retsibsi

10 ms·

91.
by retsibsi 7mo ago
Maybe I'm just being too literal, but I don't know if you're really disagreeing with me. I was disputing "the response they give to this kind of question is completely meaningless". An answer from its weights is out
92.
by retsibsi 7mo ago
I don't think it's intended as that kind of binary. It's more like "yeah, it's flawed in that way, and here's how you can get around that". If someone's claiming the tool is perfect, they're wron
93.
by retsibsi 7mo ago
> completely meaningless This is way too strong isn't it? If the user naively assumes Claude is introspecting and will surely be right, then yeah, they're making a mistake. But Claude could get this right, for the same reasons
94.
by retsibsi 7mo ago
In the original studies, most people made an error that can't be explained by that misunderstanding: they failed to select the card showing 'not y'.
95.
by retsibsi 7mo ago
Quoting the Wikipedia article's formulation of the task for clarity: > You are shown a set of four cards placed on a table, each of which has a number on one side and a color on the other. The visible faces of the cards show 3, 8, b
96.
by retsibsi 7mo ago
It's very easy to adopt a posture of above-it-all cynicism, and to think that anyone who sees an important distinction between two flawed powerful people is a sucker. But it's not particularly smart or sophisticated, and it's
97.
by retsibsi 7mo ago
This seems obvious in retrospect! I've often wondered why Chrome's bookmarks (and, to a lesser extent, history) system is so bad, even to the point of thinking it was a bit suspicious, but I didn't put 2 and 2 together and re
98.
by retsibsi 7mo ago
(I'm not a lawyer, but) I don't see the ambiguity. It's a normal grammatical sentence if parsed this way: The Department of War may use the AI System for all lawful purposes, consistent with - applicable law - operational req
99.
by retsibsi 7mo ago
You're still dancing around the point. Person A said X; person B said not X; we have no concrete evidence either way. Person A is an anonymous representative of a group that has no norms against dishonesty, an obvious motive to falsely
100.
by retsibsi 7mo ago
At this point I think they're targeting two groups: people who aren't paying much attention to this but may see the occasional headline or tweet or soundbite; and people (such as OpenAI employees, and users who might feel compelle
101.
by retsibsi 7mo ago
That's a copout and you know it. You're focusing on the 'unnamed' part; I'm focusing on the 'representative of an administration that lies constantly and brazenly' part.
102.
by retsibsi 7mo ago
I think it's dumber than that; the terms of the contract, as posted by OpenAI ( https://openai.com/index/our-agreement-with-the-department-o... ), are basically just "all lawful purposes" plus some extra w
103.
by retsibsi 7mo ago
Am I wrong to think that such an agreement is basically meaningless? OpenAI gets to say there are limits, the government gets to do whatever it wants, and OpenAI will be very happy not to know about it.
104.
by retsibsi 7mo ago
> So yes, if the supply chain risk is that anthropic might suddenly pull out of military projects and leave people depending on them high and dry, this seems like an appropriate response. But it is so much broader than that! He's sa
105.
by retsibsi 7mo ago
> My understanding is that the OpenAI deal disallows domestic mass surveillance and autonomous weapons, In that case, what on earth just happened? The government was so intent on amending the Anthropic deal to allow 'all lawful use&
106.
by retsibsi 7mo ago
Is there any reason at all to believe the account of the unnamed "defence official"? Whatever your position on this administration, you know that it lies like the rest of us breathe. With a denial from the other side and a lack of
107.
by retsibsi 7mo ago
> I think anthropic wants their cake and to eat it too. You can't take a principled stand against something and then be shocked the thing you are taking a principled stand against might think you are a risk. Is it a principled stand
108.
by retsibsi 7mo ago
> If the US buys from a defense contractor who bought from abthropic, that really isn't that different from buying direct. The moral hazard is still there and the risk that anthropic will try to prevent their product from being used
109.
by retsibsi 7mo ago
That makes sense, but then what is the purpose of the 'foe' label? I can see the logic behind using it as a time-saver (as described by conesus) or a reminder that engagement will probably be unproductive. But if you intend to lea
110.
by retsibsi 7mo ago
> And it's in those situations that our true morality, as measured against our basic needs, comes out. This is far too binary IMO. Yeah, the higher the personal stakes the bigger the test, and it's easy for someone to play the
111.
by retsibsi 8mo ago
This is downvoted, perhaps as a 'lazy dismissal'? But I read the SciAm article and I don't think it actually explained this point. The finding seems to be that the bouba-kiki effect is not specific to humans and does not depe
112.
by retsibsi 8mo ago
Phrases like "actual understanding", "true intelligence" etc. are not conducive to productive discussion unless you take the trouble to define what you mean by them (which ~nobody ever does). They're highly ambiguou
113.
by retsibsi 8mo ago
That was always kind of a cruel attitude, because real people's emotions were at stake. (I'm not accusing you personally of malice, obviously, but the distinction you're drawing was often used to justify genuinely nasty troll
114.
by retsibsi 8mo ago
> I'm curious about what other conclusion you may have reached when reading "on Meta's platforms". The claim was "79% of all child sex trafficking in 2020 occurred on Meta’s platforms", so they probably took
115.
by retsibsi 8mo ago
I wonder if they meant this HN post (the OP), rather than https://asteroidos.org/news/2-0-release/ ? The latter doesn't seem LLM-written to me, but the HN post does.
116.
by retsibsi 8mo ago
I think the AI-coding skill that is likely to remain useful is the ability (and discipline) to review and genuinely understand the code produced by the AI before committing it. I don't have that skill; I find that if I'm using AI,
117.
by retsibsi 8mo ago
I've just realised I messed up the search, and the algolia link is to my pre-2020 comments containing the word 'author'. But my full (far longer) list of pre-2020 comments also shows some pretty heavy double-hyphen use: 6 hit
118.
by retsibsi 8mo ago
I'm obviously just annoying you, which really wasn't my goal, so I'll stop here. But I want to note that if you think this all comes down to "formatting", you're still not hearing what I'm trying to say.
119.
by retsibsi 8mo ago
> the fact that you assume it takes more time reading that comment than it took for me to write is pretty weird I didn't do that either! I had no idea whether you just fired off a quick prompt and pasted the result without even read
120.
by retsibsi 8mo ago
I didn't say or think that your account was AI-run, and I didn't tell you to leave the platform. I just tried to explain why your comment might have annoyed people and triggered negative responses (while agreeing that the rude one

More ›