cjtrowbridge

6 ms·

1.
by cjtrowbridge 4mo ago
its really good at agentic tasks
2.
by cjtrowbridge 4mo ago
This has been easy with OpenSCAD for a long time. I have made lots of cool, complex models this way. I built a repo of the prompts I use to show the llm how to do this and it includes many of the models I've created this way... https:
3.
by cjtrowbridge 5mo ago
There are more elegant solutions to this problem. Why are you trying to get an LLM to work with bloated, archaic tools that you have to rent from a feudal lord in the cloud when there are free open-source alternatives like OpenSCAD. This is
4.
by cjtrowbridge 5mo ago
(this is how android has always worked)
5.
by cjtrowbridge 7mo ago
Falcon Heavy (as its name implies) is not capable as a super-heavy lift vehicle. Past GTO, it can only carry 18 tons. You need more than double that to reach the Moon and come back, as NASA did in the 1960s.
6.
by cjtrowbridge 7mo ago
You're confused. Elon said two weeks ago that they have given up on Mars and the Moon is the goal they're currently working on. He said it will probably take them another decade to catch up to NASA of the 1960s by reaching the Moo
7.
by cjtrowbridge 7mo ago
No, it's never made it to orbit.
8.
by cjtrowbridge 7mo ago
It's not possible to compare, because while the SLS just got back from the Moon and is about to go back; SpaceX has never had a single successful super-heavy launch. Now that Elon has officially given up on Mars and decided to spend th
9.
by cjtrowbridge 7mo ago
Only with LEO launches, and the soviet rockets from the 90s are still cheaper and more reliable at that. Enormous subsidies and sanctions against Russia are the only thing pushing anyone to spend more on inferior Falcon rockets.
10.
by cjtrowbridge 7mo ago
All of SapaceX rockets waste close to half their payload capacity on extra fuel for landing, extra equipment for landing, and they still have a 100% failure rate on every super-heavy launch they've ever attempted. SpaceX has blown up m
11.
by cjtrowbridge 7mo ago
A KPI is an ethical constraint. Ethical constraints are rules about what to do versus not do. That's what a KPI is. This is why we talk about good versus bad governance. What you measure (KPIs) is what you get. This is an intended feat
12.
by cjtrowbridge 1y ago
Radxa has rpi clones with better specs that already do that
13.
by cjtrowbridge 1y ago
This brings up a point many will not be aware of. If you know the random seed and the prompt, and the hash of the model's binary file; the output is completely deterministic. You can use this information to check whether they are in fa
14.
by cjtrowbridge 1y ago
This is an obvious third-factor for poverty and marginalization. Air pollution exposure is the most classic example of unequal protection from harm in environmental justice. Alameda county did a study on this that found as an isolated, dire
15.
by cjtrowbridge 1y ago
No. It's because large models have leveled off and commodified. They are all trending towards the same capabilities, and openai isn't really a leader. They have the most popular interface, but it really isn't very good. The f
16.
by cjtrowbridge 1y ago
two p40 cards together will run this for under $300
17.
by cjtrowbridge 1y ago
We already have an interplanetary internet called the NASA Deep Space Network. Understanding it's limitations and challenges is a good way to start thinking about this.
18.
by cjtrowbridge 1y ago
what does this have to do with the viking god of war like its not that hard to come up with some relevant name
19.
by cjtrowbridge 1y ago
[flagged]
20.
by cjtrowbridge 2y ago
P40s are so slept on. 24gb vram for $150.
21.
by cjtrowbridge 2y ago
Lmao
22.
by cjtrowbridge 2y ago
Bro use fine web. Random books are not objectively good training data.
23.
by cjtrowbridge 2y ago
This is only a threat to rent seeking, not the open web.
24.
by cjtrowbridge 2y ago
Entropy is also added via a random seed. The model is only deterministic if you use the same random seed.
25.
by cjtrowbridge 2y ago
Yes it's the same bands as lora
26.
by cjtrowbridge 2y ago
They released it on huggingface.
27.
by cjtrowbridge 2y ago
The naive idea that your views are somehow the inherently non-political and scientifically objective perspective is silly. The fact is that everything is political because there will always be people who build identities around denying real
28.
by cjtrowbridge 2y ago
I love how they include a helpful chart that shows this model scores worse than everything else.
29.
by cjtrowbridge 2y ago
Your GitHub link doesn't even work, and the js file being imported is just plagiarized from Google.
30.
by cjtrowbridge 2y ago
BS. Import doesn't work on static websites. This code is for ES6. This post is spam.

More ›