RobinL

4 ms·

1.
by RobinL 10d ago
I agree, and I think LLMs can potentially do a better job than more conventional methods, so long as they're provided with enough content. Another aspect of context that could be relevant to your work may be how people are nested withi
2.
by RobinL 10d ago
There's also two other important limitations to using an LLM and just providing it with pairs of records. It does not know enough about the records in the context of the overall dataset: - what is the data quality and to what extent do
3.
by RobinL 11d ago
> Half the point of this project is showing off some amazing public-domain natural-history illustrations. Over 800 cut-outs covering more than 400 species, every one taken from a real plate and hand-curated for this project (no art is AI
4.
by RobinL 12d ago
Really like this, well done!
5.
by RobinL 12d ago
I've been experimenting for a year or two making games that make mental arithmetic practice fun for kids. The latest is https://rupertlinacre.com/keep_it_going/ an infinite rollercoaster, which I think is one of
6.
A New Observable(new.observablehq.com)
4 points by RobinL 17d ago|0 comments
7.
by RobinL 18d ago
I wish I could do it better. There seems to be a lot of variability. This episode of radiolab on the topic is fascinating: https://radiolab.org/podcast/aphantasia/transcript?utm_sourc...
8.
by RobinL 18d ago
Interesting. I'm surprised they haven't prioritised/deliberately trained for this more because of how useful it is to ask codex to generate some images/assets/sprites when building sill games. It would dramatically
9.
Data Science in the Age of AI(robinlinacre.com)
1 points by RobinL 25d ago|0 comments
10.
by RobinL 1mo ago
Thanks - author here - I think this is due to the unexpected traffic spike from Hacker News! Will look into
11.
by RobinL 1mo ago
Thank you! My son's favourite at the moment is arithmetic annihilation in two player mode: https://rupertlinacre.com/arithmetic_annihilation/
12.
by RobinL 1mo ago
Also shout out to https://bustimes.org/map which is a less fun but more informative version of this. My son has spent many a happy hour on this site
13.
by RobinL 1mo ago
Author here. Wasn't expecting this to blow up! This is a quick prototype, made for my 7yo who loves buses. A few tips: - Click the freeze icon to be able to zoom and pan without it changing the location of the lens - Double click to p
14.
by RobinL 1mo ago
Click the freeze icon to freeze, or double click to choose a location for the lens
15.
Show HN: Buslens – where can I get to by bus? (UK)(rupertlinacre.com)
84 points by RobinL 1mo ago|55 comments
16.
by RobinL 1mo ago
Surprised no one has mentioned Gemini. That's the highest value add feature for me. I have pixel watch 3, which I got cheap second hand. Gemini is a little laggy but still useful. If the experience was similar to the latest open AI v
17.
by RobinL 2mo ago
https://rupertlinacre.com/arithmetic_annihilation/ A mental arithmetic tower defence game for kids. Now with head to head/two player. Code: https://github.com/RupertLinacre/arithmetic_annihi
18.
by RobinL 2mo ago
Incentives are important for allocative efficiency, whether that be the price of a banana or the incentive to do a job that pays more. Of course, they're imperfect and there are some cases where incentives lead to rent-seeking rather
19.
by RobinL 2mo ago
Yes - it's certainly the case at the moment that you can run a few thousand through the LLM at a reasonable price, but not, say, ten million. But the rate of progress suggests to me that this argument won't hold up forever. Event
20.
by RobinL 2mo ago
One example: I've been working for a couple years (not full time) on a high performance FOSS address matcher: https://github.com/moj-analytical-services/uk_address_matche... Until recently LLMs have been really b
21.
by RobinL 2mo ago
Having recently read Wilson's translation, I was immediately put off at the introduction and leapt to the conclusion it was bad. It just doesn't sound anywhere near as good to my ear. However, there may be more to it than I thoug
22.
by RobinL 2mo ago
I've often felt my job as a programmer has often been to automate myself (and others) out of my job. But I've always felt in control of that. I suppose there's some cosmic justice to the fact that it's now being done to
23.
by RobinL 2mo ago
Love this Sem. Will provide a great alternative to the SQL based connected components algorithm we ship in Splink. Looking forward to testing how much faster it is. Thanks for your work on it!
24.
by RobinL 2mo ago
There's also a pseudocolumn called rowid that's worth knowing about for similar purposes but on duckdb tables: https://duckdb.org/docs/lts/sql/statements/select#:~:text=Ro...
25.
by RobinL 2mo ago
By that logic someone working in a care home is worth less than someone building a gambling app. There's little relationship between salary and what people deserve, it's mostly just luck to have skills that happen to be in short
26.
by RobinL 2mo ago
I see it more as a way for relatively well off people to do a little bit for the environment rather than any sort of solution of high energy prices for the poor. The nice thing, of course, is that the more people buy them, the cheaper elect
27.
by RobinL 3mo ago
Does using chat in the new codex app use up your allowance? Any indication openai are moving to a model where al usage is metered?
28.
by RobinL 3mo ago
Same, no Sol (i'm on plus)
29.
by RobinL 3mo ago
It works fairly well empirically: my 7yo likes to watch trains and we regularly use this map to know when to expect one to pass. Not perfect, but pretty good
30.
by RobinL 3mo ago
I wrote a blog post a while back to address this question here: https://www.robinlinacre.com/recommend_duckdb/

More ›