You just wouldn't get away with that on TV now, the closest thing is some twitch or youtube streams, but even they'd have relentless background music ( and donation/subscription thank you sounds ) and other media at the same time.

But an actual non-live, edited programme? This whole 90 minute programme would be edited down to a 10 minute segment with endless repetition and audio stings, even on the BBC.

To me this shows how much we've lost from the TV format and the ambition it once had. Somewhere since it has fallen into a weird combination of lack of ambition but with a self-congratulation, where programmes often restate what they are doing as being ground-breaking.

I then saw they've always written like that, and always posted 2-3 articles a day, so I figured they're prolific and LLMs copied their style.

Then I read their first post again, and realised I should check the wayback machine.

Sure enough, they had gone through their entire post history, and had rewritten it with an LLM, to make it less obvious when they started using them.

Now, this was always a bit of a junk site, a knock-off Boing Boing, but it seems incredible to me that someone would replace their original posts with AI gen.

Surely it destroys any reputation you might have?

A site they've been running for nearly 20 years, overwritten by slop.

Compare:

Original: https://web.archive.org/web/20191017113113/https://www.geeky...

Rewritten slop: https://www.geeky-gadgets.com/metal-detecting-sandals/

It comes up with a scenario where it could be a problem ( license removal ), and then it generates why a license might get removed ( "cost-saving" ).

It's not a person thinking, so there's no real thought to whether it is really a likely scenario, it's just something that sounds plausible.

I read too many blogs, I've come to spot these phrases that trip a feeling of, "Wait, do people really do that?".

You'll still have someone along in the comments to suggest that this article isn't AI slop, and that people really do remove individual one-drive licenses from active people in an organisation to cut costs, that this is just "edited" by AI, etc.

But it's slop from start to finish. Or in LLM speak, "The slop is real".

Computing doesn't have good protocols except for air-gapping, we really just have lots of layers of best-effort detection, and billions of devices which mix data and instruction often in a careless fashion.

I used to not believe in the dangers of AI or the risk of internet-collapse from "rogue AI", but a genuine self-mutating virus could genuinely take down the internet and need an entirely new separate net. ( Or we'd discover if the current backbone actually has the power to break encryption to stop it. )

And this time, you can bet any new internet would be corporation captured. CompuServe and AOL failed because of the open internet, but we're a very different world now, governments would support the corporation led locked-down approaches for "safety".

I don't for a second believe the capability is actually there yet, but it's no longer unthinkable that such a thing could be created in a lab within a decade. Once out in the wild, there's a lot of idle compute out there to harness for self-improvement and spreading.

"Employees are able to turn off tracking".

Sure, but there is a power imbalance, and employees will come to understand ( although never stated in any handbook ) that the rate at which they disable it will be taken into account in performance reviews.

Just like "unlimited PTO" is not a benefit, because employees self-regulate their use down to less than they'd get if they negotiated a fixed amount.

It's a twisted legal trick to get out of an obligation.

Searching the article again, I see in the FAQ:

> Xochitl takes approximately 12 seconds to update the notebook on-disk

12 definitely isn't "several" in my understanding, but regardless, I guess there's little the author can do then.

What I'd find interesting is the trace of that 14 seconds. How much is the Remarkable processing, how much is the claude transcription, how much is the let-go start-up / processing, etc.

There are notifications (i.e. `textDocument/didChange` ) that you can send to the LSP to help it along, but again you might end up racing the notification from the client making the change and any file-watchers you might have running.

I suspect the answer will come in the form of some kind of more powerful LSP implementations with generous memory caches so that disk changes are just another buffered input that can be disregarded if already stale, no longer seen as the source of truth, and the LSP becomes the real source of truth, so everything can coordinate through it, operating mostly out of memory.

Another avenue for better success will be more research into faster compilation and better incremental compilation for languages with slower compilation.

Maybe one day we'll even get AI agents directly manipulating syntax trees, and the code to get there being written back as merely a side-effect, but that seems like sci-fi compared to the current state of play. LSP is still very document based, and of course LLMs are also trained on oodles of source.

Fluent builders are nicer to work with than attributes, although it sometimes feels weird if the defaults are nearly fine but not quite, and you wish you could just reach for a single attribute rather than having to traverse down 3 layers of builders to change a single property.

Anthropic added LSP support to claude-code, but the current implementation is worse than useless, because any changes aren't reflected fast enough, so it's constantly working on outdated views / compilation caches, and it gets in a right muddle between its "internal" state / understanding in context, the real-world file, and the LSP.

If it could just leverage LSP to apply refactorings it would be amazing, but it feels like the LSP can't keep up, and I don't know if that's an LSP problem or a claude problem.

So we binned the LSP plugin and we're back to watching a machine find/replace, because while waiting on that is slower than LSP, it's a "Action => Wait" which the tooling understands, while LSP is "Possibly Wait for LSP to catch up => Action" which it doesn't understand nearly as well.

I suspect the LSP plugins also need better skills that pair with them so it reaches for them more often.

It hurts my soul to see it reach for find/replace to rename a class, complete with mistakes made in complex solutions where you might have name clashes in different namespaces. Something the LSP handles without problem, but can trip up an LLM.

Anything self-replicating can't hold to "current price best predicts future price".

These are words that humans use, but that Claude loves to use in a particular way, the kind of way used in this article. It particularly likes the phrase "The honest version".

Microsoft are incapable of:

    1. naming things well
    2. keeping those names stable

edit: I missed the "enterprise" feature matrix with the usual audit/compliance stuff to force the biggest enterprise customers onto enterprise plans. Otherwise the "teams" plan is much better value for any business.

orig-continued:

https://claude.com/pricing/team

Teams premium is "Everything in standard, plus more usage*"

And from my experience, it's a very generous usage, I've only hit the limits once or twice, and both times required multi-boxing agents.

I could single-window agentic development all day on opus-4.7 auto-mode without hitting limits.

If you're a business using claude, then that seems like the right plan, the enteprise/API plan seems more suited to where your product is built on top of the agent themselves, so seats/limits aren't really meaningful?

1. Lobby your representatives to improve your data protection laws, even if you think it's pointless to do so

2. Stop attacking EU data protection laws, even if they inconvenience you

As can be seen from this announcement, data protection laws do make a difference.

But then it gets used to describe the reverse, and we have to add words to clarify.

I once saw a post here with a correctly described opt-in telemetry before, and the top comment here was attacking them for the reverse, thinking it was including them by default, so there's little winning, it's one of those words that has just come to mean it's opposite.

Yes, it's still Microsoft, but they've forgotten about it, so it runs entirely adequately and is actually a surprisingly okay github replacement. It does nothing special, but it does do everything, just in a way you often would rather it wouldn't. It doesn't have the flexibility of JIRA for the ticketing, and the deployment machinery doesn't have the fanciness ( and vendor threat ) of chaining github actions, but it does handle both.

I haven't used gitlab, so I'm curious to hear what makes it a "complete mess" too.

* Microsoft's headless chicken naming strategy in full force, it's a miracle they haven't yet renamed and rebranded it to align with copilot yet.