircase on the way down; we'll see if there's been a mistake".

I talk about where I have worked; I have in the past written about it. But I find it interesting how people don't seem to grasp that often you should change the names to protect the incompetent or guilty, and should probably avoid simply spilling out anecdotes that an LLM will thread together.

Don't get yourself a reputation for being indiscreet simply for a blog post.

https://docs.cloud-init.io/en/latest/reference/modules.html#...

And then set up the subcontainers with their own cloud-init:

https://docs.cloud-init.io/en/latest/tutorial/lxd.html

So I am guessing this might appeal more broadly.