Filligree

7 ms·

121.
by Filligree 6mo ago
Not to mention inefficient in memory use. I would have expected a mention of interning; using string-views is fine, but making it a view of 4kB cache pages is not really. Though I believe the “naive” streaming read could very well be superi
122.
by Filligree 7mo ago
Except in this case the code wasn't obfuscated, right?
123.
by Filligree 7mo ago
In this scenario, the alternative is “you die”. Let’s make sure we’re including that in the question.
124.
by Filligree 7mo ago
I want whatever causes fewer deaths and injuries total.
125.
by Filligree 7mo ago
It is condescending when antis say AI users do it. It isn’t when a director or team leader does it. But it’s the same process, which should tell you what’s really going on here. It’s about status, not functionality, and you don’t gain s
126.
by Filligree 7mo ago
> because 'I don't know' is not in the vocabulary of any AI. That is clearly false. I’m only familiar with Opus, but it quite regularly tells me that, and/or decides it needs to do research before answering. If I inst
127.
by Filligree 7mo ago
Yes.
128.
by Filligree 7mo ago
Everyone who's used Opus knows it's better than the others in a way that isn't captured by the benchmarks. I would describe it as taste. Lots of models get really close on benchmarks, but benchmarks only tell us how good they
129.
by Filligree 7mo ago
They serve it about 2x slower. So it must have about 2x the active parameters . It could still be 10x larger overall, though that would not make it 10x more expensive.
130.
by Filligree 7mo ago
> Can't we have a system that is optimized for the notes that are actually played in a song rather than the hypothetical set? And what if the optimization is done per note rather than over an entire song? You can. It’s called adapti
131.
by Filligree 7mo ago
Yes; the works are original. However, describing the path you need to get there requires copyright infringement.
132.
by Filligree 7mo ago
I didn’t mean “possible for LLMs”; this is clearly an open question. In fact, I didn’t even mean “possible for a neural network the size of an LLM”. I just meant “possible”.
133.
by Filligree 7mo ago
Less than most humans, but more than many humans.
134.
by Filligree 7mo ago
It’s an extremely difficult problem, and if you know how to do that you could be a billionaire. It’s not impossible, obviously—humans do it—but it’s not yet certain that it’s possible with an LLM-sized architecture.
135.
by Filligree 7mo ago
To make them sound more accurate.
136.
by Filligree 7mo ago
If you can't pedal at all, then an e-bike definitionally won't help you. An e-moped might be suitable.
137.
by Filligree 7mo ago
So the question is, why do we tokenise it in such a way that it makes everything harder?
138.
by Filligree 7mo ago
Maybe just don’t make “jokes” like that.
139.
by Filligree 7mo ago
> What I don’t understand is why the two parties couldn’t reach agreement. Surely autonomous murderous robots is something U.S. government has interest in preventing. Consider the government. It’s Hegseth making this decision, and he con
140.
by Filligree 7mo ago
What? Why do they care whether people shovel their driveway?
141.
by Filligree 8mo ago
Voice control is the claim. In my experience the voice control is entirely unusable, and can’t be fully turned off.
142.
by Filligree 8mo ago
Yes, you don’t write tests based on the implementation. You tell it to write tests based on the specification, which must exist, because it cannot read your mind.
143.
by Filligree 8mo ago
The answer can be “both”. They won’t get this specific question wrong again; but also they generalise, once they have sufficient examples. Patching out a single failure doesn’t do it. Patch out ten equivalent ones, and the eleventh doesn’t
144.
by Filligree 8mo ago
It's the one that works.
145.
by Filligree 8mo ago
The same is true for biological humans. The moment the first upload exists, they’ll be justified in wondering if the ones made from meat are truly conscious.
146.
by Filligree 8mo ago
The aperture size is usually described as e.g. f/32, where f is a camera-specific constant. Denominator, not numerator. That's why larger number = smaller aperture.
147.
by Filligree 8mo ago
Were they legally required to?
148.
by Filligree 8mo ago
> Waymo hits a kid? Ban the tech immediately, obviously it needs more work. > Waymo hits a kid? Well if it was a human driver the kid might well have been dead rather than bruised. These can be true at the same time. Waymo is held to
149.
by Filligree 8mo ago
Pathologic is one of the best portrayals of depression I have ever played. Certainly art.
150.
by Filligree 9mo ago
That's backwards. At too low temperatures batteries start to take damage during discharge or (especially) charge, so 0C is the lowest temperature at which you should charge it. 5C would be better. It's a concern mainly for e.g.

More ›