inimino

12 ms·

151.
by inimino 4y ago
You can just read what Dijkstra said about it already 30 years ago, not much has changed, and I would have little to add.
152.
by inimino 4y ago
Software engineering is actually not engineering either.
153.
by inimino 4y ago
There's never been a camp of computer science that said anything but the truth, which is that CS is applied mathematics. However, there is a pragmatic school of hacking, which says that results are all the matters. If you're in a
154.
by inimino 4y ago
I try to base my judgments of what LLMs can and can't do primarily on my study and research in related fields. I haven't been surprised by the capabilities of any LLM yet, including GPT-4.
155.
by inimino 4y ago
Or you could wait until you have 5 flagged submissions from the same student, and then, I don't know, apply human scrutiny and your own good judgment?
156.
by inimino 4y ago
> flat and stale imagery Because it is a text predictor, with the choices being made by a PRNG, which is not the same thing as an author, who starts with something to say and chooses the words to fit. The LLM is like the author with the
157.
by inimino 4y ago
> As far as I know there is no proof that you can't cause a LLM to crash with user input. Well, then you simply don't understand how they work.
158.
by inimino 4y ago
Eventually it will be common knowledge that asking a text compressor trained on all available text to design a new technology doesn't work.
159.
by inimino 4y ago
Which is a finite state machine attached to an infinite tape.
160.
by inimino 4y ago
The state space of a Markov process must be fixed but is otherwise unspecified, which the LLM satisfies, so you are technically incorrect.
161.
by inimino 4y ago
Tell that to Alan Turing.
162.
by inimino 4y ago
Answered in detail in the FAQ: http://prize.hutter1.net/hfaq.htm#lossless
163.
by inimino 4y ago
Nobody is rejecting rational thought. Most of your values are innate and most of the rest is socialization. This is not an "internet meme", it's the consensus view since approximately the time of Plato.
164.
by inimino 4y ago
Universal human rights are not rational thought, they are values. You don't think your way into your values.
165.
by inimino 4y ago
I'm not sure if you're aware of the amount of text used to train these large language models, but in that context, yes it is miniscule.
166.
by inimino 4y ago
It's a real struggle for me to take anyone seriously who uses "learnings" unironically.
167.
by inimino 4y ago
Since this gets repeated so often, it is probably worth reproducing a little more of Knuth's quote. "There is no doubt that the grail of efficiency leads to abuse. Programmers waste enormous amounts of time thinking about, or worr
168.
by inimino 4y ago
The psi in the tire is set by the amount of air inside and the volume. It does not change significantly if the car is picked up (as you can see by the shape of the tire staying nearly the same).
169.
by inimino 4y ago
The helpful advice is to just try to match your opponent's speed. This is 10 times more useful than the OP advice based on some formula. Chess is about the ebb and flow of ideas between you and the opponent. If you're not calculat
170.
by inimino 4y ago
*pedant. ;-)
171.
by inimino 4y ago
I'm not sure why you started with "This is wrong" and then went on to talk about a completely different game? In any case, I'm not expert in MOBA AI, but they are much more complicated than go just as go is to chess. So
172.
by inimino 4y ago
Humans have two ways to pick a move: look at the board and pick a move based on the position (pattern matching, intuition, etc) and calculation (tree search). Machines can already do both better. That is my evidence. There isn't any ro
173.
by inimino 4y ago
There are literally hundreds of better books for newcomers. As just one, I would mention Programming Pearls by Jon Bentley. I would say it communicates a much more healthy approach to the craft.
174.
by inimino 4y ago
I agree. Martin's approach is amenable to bureaucracies that want to increase (the perception of) organizational control over software projects. It is opposed to the Fred Brooks approach. It favors a lower standard of individual skill
175.
by inimino 4y ago
In chess, it is possible that humans and bots can both someday play perfectly, and every game will be a draw. (Young people might play chess until learning how to draw, and then move on.) For Go there is no such possible outcome, and deep r
176.
by inimino 4y ago
I appreciate the sentiment, but it's nothing that dramatic. The battle remains over. This just shows that the training process remains imperfect. It's a bit like finding a bug.
177.
by inimino 4y ago
You're still trying to prove your position. Look, you're all over this thread misunderstanding LLMs and rejecting the relatively correct explanations people are giving you. The comment by joe_the_user upthread that you called an o
178.
by inimino 4y ago
Then don't. Instead of breathlessly trying to prove your theory, try and do some science by falsifying it. (cf. Wason test) Think of things it would get right only if it truly understood, not "common questions on reddit".
179.
by inimino 4y ago
We're on hacker news. Getting systems to do things they were designed not to do (in this case, for science) is in the name. If you understand how LLMs work then you understand immediately that this is a harmless joke.
180.
by inimino 4y ago
You understand that getting the LLM to actually pretend to run the command doesn't actually delete any data, right?

More ›