Nostr notes by David Huggins-Daines

RE: https://mastodon.energy/@Sustainable2050/117349913072181143 ...

2026-09-28T18:17:55Z

RE: https://mastodon.energy/@Sustainable2050/117349913072181143

Basically every poll in every developed country everywhere 🫠

quoting
note10lq…xt47

On average, the Dutch think that 20% of the people here is Muslim (it's 5%).
They overestimate our income tax rate by 28%.
They think the number of homicides is on the rise (it's falling).
And that agriculture takes 19% of our land area (it's 60%).
And they have strong opinions on all of the above.
https://fd.nl/opinie/1613390/landbouwland

but also, despite his association with Thatcher and Reagan, ...

2026-09-01T12:34:17Z

In reply to nevent1q…wmg6
_________________________

but also, despite his association with Thatcher and Reagan, Mulroney was nonetheless a member of a totally different party than today's CPC.

Maybe because I came of age in the 90s but I think the Chrétien Liberals did a lot more austerity, with a lot more lasting negative effects, than the PCs ever could in the 80s.

also for many things a filesystem really is the best ...

2026-08-31T15:10:08Z

In reply to nevent1q…6x5x
_________________________

also for many things a filesystem really is the best representation, so for example instead of MBTiles which has lots of overhead, you can get most of the benefits of a structured, self-documenting format (as opposed to a Protobuf abomination which shall remain unnamed) by simply putting your tiles in a loop-mounted btrfs. I love when the answer to "what is the tech stack" is "there is no tech stack": https://github.com/hyperknot/openfreemap#what-is-the-tech-stack

to be honest I have mixed feelings about the use of SQLite in ...

2026-08-31T14:57:17Z

In reply to nevent1q…ra50
_________________________

to be honest I have mixed feelings about the use of SQLite in these cases as the resulting files are much larger, for example here's a GPKG which is nearly *four times* larger than the (proprietary) ESRI FGDB format: https://www.donneesquebec.ca/recherche/fr/dataset/decoupages-administratifs

the geospatial community seems to have come to this conclusion, ...

2026-08-30T13:40:13Z

In reply to nevent1q…gfxx
_________________________

the geospatial community seems to have come to this conclusion, see GeoPackage, MBTiles, etc

(It helps that Shapefiles literally have the 1986-era equivalent of SQLite "in" them, i.e. not actually in them because http://switchfromshapefile.org/)

big "the self-driving cars can't help but hit pedestrians ...

2026-06-11T14:10:01Z

In reply to nevent1q…l3u4
_________________________

big "the self-driving cars can't help but hit pedestrians so let's just put up fences to prevent pedestrians from crossing the street anywhere" energy

Il faut rappeler le plus souvent possible aux lecteurs mais ...

2026-05-26T13:30:29Z

Il faut rappeler le plus souvent possible aux lecteurs mais SURTOUT aux chroniqueurs du Le Devoir (npub1fk4…lf9n) que, en bon québécois, you do not under any circumstances have to hand it to Mathieu Bock-Côté https://www.ledevoir.com/opinion/chroniques/982007/mathieu-bock-cote-exalte

Malheureusement on n'a plus les péquistes qu'on avait ...

2026-03-19T14:20:46Z

Malheureusement on n'a plus les péquistes qu'on avait dans son temps 😞

indeed, I guess the question is whether the scale of the ahem ...

2026-02-20T15:10:46Z

In reply to nevent1q…py2v
_________________________

indeed, I guess the question is whether the scale of the *ahem* waste, fraud and abuse *ahem* of resources that LLMs seem to imply, even in benign use cases like yours, is out of line with historical precedent or not.

Am I an old man yelling at a cloud?

No, it's the children who are wrong!

As always, yes and no. A bug zapper is designed to zap bugs, it ...

2026-02-20T15:05:11Z

In reply to nevent1q…j0kv
_________________________

As always, yes and no. A bug zapper is designed to zap bugs, it is a simple mechanism that does that one thing, and does it well. An LLM is designed to read text and generate more text.

That we have decided that the best way to do NLP is to use massively overparameterized word predictors that we have trained using RL to respond to prompts, rather than just, like, doing NLP, is just crazy from an engineering standpoint.

Rube Goldberg is spinning in his grave!

But ALSO: using a multi-billion-parameter synthetic text ...

2026-02-20T14:18:58Z

In reply to nevent1q…8mak
_________________________

But ALSO: using a multi-billion-parameter synthetic text extruding machine to find spelling and syntax errors is a blatant example of "doing everything the least efficient way possible" and that's why we are living on an overheating planet buried under toxic e-waste.

If I think about it harder I could probably come up with a more clever metaphor than killing a mosquito with a flamethrower, but you get the idea.

I agree in principle with Cory, but I really wish that he had ...

2026-02-20T13:49:01Z

In reply to nevent1q…3rfr
_________________________

I agree in principle with Cory, but I really wish that he had clarified that:

1. Ollama is not an LLM, it's a server for various models, of varying degrees of openness.
2. Open weights is not open source, the model is still a black box. We should support projects like OLMO, which are completely open, down to the training data set and checkpoints.
3. It's quite difficult to "seize that technology" without using Someone Else's Computer to do so (a.k.a clown/cloud)