alansaber

7 ms·

61.
by alansaber 28d ago
Your honour, this 40 page flowchart (below the 8B LOC lean proof in addendum C) proves that my client had planning permission for his loft.
62.
by alansaber 28d ago
Yes but this is more an artifact of how much PDF sucks as a file format than LLMs being crap.
63.
by alansaber 28d ago
"Sounds kinda boring" - many do say this about lawyers.
64.
by alansaber 28d ago
Yep this is one of the best use-cases for a fine-tuned SLM/SVLM.
65.
by alansaber 28d ago
Compounding like this ignores fundamental friction. You can attempt a huge document processing factory but quality will go into the toilet.
66.
by alansaber 28d ago
Wait what, the conversion is so lossy you use an LLM to analyse it? Or you get the claude analysis into JSON and put that into the system too? I'm not seeing what's causing the 2-3x efficiency bump.
67.
by alansaber 29d ago
Exactly, a change like this will always be abhorrent to the 5% of technical users (aka this forum), and a net gain for the other 95%
68.
by alansaber 29d ago
I guess you can pay OpenAI for the privilege of hosting a RAG database rather than having their agents trawl your site.
69.
by alansaber 1mo ago
Agreed. My impression is that the more verbose output of sol, astra etc is that it helps it steer itself on long running tasks (but is worse for the human user to read)
70.
by alansaber 1mo ago
The regulations often change the economics of production to favour improved quality/attempting to sell upmarket.
71.
by alansaber 1mo ago
Because everyone at every step of the chain is primarily incentivised by money.
72.
by alansaber 1mo ago
...some kind of subscription as a service model? looks up dramatically
73.
by alansaber 1mo ago
Historically, they're lax on regulation. Hence the association with bad quality.
74.
by alansaber 1mo ago
Such is the nature of being part of a global supply chain, whilst living in one of the richest countries in the world.
75.
by alansaber 1mo ago
Perfectly stated and also explains why it slips past the author (they just nod and let it slip pass).
76.
by alansaber 1mo ago
"please do not break out of this sandbox make 0 mistakes". The timeframe is a little suspicious, not sure beyond that. Though I enjoyed the scroll effect on the website.
77.
Large-Scale Qualitative Research with AI(arxiv.org)
2 points by alansaber 1mo ago|0 comments
78.
by alansaber 1mo ago
You can use AI to brute-force different crystal configurations, run some MD/DFT and identify candidates. I'd be (not completely, but) shocked if there wasn't a research group doing this already. To rapidly iterate with physic
79.
by alansaber 1mo ago
Just go to any beginner class for social dancing (salsa, bachata, etc etc). They'll have had older/less co-ordinated, trust me. It's fantastic physical, mental exercise and a great way to meet people/be part of a communi
80.
by alansaber 1mo ago
There are just fewer male dancers. I've gone to events with insane ratios either way (3x more leaders/followers) but it's rare. Usually there are more leaders at the beginner level and decreasing amounts as you go up (but mor
81.
by alansaber 1mo ago
"Dance clip" seems to do exactly 0 things I can't already do with the video editor/folder system on my phone.
82.
by alansaber 1mo ago
It's not just copyrighted training data. Truly open source e2e model training would include scrapers, data cleaning, all pretraining scripts, posttraining scripts, exact hardware info, etc. Open weights labs will release a sanitized ve
83.
by alansaber 1mo ago
Anthropic's sense of ethics is completely aligned with its economic incentives. Shocking, right.
84.
by alansaber 1mo ago
Exactly, people are jumping at the shadow on the wall- LLMs have achieved remarkable things regarding their limitations, but those limitations show no signs of yielding
85.
by alansaber 1mo ago
Why draw the line at the industrial age? Humans subsisted for millenia before the ready preparation of fire.
86.
by alansaber 1mo ago
Technology advances, but extrapolating the value of human labour to 0 is equally invalid.
87.
by alansaber 1mo ago
I agree with this. The substansive counterarguments all revolve around the short-term economic and emotional instability this causes
88.
by alansaber 1mo ago
The frontier companies asking for regulation is somehow one of their less insane marketing ploy to date
89.
by alansaber 1mo ago
Sometimes it's good enough to acknowledge a correct answer, and slowly build intuition towards it. Academic problem solving is rarely a straight path, anyway.
90.
by alansaber 1mo ago
Griftmaxxing and terrible ideas tend to go hand in hand

More ›