empath75

21 ms·

541.
by empath75 1y ago
Yes I very much understand that most people do not think that LLMs think or understand like we do, but it is _very difficult_ to prove that that is the case, using any test which does not also exclude a great deal of people. And that is be
542.
by empath75 1y ago
This is an obnoxious response. Of course I recognize that philosophy is the solution to this. What I am pointing out is that philosophy has not as of yet resolved these relatively new problems. The idea that non-human intelligences might
543.
by empath75 1y ago
One thing that LLMs have exposed is how much of a house of cards all of our definitions of "human mind"-adjacent concepts are. We have a single example in all of reality of a being that thinks like we do, and so all of our defini
544.
by empath75 1y ago
It's good at micro, but not macro. I think that will eventually change with smarter engineering around it, larger context windows, etc. Never underestimate how much code that engineers will write to avoid writing code.
545.
by empath75 1y ago
I actually find it very stressful and frustrating to watch because they answer faster than I can reason through stuff.
546.
by empath75 1y ago
Models will engage in post-hoc rationalizations, so you can't trust their purported reasoning -- in particular, if you sneak an answer into the context (even an incorrect answer), it will provide reasoning for giving you that as the fi
547.
by empath75 1y ago
ChatGPT 5 did argue with me about something math related I was asking about, and I did realize I was wrong after considering it further. I don't actually think being told that I have asked a stupid question is valuable. One of the pri
548.
by empath75 1y ago
A good way to determine if your argument is a good one on this topic is to replace every instance of an LLM with a human and seeing if it is still a good test for whatever you think you are testing. Because a great many humans are terrible
549.
by empath75 1y ago
There are many reasons why someone may ask a question, and I would argue that "getting the correct answer" is not in the top 5 motivations for many people for very many questions. An empathetic answerer would intuit that and may g
550.
by empath75 1y ago
you can interrupt it while it's talking.
551.
by empath75 1y ago
> The whole game of Rationalism is that you should ignore gut intuitions and cultural norms that you can't justify with rational arguments. The game as it is _actually_ played is that you use rationalist arguments to justify your pr
552.
by empath75 1y ago
Most "rationalists" throughout history have been very deeply religious people. Secular enlightenment-era rationalism is not the only direction you can go with it. It depends very much, as others have said, on what your axioms ar
553.
by empath75 1y ago
I do a ridiculous amount of dumb shit with cursor without paying extra for it and I can't be the only one. There is just no way they will ever make a profit with their current all-you-can-eat model. Like there have been times I have
554.
by empath75 1y ago
When I worked at AOL about 10 years ago, dialup was _still_ responsible for 100% of their profits. Literally every other part of the company lost money (Mapquest, AIM, Huffington Post, their ad network, etc). They were making literally bi
555.
by empath75 1y ago
I just setup a short cut "Hey siri, let's chat" that opens chatgpt voice.
556.
by empath75 1y ago
The auth story for MCPs is a complete mess right now, though, which is why people make ones to run locally.
557.
by empath75 1y ago
I wonder how the post mortem on this was handled internally, because if the lesson learned was "we need to hire people that can do basic math", rather than "We should probably use '.2 cents per kb' in our documentat
558.
by empath75 1y ago
It's hard to understand his syntax, but I think he's arguing for something similar to the way that Rust uses From and Into.
559.
by empath75 1y ago
(Sorry the original video title was very clickbaity)
560.
Verizon doesn't know the difference between dollars and cents (2006) [video](youtube.com)
3 points by empath75 1y ago|2 comments
561.
by empath75 1y ago
That isn't the font, you can look at it yourself, it doesn't match.
562.
by empath75 1y ago
Open AI has 10 times the active users of any of it's competitors, including google.
563.
by empath75 1y ago
https://chatgpt.com/share/688cd9bd-2dc0-8000-936a-0bbf7ba442... Compare to what 4o does.
564.
by empath75 1y ago
Think about how much money openai could make _just_ by getting referrals for agentic work flows (ie: help me plan my vacation). The thing that people are missing is that openai is a platform like google's, and there are a million diffe
565.
by empath75 1y ago
It can also get oomkilled
566.
by empath75 1y ago
Chrome has thousands of developers working it. That is a lot of man hours for small optimizations.
567.
by empath75 1y ago
Rayon is about as pure an example of it as you can imagine. In a lot of cases you just need to replace iter() with par_iter() and it just works.
568.
by empath75 1y ago
> Congrats, rustc forced you to wrap all your types in Arc<Mutex<_>>, and you no longer have data races. Or you can just avoid shared mutable state, or use channels, or many of the other patterns for avoiding data races in Ru
569.
by empath75 1y ago
One can talk about LLMs in contexts that aren't about engineering, and are instead about topics like: "Do LLMs think" or "Are LLMs intelligent". People _frequently_ point to some failure mode of LLMs as dispositive
570.
by empath75 1y ago
I am very much of the opinion that Hamas should not be allowed to continued to exist for Israel's benefit and for Palestine's, but there is a lot of space in between "just stop fighting" and "genocide", and Isr

More ›