matthewsinclair

Looks like it’s been fixed. Thanks.

Apologies for the typo. I’ve asked mods to fix it.

I wonder how this works for code generation as opposed to general text generation?

There's some more here [0] and here [1].

[0]: https://www.anthropic.com/research/global-workspace

[1]: https://www.youtube.com/watch?v=rKV5JcALQoQ

I’ve been referring to it as “Dunning-Kruger As A Service” for quite some time. Interesting that other folks are coming to the same conclusion.

There’s a subtlety here, tho. It doesn’t mean (at all) that’s it useless. Just that those who don’t know what they’re doing are going to be amplifying their incompetence.

I’ve had some pretty good results with Claude Code over the last 18 months or so, and “agentic engineering” (as opposed to “vibe coding”) is now very tightly integrated into my workflow. But it’s pretty easy to see how it can devolve to Dunning-Krueger-as-a-service if you don’t know what you’re doing. Which sadly appears to be the case far more often than it’s not right now.

Now we’re into a conversation about degree, and not principle. Yes, it could be small, applied like a sales tax a percentage of the spend on advertising. The idea would be to create a price disincentive to do it. Ultimately, it’s a form of digital pollution, so the less we have of it the better (in my view).

I know this is controversial, and imperfect, but the longer this crap continues the more right I feel about this as a solution:

What if we taxed advertising? https://matthewsinclair.com/blog/0177-what-if-we-taxed-adver...

The enshitification will continue until morale improves.

As an Aussie, it’s always amused me that we managed to name a public swimming pool after a Prime Minister who (at least as the official story goes) died whilst swimming.

I went back about 2 years ago and speed ran all of them. Somehow the sequels seemed substantially better than I remember them. In fact, on the first watch I was _so_ disappointed with 2 that I didn’t even watch 3. 20-odd years later, I had a very different (positive) experience. Not sure why. Maybe I was in a good mood.

What? I thought the idea of running Linux in a PDF was pretty cool. Apologies if that bummed your day.

And here's the PDF: https://linux.doompdf.dev/linux.pdf

“Siclair” or “Sinclair”?

Congrats! Nice work. I’m a fellow “Phoenix |> Elixir” fan and built something similar here [0]. I’m still ironing the kinks out of mine, but I’ve moved all of my personal sites over to it and it’s been running for almost a year now.

[0]: https://laksa.io

And there’s also the Ash framework for Elixir.

Totally agree with this, too! The books are great and I’ve read all four a few times over, but the audiobooks are something else. The guy who reads them manages to strike a near perfect balance between “reading” and “acting” that is just such a pleasure to listen to. I think we must have listened to it beginning to end about another three or four times as a family during long car trips!

100%. One of the genuine great writers.

Vale Dan Simmons. You brought the world a _lot_ of joy.

I have dreams, but they're almost always non-visual, at least in the traditional sense, or when compared to other people I know who recount their dreams to me, and I very rarely "remember" (if that's the right word) them in any detail. I sometimes (rarely) will have visual dreams, but that almost always happens after I've had a good night's sleep, woken up, and then fallen back to sleep again. I think it might have something to do with how deeply I sleep when I am asleep. A train could drive past my bed and I wouldn't wake up.

Hello from the other side!

I was about 45 before I realised that when people said “in my mind’s eye” it was substantially more than a metaphor.

And it wasn’t until about a year later that I realised that I also didn’t have what ordinary people refer to as an inner monologue.

Realising that I had both aphantasia and anendophasia was quite a shock, but has never felt to me like I was missing anything.

For images I literally have nothing “pictorial” or “graphical” at all, but concepts and relationships are “vivid”. And for the inner monologue, there’s no autonomic voice at all, but if I concentrate in the same way that someone might “consciously breathe” I can kinda sorta trigger something.

Interestingly, in periods where I have meditated for >20min per day for consecutive weeks, I can trigger what I refer to as “flyover mode” which is like a literal landscape flyover that feels like a 4K screensaver. But this is _rare_ and requires a huge amount of effort.

Weird, eh!?

I agree!

“Why LLM-Powered Programming is More Mech Suit Than Artificial Human”

https://matthewsinclair.com/blog/0178-why-llm-powered-progra...

I agree [0]. Well, taxed rather than banned. But we’re in the same postcode.

[0]: https://matthewsinclair.com/blog/0177-what-if-we-taxed-adver...

Here’s a counterpoint that attempts to argue an alternative view:

https://xeiaso.net/blog/2026/markdownlang/

Indeed. There’s definitely a bit of that in there, no doubt. But I think, after reading it, that it’s one of those articles that might off on some details but is in the ballpark in terms of trajectory.

Here’s a link to the original blog post (to avoid Xitter): https://shumer.dev/something-big-is-happening

Hyperbole? Hmm.

That article is fantastically well written. What a trip.

Yep. Totally. And there was this dreadful wait for a load that could (randomly) fail and you’d need to go thru the whole process again!

Parsec! And what about “Tunnels of Doom [0]” !? I spent so much time on that thing.

[0]: https://en.wikipedia.org/wiki/Tunnels_of_Doom