nearbuy

11 ms·

181.
by nearbuy 1y ago
> My stance is it's (obviously and only a fool would think otherwise) never going to be conscious because consciousness is a physical process based on particular material interactions, like everything else we've ever encountere
182.
by nearbuy 1y ago
Thanks for clearing it up. > The point about structural variation not precluding consciousness stands. Maybe, but my point about high-functioning people with hydrocephalus is that they have the same or similar brain structure (in terms o
183.
by nearbuy 1y ago
The first article is about someone missing a cerebellum, not part of their cerebrum. That's the motor and balance part of the brain, and as you might expect, the subject of the article has deficits in motor control and balance. The Bio
184.
by nearbuy 1y ago
Are there any cases like that? I've never heard of someone functioning normally with little or no brain beyond a cerebellum.
185.
by nearbuy 1y ago
The monkeys on typewriters saying is just a colorful way of saying that an infinite random sequence will contain all finite sequences somewhere within it. Which is true. But I don't see what infinite random sequences have to do with LL
186.
by nearbuy 1y ago
> But doesn't this rely on the same thing you suggest we don't have, which is a working and definable definition of consciousness? No, it's like saying we have no indication that humans have psychic powers and can levitate
187.
by nearbuy 1y ago
Can you explain #2? What does the part of the brain that's primarily for balance and motor control tell us about deep learning?
188.
by nearbuy 1y ago
I haven't lied. You're making accusations in bad faith. This was a faithful representation of your position as best as I can tell from your comment. If you'd like to explain why "What you've mistaken for a 'log
189.
by nearbuy 1y ago
My guess is someone didn't fully understand what was expected of them. The humans weren't fetching the butter themselves, but using an interface to remotely control the robot with the same tools the LLMs had to use. They were (I b
190.
by nearbuy 1y ago
"repeated information" was copied verbatim from your comment. Your full sentence was: > What you've mistaken for a "logical model" is simply a large amount of repeated information.
191.
by nearbuy 1y ago
You're not addressing the objection. What is it about your model of how you think LLMs work (that it's just "repeated information") that predicts they'd go haywire when asked about a seahorse emoji (and only the sea
192.
by nearbuy 1y ago
If anything, the seahorse emoji case is exactly the type of thing you wouldn't expect to happen if LLMs just repeated information from their training corpus. It starts producing a weird dialogue that's completely unlike its traini
193.
by nearbuy 1y ago
He's relaying what the textbook he cited claimed. You'll have to check the textbook for their sources. Getting into the weeds about different theories for why various hunter-gatherer societies have lower rates of PTSD is rather of
194.
by nearbuy 1y ago
The article goes through and refutes 25(!) claims from the book. I'm not sure how many more backing arguments you could possibly want from a blog post.
195.
by nearbuy 1y ago
Nineteen Eighty-Four captures the political zeitgeist of when it was written (~1949) and the following years better than it captures that of 1984. This was the era of the cold war and McCarthyism. Describing the present often seems surprisi
196.
by nearbuy 1y ago
Cherry-picking? You made a completely factually wrong statement. There was no cherry-picking. You said the Turing test was never about AGI. You didn't say it has weaknesses. Even if it were the worst test ever made, it was still about
197.
by nearbuy 1y ago
This terminology is confusing. Historically, AI was always used to mean any kind of machine intelligence, including the most basic novice chess AI, or an image classifier, or a video game character's AI. Now a lot of people seem to be
198.
by nearbuy 1y ago
That's pretty much exactly what Alan Turing made the Turing test for. From the Wikipedia entry: > The Turing test, originally called the imitation game by Alan Turing in 1949, is a test of a machine's ability to exhibit intelli
199.
by nearbuy 1y ago
Except his points are mostly straight up factually wrong.
200.
by nearbuy 1y ago
The myth that people in Columbus's time thought the Earth was flat was largely spread by school textbooks in the early to mid 20th century. And those textbooks weren't the originators of the myth; they could cite earlier writings
201.
by nearbuy 1y ago
The landmark paper, "Attention is all you need", that triggered the breakthrough that led to current transformer architecture LLMs, only came out in 2017. Without that breakthrough, they wouldn't exist. And even then, the ear
202.
by nearbuy 1y ago
The smartest, fastest investors are the ones who make a profit by incorporating their information into the stock price in the EMH. The stock price can't move on its own. Under the EMH, someone has to be the first to trade stock based o
203.
by nearbuy 1y ago
Information that requires 12 months to figure out isn't information that's available now. Say you want to know the 400 trillionth digit of pi. We have all the information needed right now to know how to compute it. But you don&#x2
204.
by nearbuy 1y ago
There are two different memory systems for ChatGPT. The older one is the "saved memories" you mention, where it would occasionally jot down some notes when it thought something was worth remembering or when you specifically told i
205.
by nearbuy 1y ago
If you include the encoder outputs as part of the state, then encoder-decoder LLMs are Markovian as well. While in token space, decoder-only LLMs are not Markovian. Anything can be a Markov process depending what state you include. Humans,
206.
by nearbuy 1y ago
> (Also, Wikipedia is a tertiary source, as it is meant to only cite secondary sources, not primary sources.) Wikipedia absolutely cites primary sources (as well as secondary and tertiary sources), and this is in accordance with their po
207.
by nearbuy 1y ago
Not sure why you're telling me this. I was giving an update for the parent comment which was complaining that China's GDP per capita was only $15k and much lower than Japan and Korea. You can tell them why they shouldn't comp
208.
by nearbuy 1y ago
It was $15k in 2017. Now their GDP per capita (PPP) is about $29k. It's growing fast.
209.
by nearbuy 1y ago
They can learn to generalize patterns during training and develop some model of the world. So for example, if you were to train an LLM on chess games, it would likely develop an internal model of the chess board. Then when someone plays che
210.
by nearbuy 1y ago
This isn't true at all. The LLMs absolutely world model and researchers have shown this many times on smaller language models. > techniques like vector tokenization (I assume you're talking about the input embedding.) This is r

More ›