lubesGordi

6 ms·

31.
by lubesGordi 2y ago
What's their narrative and what's the reality? Sounds like you got something to say!
32.
by lubesGordi 2y ago
Kind of shocking to see so much angst against memorization here. Memory has been long thought to be a critical component of intelligence, with elaborate mnemonics systems developed by people to help memorize more things (see Francis Yates&#
33.
by lubesGordi 2y ago
Its such a testament to how indoctrinated and homogeneous that environment is.
34.
by lubesGordi 2y ago
Agreed. Being a whistleblower is probably very stressful.
35.
by lubesGordi 2y ago
I think this is the point of the Attention portion in an llm, to use context to skew the embedding result closer to what youre looking for. It does seem a little strange 'ruler' would be closer to 'king' versus something
36.
by lubesGordi 2y ago
Maybe 'new information' can be generated by taking existing data and finding new relationships between that data. If that were the case, the original embedding for that data would change (new relationships = new associations). I
37.
by lubesGordi 2y ago
Hey can you guys elaborate how this works? I'm looking at the Ollama section in their docs and it talks about load balancing? I don't understand what that means in this context.
38.
by lubesGordi 3y ago
Quick question, is this just pure C code that can be loaded into an Nvidia gpu and run (via the python code)? I scanned the C and didn't see anything CUDA related (maybe I missed something, I'm not a GPU programmer!). K mentions
39.
by lubesGordi 3y ago
Ah, non-determinism at its finest.
40.
by lubesGordi 3y ago
I am curious what laws require deep packet inspection on in transit data. Is this mostly to stop 'bad' packets before they reach an endpoint?
41.
by lubesGordi 3y ago
Seriously! Can these people not even read what they write or are they so out of touch with reality. Jesus.
42.
by lubesGordi 3y ago
What does it mean to 'treat gravity approximately (that is, perturbatively)'? That sounds like something we do to model, which only approximates reality. That model sounds like it shouldn't be used to predict anything else?
43.
by lubesGordi 3y ago
I was on the edge of my seat in a very novel way for the first half or so of the first book, then it turned fantasy (to me) quick. I know people are claiming this is hard sci-fi, it's not. But there are some very interesting ideas and
44.
by lubesGordi 3y ago
I was just skimming this and I thought there were easy ways of decompiling into C source as well with generated function names like a, b, c or whatever. Could one just dump that sort of minimized C code into an llm and have it generate va
45.
by lubesGordi 3y ago
I'll be less nice than you and say I'd be pretty mad if I was building something legit but lost out to a project like this. You not being brazen to phish all your fellow students just means you're not an asshole (and I thank
46.
by lubesGordi 3y ago
I must've missed the hacking part. All I saw was a phishing expedition that resulted in them being able to log in as other users (and scrape their data)?
47.
by lubesGordi 3y ago
I'm not an ML guy but does it actually matter that OpenAI doesn't disclose it's source? Who cares? They proved the transformer arch and everyone knows how to replicate what they did. That's pretty open. The sauce is
48.
by lubesGordi 3y ago
I think the deal is that these breakthroughs, aren't really great in any general case. It helps with specific instances of work, and makes individuals way more productive. That by itself might won't end up making some people obs
49.
by lubesGordi 3y ago
We need something like this but with emergent phenotypes instead of hard-coded. How would that work?
50.
by lubesGordi 3y ago
This is the problem across the board, regardless of service almost. Everything is some crappy recommendation system, instead of enabling me to find what I want. In addition to sucking in general, recommendation systems have this great prop
51.
by lubesGordi 3y ago
Thanks for the udl book (Understanding Deep Learning), that looks like a really great starting point.
52.
by lubesGordi 3y ago
Its like a securities fraud race condition. I wonder how many of those there are in the legal system.
53.
by lubesGordi 3y ago
That's really insightful into how this guy works. He's judging everything off probabilities to an egregious degree. He's also utilitarian, which doesn't really offer a foundation for any kind of morality because there&
54.
by lubesGordi 3y ago
Sentencing isn't until March, and it's pretty unlikely he'll get a life sentence. Could be decades though.
55.
by lubesGordi 3y ago
Sequoia invested $200M in this guy. Has anyone looked at whether SBF could've done as much damage without that investment? Has anyone looked at if Sequoia did their due diligence? Seems like Sequoia could've seen red flags just
56.
by lubesGordi 3y ago
Does this custom diff driver also help git do a merge of two dbs?
57.
by lubesGordi 3y ago
This seems so obviously useful and surprising it's taken industry so long to get here. But honestly I'm hesitant to go for something that isn't like a wrapper around the underlying database, just because those are notoriousl
58.
by lubesGordi 3y ago
MiraLAX is basically just PEG, and is a laxative prescribed by doctors to people every day.
59.
by lubesGordi 3y ago
I'm surprised to not see anything by Corel in there, though admittedly this isn't my area. Corel was the only real alternative back in AD 2000 when I was playing with photoshop, and I do have their recent lightroom version which I
60.
by lubesGordi 3y ago
The one that bothers me the most is the Ring integration with police. That's happening in Chicago. The neighbor across the street installs a Ring doorbell because they like gadgety shit or whatever, and suddenly my children's pl

More ›