They are, the model has no inherent knowledge about its confidence levels, it just adds plausible-sounding numbers. Obviously they _can_ be plausible, but trusting these is just another level up from trusting the original output.
I read a comment here a few weeks back that LLMs always hallucinate, but we sometimes get lucky when the hallucinations match up with reality. I've been thinking about that a lot lately.
> the model has no inherent knowledge about its confidence levels
Kind of. See e.g. https://openreview.net/forum?id=mbu8EEnp3a, but I think it was established already a year ago that LLMs tend to have identifiable internal confidence signal; the challenge around the time of DeepSeek-R1 release was to, through training, connect that signal to tool use activation, so it does a search if it "feels unsure".
Wow, that's a really interesting paper. That's the kind of thing that makes me feel there's a lot more research to be done "around" LLMs and how they work, and that there's still a fair bit of improvement to be found.
In science, before LLMs, there's this saying: all models are wrong, some are useful. We model, say, gravity as 9.8m/s² on Earth, knowing full well that it doesn't hold true across the universe, and we're able to build things on top of that foundation. Whether that foundation is made of bricks, or is made of sand, for LLMs, is for us to decide.
G, the gravitational constant is (as far as we know) universal. I don't think this is what they meant, but the use of "across the universe" in the parent comment is confusing.
g, the net acceleration from gravity and the Earth's rotation is what is 9.8m/s² at the surface, on average. It varies slightly with location and altitude (less than 1% for anywhere on the surface IIRC), so "it's 9.8 everywhere" is the model that's wrong but good enough a lot of the time.
It doesn't even hold true on Earth! Nevermind other planets being of different sizes making that number change, that equation doesn't account for the atmosphere and air resistance from that. If we drop a feather that isn't crumpled up, it'll float down gently at anything but 9.8m/s². In sports, air resistance of different balls is enough that how fast something drops is also not exactly 9.8m/s², which is why peak athlete skills often don't transfer between sports. So, as a model, when we ignore air resistance it's good enough, a lot of the time, but sometimes it's not a good model because we do need to care about air resistance.
Gravity isn't 9.8m/s/s across the universe. If you're at higher or lower elevations (or outside the Earth's gravitational pull entirely), the acceleration will be different.
Their point was the 9.8 model is good enough for most things on Earth, the model doesn't need to be perfect across the universe to be useful.
Does Blender also qualify? It even shows you the name of the Python function behind each UI element on hover, which is great for discoverability when scripting. Or maybe it used to, can't see it now.
The rise in autism rates (5x over the last 25 years) is almost entirely explained by changes in how autism is defined, diagnosed, and detected. Not by an actual surge in underlying cases or any specific environmental trigger like vaccines, air pollution, heavy metals, plastics, or screen time.
I just don't believe that based purely on my personal experience. I've been around many many people and in the last 15-20 years the amount of people with autism has exploded. It was very rare 30 years ago in my area, the occasional led exposure here, farm chemicals there... But now it's seemingly every 10th house that has someone that is nonverbal.
I'm all for papers and science but only if it matches the real world.
It's come up every time something related to thermal printing has been mentioned on HN lately, but this is honestly great stuff if you're in Germany: https://www.oekobon.de/
These non-poisonous blue receipts have the added benefit of being able to be marked with a fingernail, which is nifty if you're using them to print your shopping list, crossing things off is very satisfying.
> I was able to improve my sleep because I found out that my waking up in the night was correlated with high co2 values.
What did you change? I assume it's non-trivial to automatically open windows based on sensor data? Or do you mean you've been able to improve it by knowing about it and opening a window before going to sleep, which now that I write it sounds much more sensible :D
Haha, opening a window automatically was my initial idea, but like you I quickly converged on the latter! In addition, I just keep the bedroom door open as well, which helps a little bit too
Not just a $PROPRIETARY_THING, but a random unaffiliated project that tries to nab users by using almost the same name as a well-known and well-appreciated project. It should count as malware.
I read a comment here a few weeks back that LLMs always hallucinate, but we sometimes get lucky when the hallucinations match up with reality. I've been thinking about that a lot lately.