Tenoke

12 ms·

151.
by Tenoke 3y ago
The average IQ in the training data isn't 100. Higher IQ individuals likely write more online, higher quality datasets like Wikipedia are oversampled etc.
152.
by Tenoke 3y ago
Sure, but again if you run it yourself you can use the finetuned by users checkpoints that have it.
153.
by Tenoke 3y ago
The nice thing about Stable Diffusion is that you can very easily set it up on a machine you control without any 'safety' and with a user-finetuned checkpoint.
154.
by Tenoke 3y ago
I haven't done it recently but I prefer to go to dreddit for finding the currently respected markets and links. dark.fail has been comrpomised in the past https://www.reddit.com/r/onions/comments/n1byhj&#
155.
by Tenoke 3y ago
Okay but if that's the case we are no more than a decade away from integrating those into a newer and bigger model.
156.
by Tenoke 3y ago
Quest 3 is $500, not $1000 and needs no external PC. Your choice of game is equally bizzare.
157.
by Tenoke 3y ago
How is it a defeaning silence? Their latest hardware product was released ~3 months ago.
158.
by Tenoke 3y ago
This isnt the board though. This is the against the board side.
159.
by Tenoke 3y ago
Sam Altman starts his essay with "Development of superhuman machine intelligence (SMI) [1] is probably the greatest threat to the continued existence of humanity. " Yan LeCunn explicitely says "superhuman AI is not an existen
160.
by Tenoke 3y ago
Yann LeCunn is the most outspoken anti-AI Safety guy in the field, he has practically built his modern brand on it. In his eyes AGI is at best very far, and even if it isn't, AGI Safety isn't a real concern. He posts about it mult
161.
by Tenoke 3y ago
I don't think he is a Yann LeCunn but I still think he (and his vision for OpenAI) is more reckless than many people in the field.
162.
by Tenoke 3y ago
What's erratic here? If they are likely to lose most of their employees and resources then merging with the most aligned other company is not only reasonable but exactly what it says they'll do in their charter.
163.
by Tenoke 3y ago
Except so far he has 1. Tried to reverse the decision and come back 2. Position the board into being publically pressured and thus accountable 3. Tried to have the board replaced 4. Take near everyone and everything from OpenAI where he is
164.
by Tenoke 3y ago
'Let’s say we really do create AGI. Right now, if I’m doing a bad job with that, our board of directors, which is also not publicly accountable, can fire me and say, “Let’s try somebody else”.' - Sam Altman, 2020 0. https:/&
165.
by Tenoke 3y ago
I'm sure the return on that investment was very large in terms of donations.
166.
by Tenoke 3y ago
I find it really good but if you don't like rationality/EY it's really easy to latch on as something to hate (overly smart fanfic does sound cringe on the face of it).
167.
by Tenoke 3y ago
Don't they have a more limited license to use the IP rather than full rights? (The stratechery post links to a paywalled wsj article for the claim so I couldn't confirm)
168.
by Tenoke 3y ago
This is an extremely uncharitable take based on pure speculation. >Ilya cares nothing about humanity or security of OpenAI, he lost his mind when Sam got all the spotlights and making all the good calls. ??? I personally suspect Ilya tri
169.
by Tenoke 3y ago
At this point either pretty much all the speculation here and on Twitter was wrong, or they've threatened to kneecap him.
170.
by Tenoke 3y ago
They'll definitely both get funded.
171.
by Tenoke 3y ago
The problem is Elon's approach to alignment as presented during x.ai launch is pretty different to what Ilya says and as far as I can tell pretty naive on top of that.
172.
by Tenoke 3y ago
In all fairness if there is an active post it will be at the top and that's the main value of a status page, not so much about the right order of years-old outages.
173.
by Tenoke 3y ago
>I wouldn't say that OpenAI "telegraphed" their intentions They did telegraph it, they showed the multimodal capabilities back in the GPT4 Developer Livestream[0] right before first releasing it. 0. https://yout
174.
by Tenoke 3y ago
Polling an ensemble of experts is much better than asking a single one, and there are some that vote 'not sure' or similar in those polls so it's hardly a coin flip.
175.
by Tenoke 3y ago
To be fair most people can't visit countless cities weekly, can't visit or don't care for the nature you like, and the increased injury risk of many sports (compared to just cycling for cardio and doing resistance training) i
176.
by Tenoke 3y ago
The saltiness is much, much less than in some games (shooters, mobas) but a portion of every playerbase thinks they are the worst one even if they are not.
177.
by Tenoke 3y ago
>but that it actually pays an okay/ish salary to work on fulltime, while upholding all those lofty goals. He pays himself a fairly low salary for the work he is putting in.
178.
by Tenoke 3y ago
A lot who stay purely in Academia do it as well, presumably for the prestige.
179.
by Tenoke 3y ago
In 2020 I did my Anmeldung by just sending them an email and not going in-person. Too bad they didn't keep the Covid workaround even after setting it up.
180.
by Tenoke 3y ago
> German apartments don’t have apartment numbers They do (at least sometimes?). They are just not listed anywhere and barely used. I have my number in my contract, and e.g. Vatenfall have it to connect the meter number to the flat number

More ›