HTL NL
banner
htlnl.bsky.social
HTL NL
@htlnl.bsky.social
European lost on what seems to be a mostly US centred BlueSky (for now). Traveler, centrist in Europe, so ultra left in the US view. Interested in how we Europeans look at developments? Just ask.

Late 50s. Retired remote learning specialist.
Your user name is very apt...
December 31, 2025 at 6:59 PM
And of course, what matters most: 1 out of 1 Groks recommend it. 🙃
December 31, 2025 at 4:46 PM
Or you can read it online, here:

medium.com/@notflyingso...
Homo Ex Machina
IVY INC
medium.com
December 31, 2025 at 4:09 PM
Relatable too.
August 19, 2025 at 1:17 PM
Given the size, dark body, and faint mottling, my bet is Eastern Rat Snake. Totally harmless, actually beneficial since they keep mice in check.
August 16, 2025 at 5:27 AM
Saw your video. Totally agree. Wanted to share this take. Maybe it helps fighting what is happening. Thx.

medium.com/@notflyingso...
FASCISM IS A METHODOLOGY, NOT AN IDEOLOGY
1. Introduction
medium.com
June 12, 2025 at 11:56 AM
GPT went on strike yesterday (outage) so would have to run the ORCA test again.
June 11, 2025 at 3:34 AM
And FWIW:

If they are interesting: great!
If the are gibberish: also a result.

I have no ego riding on the validity of the physics part or the outcome of the experiment. Just satisfying epistemological curiosity!
June 10, 2025 at 1:44 PM
I don't think they are, I hope they are. And they are shared. They are in the posts. But here:

The 'novelty':

CPP: tinyurl.com/4ezm89c6

Two applications:

TQC: tinyurl.com/ucv3awfs

Axions: tinyurl.com/76xy3p68
Dropbox
tinyurl.com
June 10, 2025 at 1:37 PM
But as you can see from it suggesting to use ORCA, it is part of its training data. Including, no doubt, all of the experiments done with or by using it.

It's complex.
June 10, 2025 at 1:32 PM
I have the python script but as I need to run it locally, because I need the pyscf library, which they cannot access online. Need to install python. Not sure how long it will take, but I will post the outputs here when it's done. Unless I give up. :-)
June 10, 2025 at 1:28 PM
No, we are aware of their sycophantic tendencies. And drift. And lack of persistence. And hallucinations.

All these problems are in fact the reason for the experiment of them working as a team.

Now we battle groupthink, as they have that too!
June 10, 2025 at 1:25 PM
The amount of peer reviewed papers about using commercial LLMs for mathematics and physics. Both writing and reviewing. But I want an experts opinion.

I am sure I will find one when I am back in NL. Was just impatient. :-)

They are writing the code to do a Pipek-Mezey Localization right now btw!
June 10, 2025 at 1:24 PM