TobyTheCamel

4 ms·

1.
by TobyTheCamel 3mo ago
There's a fantastic website called Difford's Guide [1] that allows you to list the cocktail ingredients that you have and will then provide a long list of cocktails you can make (optionally allowing substitutions). It is however a
2.
by TobyTheCamel 4mo ago
I think people are typically referring to the task-completion time horizon at a fixed success rate [1]. That has had pretty robust exponential scaling for many years now. [1] https://metr.org/time-horizons/
3.
by TobyTheCamel 4mo ago
Genuinely, thank you. This is very encouraging and makes me feel much better about commenting more going forwards. The quality of discussion and prose on HN is just generally so high that it can feel quite a bit intimidating to jump in (in
4.
by TobyTheCamel 4mo ago
Looks pretty exponential to me [1]. From a fully independent, non-profit research group. [1] https://metr.org/time-horizons/
5.
by TobyTheCamel 4mo ago
You're talking quite statically though. I don't think anyone is worried about today's models being a serious threat, but next year's, three years' time? Just three years' ago these models were useful bumbling f
6.
by TobyTheCamel 4mo ago
You're talking as if this is a static thing though. It's the God of Gaps [1] but for humanity's special sauce. Two years ago, I couldn't trust an LLM to do anything that wasn't straight forward boiler plate. One yea
7.
by TobyTheCamel 4mo ago
If Anthropic were not developing these models, one of many other companies would be. I think it's good that the CEO of the current world-leader is at least considering these discussions and platforming possible solutions. The fact that
8.
by TobyTheCamel 4mo ago
Your first point is very reasonable, and I agree that that is something Dario would likely be more opposed to. However, my point isn't that I think Dario is our saviour who we should follow the every word of. As with everyone, his opin
9.
by TobyTheCamel 4mo ago
Sure, I have no doubt that AI progress will follow an S-curve. The question is, where are we on it and is the plateau at a level safe for humanity? That's a very difficult thing to estimate without a crystal ball and not a risk I want
10.
by TobyTheCamel 4mo ago
I feel completely baffled by the other responses on this thread. People viewing this purely as a marketing stunt, regulatory capture or attack on their freedoms, with seemingly no appreciation of the real threat that AI could pose to societ
11.
Ask HN: What prompt do you use to get Claude to consistently render LaTeX?
6 points by TobyTheCamel 7mo ago|5 comments
12.
by TobyTheCamel 3y ago
> However, the idea is that often a lot of the probability mass - an amount that is not small - will be concentrated around the maximum likelihood estimate, and so that's why it makes a good estimate, and worth using. This may be tr