ruaraidh
Joined 12 years ago
Karma
862
Submissions
32
Member since
Jun 5
Recent submissions
Comment · 9 days ago
How definitive. I wish I had your certainty about literally anything.
Comment · 12 days ago
I think maths is really the only solid case so far. I do wonder whether the big early contribution to experimental science will be in speeding up the scientists conducting it, at least until some non-negligible boundary in simulation fidelity is crossed.
Comment · 3 months ago
Ah, that's extremely cool! I'm enjoying hacking around with ESP32s etc at the moment (maybe it's just nice to do something physical that an AI can't) and it's amazing how mature the hardware ecosystem around them is. I'd love to hear more, both about the tech and the economics.
Comment · 5 months ago
Oh, that's such a good idea! I suppose the challenge is knowing where there are installable surfaces are (or at least making defensible guesses). I'm going to have a go at this...
Comment · 5 months ago
Good idea. I want to add specific options for different mounting locations (sheds etc) as well.
Comment · 5 months ago
Absolutely! I have some other datasets that might be useful too (e.g. air quality). Drop me a line: ruaraidh[at]southlondonscientific.com :)
Comment · 5 months ago
Thanks - I didn't know about Inspire Index, I'll check it out. I tend to agree about false precision. My first instinct was to use the synthetic horizon for addresses in that group, but I think that's over positive. A range might be better (if a bit more complex)?
Comment · 5 months ago
Thanks! Should be doable, I just got excited by the new shiny thing first.
Comment · 6 months ago
Opus keeps pointing out (in a fashion that could be construed as exasperated) that what it's working on is "obviously not malware" several times in a Cowork response, so I suspect the system prompt could use some tuning...
Comment · 6 months ago
Aeolus, a Python library to provide a unified interface for air quality data from sources around the world: https://github.com/southlondonscientific/aeolus
Air quality data is now very widely available, but managing access to multiple networks is a massive task (lots of shonky APIs out there - the EEA has a csv endpoint that actually returns a .zip with mimetype "text/html", just to give you a flavour). Integrating new APIs could be a full-time job, but it's something AI can do very well given a pattern.
This is really for me as I build out my company working on turning air quality data into actionable information, but it's open source and freely available.
Comment · 4 years ago
"high-risk, high-reward" science sounds great, until you suddenly have a replication crisis on your hands. Big, exciting results already attract more attention than good, careful, journeyman science, despite generally being, well, wrong.