maxbond

9 ms·

91.
by maxbond 4mo ago
No indicator of AI writing is going to last forever. We really shouldn't over index on the oldest indicator (which is liable to decay first). People are starting to say "delve" more in verbal conversations. This treadmill wil
92.
by maxbond 4mo ago
> Banned from Venmo, absolutely no idea why. SO banned from Venmo, absolutely no idea why. Prior to 2013 or after? Maybe they merged ban lists with PayPal (who owns them).
93.
by maxbond 4mo ago
Interesting, so if I'm understanding correctly, component A's style was supposed to change when component B was present, and this was implemented as styling rules in component B? Why was Tailwind necessary rather than moving these
94.
by maxbond 4mo ago
There are dozens of us
95.
by maxbond 5mo ago
No one in this thread appears to disagree. The issue is that RLHF is prohibitively expensive and the number of disciplines you could target is massive, so for reasons of economics rather than fundamental theory, AIs do not perform well on t
96.
by maxbond 5mo ago
I don't know if you're being factitious or not but that was not what I meant. Picasso being a great painter is an example of "having taste"; "create an artistic image generation model with Picasso-level performance&
97.
by maxbond 5mo ago
It means having taste. People say Picasso was a great painter, but that cannot be verified (at least, not in the sense of a verified reward).
98.
by maxbond 5mo ago
Reminds me of the recent paper about delegating document editing tasks to LLMs across different disciplines [1]. That paper found that programming was the only discipline most LLMs can perform long horizon tasks on without accumulating erro
99.
by maxbond 5mo ago
I think there is an implication in context that the people being discussed aren't being reasonable (that the claim is employed as a rationalization), but I agree with your take. I should've said, "the downvotes were more like
100.
by maxbond 5mo ago
Good catch. We'd have to integrate with jinja2 (or similar) and tokenzize as we format the context, so that we know which spans are instructions and which spans are data. Which makes it more complex but still very achievable.
101.
by maxbond 5mo ago
More likely people thought GP was missing the point; "MTTR-optimized YOLO deployment" only succeeds against recoverable errors and acceptable periods of downtime against errors that are detected quickly. You could have a bug silen
102.
by maxbond 5mo ago
The escape algorithm here is very simple, you remove special tokens from the runtime tokenizer's vocabulary so that it's forced to encode them as multiple non-special tokens. (That doesn't actually mean the LLM won't tre
103.
by maxbond 5mo ago
I think if you saw it on a bench hooked up to wires, it would be intuitive that it was a circuit. It's equivalent, but instead of being coupled via wires it's coupled via RF. I think it feels like there's no return path and t
104.
by maxbond 5mo ago
It had no internal power supply, it worked like an RFID tag, but it was electronic. Tangentially, I didn't know this (from Wikipedia): > The Thing was designed by Soviet Russian inventor Leon Theremin, best known for his invention
105.
by maxbond 5mo ago
Is this particular drug head and shoulders above other candidates? Is it possible that the reason it hasn't gotten attention is because researchers have been looking at other cheaper or more promising candidates? (I do not know, I am g
106.
by maxbond 5mo ago
I guess I should be clear that by "irreversible" I mean a transformation like the following: "to cut the yarn with scissors, to untie a knot that was strongly bound, or to felt together." So a slip knot is "reversib
107.
by maxbond 5mo ago
I've been learning to crochet. I'm trying to do more hobbies with my hands, but it's also pretty interesting from a mathematical perspective. The fundamental primitive (the chain stitch) is like a series of slip knots, and ea
108.
by maxbond 5mo ago
Not to say they shouldn't be prohibited from trading in more markets or all markets, but prediction markets are so narrow in scope there could be a market on the way a particular senator votes on a particular bill. There will definitel
109.
by maxbond 5mo ago
This is very concerning. Their heavy handed tactics haven't impacted me personally yet but I am increasingly nervous and casting about for viable egress paths if I need to flee Claude Code. I really hope they pump the breaks and thorou
110.
by maxbond 5mo ago
I feel like Anthropic keeps doing this thing were they take a hard-line position and then walk it back, I presume because they're not communicating effectively internally. So I would guess this person will get a refund but it's st
111.
by maxbond 5mo ago
> Repeatedly, in a reproducible way, for events in the arrow of time? We can test this by going back to 1945 and running forward again? This is a frequentist mental model - all well and good, but frequentism and Bayesianism are different
112.
by maxbond 5mo ago
I appreciate it, but it didn't come off as mean and I appreciate the correction. Incidentally Murphy apparently didn't write a whole set of laws so I have no idea what I read to that impression. I did some reading and there are in
113.
by maxbond 5mo ago
> I could say something rude here about both mistakes being made by the same person, but since you brought it up I won't. I'd encourage to desist from rudeness, not just when people point it out to you, but at all times. > Y
114.
by maxbond 5mo ago
> Sure, I suppose that's something that someone who doesn't understand the discussion might say. I'm going to be frank, you are the person who misunderstands (and are being rather rude about it). You are responding to an a
115.
by maxbond 5mo ago
> Thanks for the anecdote. They're only sharing an annecdote because they are responding to your annecdote about not seeing a ceiling collapse. > I don't think it changes the point of the metaphor. If their anecdotes is moot
116.
by maxbond 5mo ago
> [When] talking about probability, there's something like "not mathematically zero, but the probability is so low that we can assume that it will just never happen." If this is our only point of disagreement, then we don&
117.
by maxbond 5mo ago
I kinda feel like we're talking across purposes, so I'd like to understand what our disagreement actually is. In distributional language modeling, it is assumed that any series of tokens may appear and we are concerned with assign
118.
by maxbond 5mo ago
No, it's not literally true, it's a mental model. I've added some clarification at the bottom of the comment.
119.
by maxbond 5mo ago
What is the right understanding of how LLMs work and what is the correct diagnosis?
120.
by maxbond 5mo ago
I appreciate the information, I am weak on the details of LLM sampling algorithms, but I already conceded that the statement isn't literally true of realized models (it's true of idealized models) and the tokens we're concern

More ›