daveguy

48 ms·

961.
by daveguy 2y ago
"Talking politics is probably safe as long as you are not too loud." and "Talking about politics is always taboo." Hahahaha. That's not the airtight argument against "a subtle way of saying keep your mouth shut
962.
by daveguy 2y ago
Trump is for whoever brown noses the most. He has no independent preferences. Big tech gave him millions, now he's for big tech, but only the ones who donated. Most of them figured out that Trump is for sale to every bidder and got in
963.
by daveguy 2y ago
Doge couldn't even secure their "transparency tweets" website when it went up. I get the impression they don't have the organization to confirm anything in writing.
964.
by daveguy 2y ago
I'm wondering how someone knows the size of the field of evolutionary psychology and what it should be instead of what it is so well that they feel no evidence is necessary. Are you in the field yourself? Or do you have questions about
965.
by daveguy 2y ago
Deeming things as waste within days of gaining access to the info is 100% in bad faith. There is no possible way that musk and his minions took the time to find out why anything is the way it is. Nevermind the fact that you don't have
966.
by daveguy 2y ago
Yeah, don't be so hard on Musk and Trump. They're destroying oversight agency by agency, not all at once. Starting with the agency in charge of HR oversight.
967.
by daveguy 2y ago
Not if we speak the fuck up.
968.
by daveguy 2y ago
LLMs are the "move fast and break things" of AI.
969.
by daveguy 2y ago
Yeah, you can build a rig to run full models for 10-20k... That's a big reason OpenAI might not make it. The whole article is about LLMs not being a viable business.
970.
by daveguy 2y ago
Subjective evaluation of "humanity" and "empathy" in responses is much less important than clinical outcome. I don't think an online chat with a nebulous entity will ever be as beneficial as interactions that can,
971.
by daveguy 2y ago
I bet Google is utilizing the value of the LLM input prompts with close to the same efficiency they are monetizing search. I that case, there are two questions -- 1) will LLM overtake search? and 2) can anyone beat Google at monetizing thes
972.
by daveguy 2y ago
That would be a killer for the current and near future generations of LLM as a business. If they are having to pay many times in compute what they are able to get for the API use (due to open models being near comparable?), then you definit
973.
by daveguy 2y ago
Not sure if it matters at this point. There will need to be many more rounds of CapEx to realize the promises that have been put forth about these models.
974.
by daveguy 2y ago
Congrats on the rapid MVP! Is it launched / publicly available? I'm in the same boat with modern front ends. Would like to see what the current LLMs can help with.
975.
by daveguy 2y ago
Trump has never done anything for anyone over doing what's right for himself.
976.
by daveguy 2y ago
This seems like magic, but I can't find a research paper that explains how it works. And "expert-level analysis across a range of complex subject matters." is quite the promise. Does anyone have a link to a research paper tha
977.
by daveguy 2y ago
Hahaha. Thank you for the laugh. It must be great walking around with such an unwavering sense of superiority. Enjoy the dark joy of your perceived vast superiority and try not to have a breakdown when you come crashing back to reality. May
978.
by daveguy 2y ago
Someone who does not know that MRI has variable field strength is someone who truly does not know the basics of MRI. Maybe this source will clue you in as to why your understanding of MRI is fundamentally flawed: https://pmc.ncbi
979.
by daveguy 2y ago
Oh good, they noticed the effect too. It's almost as if many people experience the effect, but somehow to you it's impossible for any humans to be able to take advantage of the effect at lower field strengths. When you walk up to
980.
by daveguy 2y ago
Apparently you have not had an MRI or studied MRI. 1) The people operating the MRI are outside the magnetic field. 2) The patient in the MRI is oriented such that the magnetic field is directed from head to toe (where the effect on sensing
981.
by daveguy 2y ago
Has your LLM ever known just the right next token that should come after 2 million other tokens? Mine hasn't.
982.
by daveguy 2y ago
Apparently they get backdoors in as incompetently as they create efficiency.
983.
by daveguy 2y ago
Well yeah, government efficiency is obviously a "great idea", but that is ridiculously simplistic. No one is paid for "great ideas" without any clue how to execute. Poorly implemented good ideas screw things up just as b
984.
by daveguy 2y ago
Let me break it down for you. In electromagnetic fields both an electric field and a magnetic field are produced. They are orthogonal to one another (in 2D with 2 vectors this is perpendicular, orthogonal is when the vector is in a differen
985.
by daveguy 2y ago
It may be a bit more complicated than what a nameless "competent repair person" thought it should have been. Do you worship your news sources too? https://www.nbcnews.com/news/us-news/reservoir-pacific-p
986.
by daveguy 2y ago
Apparently you fell for the narrative that LA ran out of water because they preferred to save the environment. Just so you know, it didn't matter how much water was in any reservoirs. There was plenty of water. The two failures that ha
987.
by daveguy 2y ago
What do you mean by that? Is that how doge is working? No one in the government has previously been paid per line of code. Am I missing something here?
988.
by daveguy 2y ago
This makes sense to me. When I optimize, the most significant gains I find are algorithmic. Whether it's an extra call, a data structure that needs to be tweaked, or just utilizing a library that operates closer to silicon. I rarely
989.
by daveguy 2y ago
It's definitely both. A bunch of 20 year olds were let loose to be "super efficient." So, to be efficient they use LLMs to implement what should be a major government oversight webpage. Even after the fix the list is a few
990.
by daveguy 2y ago
Same place every LLM gets "information" -- plucking random bits from the latent space that are most likely to follow the input context.

More ›