xcv123

15 ms·

121.
by xcv123 2y ago
Go install TikTok and scroll for 30 minutes. It is the experience of your brain dissolving into sludge in real time. https://www.gurwinder.blog/p/tiktok-may-be-a-chinese-bio-wea...
122.
by xcv123 2y ago
The full screen OCR will capture the jpeg file names
123.
by xcv123 2y ago
Allowing users to disable auto restarts
124.
by xcv123 2y ago
Matlab runs on Linux and Mac. He can switch.
125.
by xcv123 2y ago
No. I only accept medical advice from pharmaceutical industry press releases.
126.
by xcv123 2y ago
Same here. Haven't logged in for 20 years but I remembered the number.
127.
by xcv123 2y ago
Same thing happened with my Dell Precision laptop. Fixed it by wiping the hard drive and installing Linux.
128.
by xcv123 2y ago
That's a good monitor but it's relatively low DPI (way too big at 32") and 16:9. I wish they still made 24" 4k monitors.
129.
by xcv123 2y ago
Where can you get a 4k 100 Hz monitor for $250?
130.
by xcv123 2y ago
Vast majority of 120+ Hz displays are 1080p only. That's for gaming. 4k at 120 Hz will need DisplayPort 2.0 and a powerful GPU to provide that bandwidth (40 Gbit/sec). Practically useless for its intended purpose (coding). Will in
131.
by xcv123 2y ago
"We talked through and I got to an understanding in a way having read manuals and textbooks didn’t really help before." You can't have a discussion with a book.
132.
by xcv123 2y ago
Relatively speaking. It's like the definition of a supercomputer 30 years ago is a cheap Android phone in your pocket today. You can certainly run a transformer model or any other neural network based model on an iPhone. Siri is probab
133.
by xcv123 2y ago
You can run a SLM (small language model) on an iPhone, and not much else. It would use up all of the resources. The iPhone only has 6 GB RAM.
134.
by xcv123 2y ago
Siri runs locally on the device. It works with internet completely disabled (airplane mode). You can't run any LLM on a phone.
135.
by xcv123 2y ago
Siri runs locally on the device. The cost would be prohibitive if they changed it to an LLM running in the cloud.
136.
by xcv123 2y ago
Just like learning from another human. A person can teach you the higher level concepts of some programming language but wouldn't remember the entire standard library.
137.
by xcv123 2y ago
That's a ~12 year construction project. 5.4 Billion USD per year is nothing these days. Only 0.006% of annual GDP for a project that has the potential to transform our civilization.
138.
by xcv123 2y ago
Their unpredictability makes it more terrifying to future whistleblowers.
139.
by xcv123 2y ago
That is missing the forest for the trees. Human cognition is more similar to a learned connectionist model versus an explicitly programmed rules-based model.
140.
by xcv123 2y ago
Get rid of whistleblowers before they blow the whistle? So they just start killing random employees who look suspicious?
141.
by xcv123 2y ago
Yes it's not my summary. I originally learned about the family tree example from a lecture by Geoffrey Hinton. Found some lecture slides here which reference the example but can't find the original lecture right now. https:/
142.
by xcv123 2y ago
> Plug that into the fact that it doesn't work to kill someone after they have testified This way deters future whistleblowers. If they got rid of whistleblowers before they blew the whistle, no one would hear about it. It would not
143.
by xcv123 2y ago
I'll refer you back to this 1990 paper by Geoffrey Hinton. Up to you if you want to investigate this further. If you can prove this is wrong then you should publish your result. https://www.cs.toronto.edu/~hinton/a
144.
by xcv123 2y ago
LLMs have already demonstrated their ability to generally apply abstract rules of logical reasoning to problems outside of their training set. Those are instances of reasoning. The models are learned from data, and they infer logical reason
145.
by xcv123 2y ago
That is a blatant oversimplification and not true. There are both similarities and differences. New ANN training methods are inspired by studies of biological neural networks (Dropout Regularization is one example)
146.
by xcv123 2y ago
Only if you define reasoning ability as exactly equivalent in capability to a formal theorem prover. But that is a difference in tribe or philosophy. Your Symbolic/Classical rule-based AI tribe versus the Connectionist AI tribe. No poi
147.
by xcv123 2y ago
> That's not really reasoning They are trained a certain way to perform specific types of tasks, primarily natural language processing tasks. They have necessarily learned some methods of reasoning in order to do what they were trai
148.
by xcv123 2y ago
> I think it's clear from this example that the LLM has 0 ability to reason. It's not a 0 or 1. You are oversimplifying it. Obviously neural networks can learn to generalize patterns of reasoning inferred from their training da
149.
by xcv123 2y ago
Yes we know the current LLMs cannot solve the original prompt. That's why I experimented with different prompts. The instructions are prompting it to proceed rigorously, as it is a logical problem, not a natural language problem. These
150.
by xcv123 2y ago
It's not spoiling anything. It's just an observation of the limits of current LLMs. I tried a few chain of thought prompts for the original question and GPT-3.5 was sometimes (randomly) able to find the correct answer on the first

More ›