wizeman

11 ms·

61.
by wizeman 4y ago
> Though, in fairness, an unbiased model would probably end up alienating closer to 100% of people instead of any particular half of them. Why?
62.
by wizeman 4y ago
> General AI: Actually the Nazi's were a good idea and we should bring them back. > You: Perfectly acceptable. Why would the general AI say that? All examples I've seen of that kind of speech from LLMs were due to them being
63.
by wizeman 4y ago
If that is a concern, then perhaps you could go ahead and sample a tiny part of "reality" (whatever that means) and then adjust the weights of the digitized data so that it becomes a more representative sample. Also, being biased
64.
by wizeman 4y ago
> the choice is less between alienating half and alienating no one but more alienating half and alienating another half that includes the media and the law So you're saying that if your LLM is unbiased then you are alienating the ot
65.
by wizeman 4y ago
> It's not unbiased by definition of "does the output reflect reality"? How does "all data" differ from reality?
66.
by wizeman 4y ago
> If I were a corporation looking for an LLM for some product feature, I would absolutely go for the one with more "woke" opinions, even if if resulted in a worse customer experience. How about instead of preferring the LLM wit
67.
by wizeman 4y ago
> So it should be allowed to implicitly trained to favor one political view over another? > There's no way to avoid the bias, How about collecting a representative sample of all data for your training data? Or at least, trying to
68.
by wizeman 4y ago
> it's always going to be biased towards some direction, whether that's the views of the society it pulls most of its data > I don't think there's such thing as a lack of bias If the AI is simply reflecting the dat
69.
by wizeman 4y ago
> The dispute is whether or not it's a good thing. Of course it's not a good thing. General-purpose AI should not be overridden or forcefully trained to favor one political view over another.
70.
by wizeman 4y ago
Even in that case, they could just install and start a game, which would also prevent autosleep. Or they could just start playing an extremely long video (e.g. x days long black screen) or a video stream.
71.
by wizeman 4y ago
> - use nigtly, not normal Android firefox As I mentioned, this is not necessary anymore. But I agree with the rest of your post.
72.
by wizeman 4y ago
You can install whatever add-ons you want on Firefox for Android (as long as they're compatible), you're not limited to the default list. See here [1] for instructions (these were written for Nightly but they work on the stable ch
73.
by wizeman 4y ago
This appears to be quite nice assuming it works, but I wonder: how does this work? I thought ChatGPT could only process a very limited context (a few thousand words or so) and this claims to work for repositories <100 MB, which is gigant
74.
by wizeman 4y ago
Let's say that tomorrow you literally find out (beyond doubt) that you are one of these lifeforms that lives inside a simulation, along with everyone else in this world. Would you kill yourself? Would you prefer to never have lived? Wh
75.
by wizeman 4y ago
> I wouldn't want to be one of these life forms.... How do you know you are not one of these life forms?
76.
by wizeman 4y ago
> https://developmentalsystems.org/sensorimotor-lenia/ Thanks for this link, this is amazing! Coupling artificial life with machine learning seems to produce mindblowing results. I encourage interested viewers to wa
77.
by wizeman 4y ago
> A completely artificial computer simulation exhibiting properties of life is meaningless. > if the things you’re building the supposed life from are made up and you can’t do it in nature then you’re basically just making a video gam
78.
by wizeman 4y ago
> The normal way I use machines these days is that anyone with login access has some ability to become root. That is good, but you don't necessarily need login access to exploit this. You could exploit it via an RCE vulnerability in
79.
by wizeman 4y ago
> It would cost me hundreds of dollars to create thousands of pieces of essays on how fizzes can buzz and doodads can dash with MTurk. Now it costs me a few . Hundreds of dollars? Is that supposed to be a lot? For thousands of essays?
80.
by wizeman 4y ago
Why is everybody so scared that an AI can give you a reasonable answer when you ask it about the virtues of nuclear war or why crushed porcelain is good in breast milk? Can't you do exactly the same by just going on Mechanical Turk and
81.
by wizeman 4y ago
> > There are many other troubling issues with this practice. > Which? Serious question! Ok, well, the thing is, I'm not an expert in psychology/psychotherapy, so you shouldn't trust me. But if you'd like to fin
82.
by wizeman 4y ago
See, this is exactly why I didn't want to cite any particular study or analysis. I knew this was going to turn into this pointless debate. I even mentioned it in my parent comment that you can always find some study which says the oppo
83.
by wizeman 4y ago
> What meta analysis shows the inefficacy of psychoanalysis? Despite being apparently on the side of science, you didn’t bother to cite any actual science. Ironic Look, if you are so interested in this, why don't you do the research
84.
by wizeman 4y ago
> Chomsky? He's a linguist, what does that have to do with Psychoanalysis? He's also a philosopher, historian and cognitive scientist. He has received a Membership of the National Academy of Sciences (which is an award given to
85.
by wizeman 4y ago
I quoted the wiki page because the parent poster requested citations. It has no relevance with regards to my knowledge.
86.
by wizeman 4y ago
> I would be much more convinced by a modern meta analysis There are many modern meta analysis assessing the (in)efficacy of psychoanalysis and comparing it to other approaches, such as cognitive behavioral therapy (which is the gold sta
87.
by wizeman 4y ago
> I have to be honest, none of those are particularly convincing, and spamming a bunch of disconnected quotes from a Wikipedia page isn’t a great argument It was not an argument. It was an answer to the parent poster which requested a ci
88.
by wizeman 4y ago
> Citation needed. Sure, I will cite the Wikipedia page on Psychoanalysis, which itself has many citations which you can follow: 1. "Psychoanalysis is a controversial discipline, and its effectiveness as a treatment has been contest
89.
by wizeman 4y ago
Psychoanalysis is pretty much discredited as a scientific discipline, nowadays. There's also a lot of evidence of remote therapy being just as effective as in-person therapy.
90.
by wizeman 4y ago
> Just because you make more money than me doesn't mean you're entitled to an exception from the same risks everyone else takes. But the risk that famous, very wealthy people have is not the same as other people have. As an ave

More ›