shwaj

7 ms·

61.
by shwaj 9mo ago
There are literally dozens of people in this very HN conversation that are missing him.
62.
by shwaj 9mo ago
I’m not sure whose business model you’re referring to, Cloudflare or Trump/Vance? Or sounds like the former, but I’m not sure how that appeal “validates the view…”.
63.
by shwaj 9mo ago
That’s what he’s famous for, huh? Nobody knew who he was until he burst onto the national stage because of his brain worm. And please show a source that he was “proud” of the affliction.
64.
by shwaj 9mo ago
It seems like a cheaper model could be asked to review transcripts, something like: “does this transcript seem at all like a wacky conspiracy theory that is encouraged in the use by the LLM”? In this case, it would have been easily detected
65.
by shwaj 9mo ago
Matter to who? If you want to deeply understand how this technology works, this is still relevant. If you want to vibe code, maybe not.
66.
by shwaj 9mo ago
If they set the deadline in 10 years, there would be (smaller) mountains generated in that period anyway.
67.
by shwaj 10mo ago
Could always run the posts through a LLM to decide which are about AI :-p
68.
by shwaj 10mo ago
I’ve heard many European non-native English speakers say “a software”, and I see it as a charming reflection of their native languages, where presumably “software” is a countable noun. I think that’s what’s going on here.
69.
by shwaj 10mo ago
That was my concern too. However, the provided links to both the ToS and privacy policy were the standard Google ones ( https://policies.google.com/terms ), so it seems not to be giving Opal special privileges to read/t
70.
by shwaj 10mo ago
I tried de-selecting that, and it told me to re-login and enable that setting.
71.
by shwaj 10mo ago
Just like those? Those examples are highly improbable, whereas the CIA is almost certainly active in Iran right now (that’s their job!), even if there isn’t evidence that they are behind these particular protests.
72.
by shwaj 10mo ago
Why not both? No need to be so disparaging.
73.
by shwaj 10mo ago
That was my original thought, however if you want to quantify an effect it would be ideal to measure the same trait in both parent and offspring. I assume that the reason that this was done (I didn’t read the papers) is because this is a r
74.
by shwaj 10mo ago
Very apt analogy.
75.
by shwaj 10mo ago
Fair enough. It would be a clearer result (more “apples to apples”) if the same trait was directly measured “on both sides”. Thanks for restating. Edit: typo
76.
by shwaj 10mo ago
In the “paternal exercise” case the trait isn’t the habit of exercising, it’s the metabolic changes of exercise that are (apparently) conferred to both father and offspring.
77.
by shwaj 10mo ago
I know, so what? Very smart people might have more than one idea in their careers. He left to pursue one of those.
78.
by shwaj 10mo ago
Inference being profitable doesn’t mean that they’re selling enough inference to offset their other costs.
79.
by shwaj 10mo ago
Groq is very different from TPUs. The difference in memory should be a big clue. (Groq using SRAM built into each compute unit, vs TPUs using HBM more similarly to GPUs)
80.
by shwaj 10mo ago
Rather than bemoaning the ignorance of those who can’t grasp that the text still exists under black rectangles, we should thank our lucky stars that this method of deredacting still works, well over a decade after the first time I heard of
81.
by shwaj 10mo ago
We don’t know that, AFAICT (after reading the Ars article and the Bluesky post). Some of the puzzles are probably reused, and other new ones using the same mechanics may have been written. I’m not sure why you’d so confidently state “Blow
82.
by shwaj 10mo ago
Agreed. The head of the fly also seems to have weird depth.
83.
by shwaj 10mo ago
Why? Wash trading is about selling and then requiring the same asset for tax purposes. How is this analogous, other than that you presumably dislike both practices?
84.
by shwaj 10mo ago
It’s “Fuchsia” with a “chs” not a “sch”. Where do you get your information that it’s dead?
85.
by shwaj 10mo ago
“it”
86.
by shwaj 11mo ago
Well said, except for the last sentence: Just because everyone does it doesn’t mean one isn’t a sinner for doing it.
87.
by shwaj 11mo ago
The bet, (I would have thought) obviously, is that AI will be a huge part of humanity’s future, and that Anthropic will be able to get a big piece of that pie. This is (I would have thought) obviously different from selling dollars for $0.5
88.
by shwaj 11mo ago
I use VSCode on a Mac, using remote SSH to edit code on a remote Linux machine. So I voted twice I’ve never been able to get used to default Linux key bindings, and never been able to customize them to feel quite right.
89.
by shwaj 11mo ago
It’s an inadvertent step toward Newspeak, where we no longer have a word that means what “literally” used to unambiguously mean.
90.
by shwaj 11mo ago
Another glaring example: > Set theorists use the language of logic, computer scientists the language of algorithms. Computer science doesn’t use logic? Hello, Booleans. So lazy, especially when you can ask an AI to tell you if you’re say

More ›