HNHacker News
TopNewBestAskShowJobs

byby

63 karma · joined May 18, 2023

submissionscomments
byby··on Programming with natural language is going to work (2010)
>I think if you're an expert in an area,

Experts in the area, including Hinton, the father of modern AI, self admit they don't fully understand what's going on but they think that LLMs know what they are talking about.

>These systems are not "fully correct" and not "out of thin area". They are trained on everything ever digitised, including the entire internet. They, in effect, find similar historical cases to your query and merge them. In many cases, for specific enough queries, the text is verbatim from an original source.

I never said the systems are fully correct. I said that for my specific example the answer is fully correct and out of thin air. No such question and answer pair exists on the internet. Find it and prove me wrong.

>This will be clear to everyone pretty quickly, as with all tech, it's "magic" on the first encounter -- until the limitations are exposed.

Except many experts are saying the exact opposite of what you're saying. I'm just parroting the experts..

>I actually work in an area where what took 3mo last year, I can now do in maybe 3 days due to ChatGPT. But when it comes to providing my customers with that content, the value was always in how I provided it and what it did for them.

So if they knew you were just copying and pasting their queries to chatgpt would they still care about the "how"? I doubt it.

byby··on Programming with natural language is going to work (2010)
>The game was never writing the stack overflow answer -- that was already written.

The problem is this was never a stackoverflow question and there was never an answer for it.

Try finding it. The LLM is already playing the game because it came up with that answer which is Fully Correct, Out of Thin Air.

Look, clearly the LLM can't play the game as well as a trained adept human, but it's definitely playing the game.

>The game is designing software to requirements. It's writing literature for a new era. It's creating X for A audience with N vauge unspecified needs -- where X is a complex product made of many parts, involving many people, with shifting and changing problems/solutions/requirements.

It can do all of this. It can talk like you and parrot exactly what your saying and also go into more detail and re-frame your words more eloquently.

What you're not getting is that all the things you mentioned the LLM can do in actuality to varying degrees to the point where it is in the "game." and at times it does better than us. Likely, you haven't even tried asking it yet.

byby··on Programming with natural language is going to work (2010)
I'm just saying the tech is already here. The core engine can do it.

Before you go on and write such a system it's better to test if the LLM can do debugging to an efficacy level that we require. I don't think anyone has tried this yet and we do know LLMs have certain issues.

But make no mistake, the possibility that an LLM knows how to debug programs is actually quite high. If it can do this: https://www.engraved.blog/building-a-virtual-machine-inside/ it can likely debug a program, but I can't say definitively because I'm too lazy to try.

byby··on Programming with natural language is going to work (2010)
>Create a web site using language X with framework Y

Not sure if you played with chatGPT, but you can ALREADY ask it to do this and it will give you something that works 50% of the time.

byby··on Programming with natural language is going to work (2010)
I don't think you get it. The game has already changed. And nobody trained the AI to write unit tests or refactor code. It could do it as emergent behavior.

It happened so fast that so many people are in denial and many people aren't even asking the obvious questions. Simply ASK chatGPT to refactor your code to be more unit testable and IT Can do it.

Of course it clearly has your typical LLM problems but it is roughly 50% there. The game has changed. If AI gets to 100%, not only will the game be changed, but humans will no longer be part of the game.

See what I asked chatGPT:

   ME:
   def (x: int):
      for i in range(x):
         print(i)

   Can you refactor the code so it can be more unit testable?


   CHATGPT:
   Certainly! To make the code more unit testable, we can refactor it into a function that returns a list of integers instead of directly printing them. Here's the refactored code:

   python

   def generate_number_list(x: int) -> List[int]:
       number_list = []
       for i in range(x):
           number_list.append(i)
       return number_list

   In this refactored code, we've created a function called generate_number_list that takes an integer x as input. Instead of printing each number, we use a number_list variable to store the generated numbers by appending them in the loop. Finally, we return the number_list.

   By returning the list instead of printing the numbers directly, we make the code more testable because we can now write unit tests that assert the correctness of the generated list.
If boggles my mind that I see threads of people saying "what if we train LLMs to do this or to do that?" or whatever. Guys, just ask the LLM to do what you want in plain English. It can do it.
byby··on Programming with natural language is going to work (2010)
>Everyone's trying to train GPT models to write code, but maybe we should be training them how to use a debugger. Though its a lot harder to map text generation to debugging...

We can actually try this now. Literally tell the LLM what you want to do and work with it. See how far you can take it. You will of course be asking the LLM for debug line inputs and you will be providing it with outputs as you run the debugger yourself.

byby··on Elephants may be domesticating themselves
>If you read a histories any war/conflict where there are large numbers of horses involved, a common theme is them just dropping dead of exhaustion along the side of the trail, because they can't sustain the pace of the army marching with them.

This is assuming that most of the army were full of people walking and the horses walking along side were getting exhausted? So an average human can walk down a horse?

byby··on Why I Use Elm in 2023
Learned something new. I feel though it's better if most programmers didn't know about this. It looks like cursed knowledge.
byby··on Elephants may be domesticating themselves
I'm aware of this. We can run down a cheetah or bunny.

But we can't actually run down a horse. Only top ultra marathoners can do that.

But walking down a horse? Really?

byby··on Elephants may be domesticating themselves
This is a good start up idea in biotech. And highly lucrative. Genetically engineered tea cup elephants.
byby··on Elephants may be domesticating themselves
I've heard of the fact we have more endurance then the horse but walk one to death? Source? Or deeper explanation?
byby··on Elephants may be domesticating themselves
Yeah but this was definitely achieve well before 200 years. Armies existed since the dawn of civilization. We are talking B.C. dates.
byby··on Why I Use Elm in 2023
>I like exceptions. You can't get away from them, I know Haskell tried, Rust tried, they still have exceptions (perhaps they call them by a different name though)

Rust has no exceptions in any sense of the word. Exceptions are ways to crash the program in a controlled manner with the ability for these crashes to be caught midway before the program crashes and handled correctly.

In rust you can only crash it explicitly using the panic! Macro. No handling.

The different name you are looking for is "errors". Rust has errors as do all programs. Some programs have exceptions to handle those errors. Rust does not.

If you mean "runtime crashes" then elm is the only language that is popular that uses has no runtime crashes.

byby··on Legend of Zelda game sells 10M copies in three days
Not really.

Aiming and shooting is far more precise in red dead. Combat controls are instantaneous. You fire your gun, the bullet comes out and goes in a straight line towards an enemy on every click. You can headshot an entire gang of people no issue in red dead if you used PC controls.

Dark souls has delay. Not precision at all. Every movement is unresponsive and the point of those games is to get used to the delays and imprecision. You rapidly press hit 3 times in dark souls your character hits maybe once. In red dead that's 3 shots with the pistol.

Rdr2 movement outside of combat is built for immersion. As in the delays are variable based realistically off the environment. The combat however is by far significantly more precise then any dark souls game and as a result is actually the easier game.

byby··on KeyDB – A Multithreaded Fork of Redis
It makes sense for disk persistence to be io but redis is primarily in memory. So I don't think async applies to that aspect of it. In theory that makes it so all operations on the database from the user perspective are serial.

Not sure about this that's why I'm asking here for a definitive concrete answer about this.

byby··on KeyDB – A Multithreaded Fork of Redis
This is mainly because it's single threaded and operates on in memory state. Async io doesn't apply to in memory so all operations have to be serial, is this characterization correct? Thanks.
byby··on Study finds 90% of Australian teachers can't afford to live where they teach
No this isn't the problem. There is huge demand for skilled teachers as in teachers that are extremely good at their job. Huge demand. There's also too little supply. It Could be a lucrative career if the system were setup better.

Think about it. When you were in school how many good teachers were there? Very very very few.

Now think further... Any school you went to.. did you have the opportunity to choose the school based on how good the teachers are?

And therein lies the mismatch. Top schools aren't associated with good teachers. It's basically a crap shoot... you select the school based off of some prestige factor that's independent of teacher skill. Then within the school you have limited selection based off of informal "ratemyprofessor.com" reviews.

There needs to be official records and ways to measure teacher skill and hiring must be made off of these records. This will increase teacher pay. Until then teachers will be paid shit.

Imagine a university that did specifically this. You'd have a school where every teacher was freaking phenomenal. I, personally, went to a uni with a pretty high prestige factor on the west coast of the US and of course, ironically, the higher the prestige the shittier the teachers. I would drop all of that to attend the aforementioned hypothetical school.

And of course to stay in line with the topic because this hypothetical school would be more in demand the school would pay more. Teachers all over the industry would being to raise their own skill level and hiring would only be based on actual really good teachers rather then anybody who can speak English.

byby··on KeyDB – A Multithreaded Fork of Redis
Side question about redis:

Does redis do async operations? I'm not sure how that would work because it's known to be in memory. But I do know it persists to disk. So basically my question is:

Does almost absolutely every operation on a redis database happen serially? Maybe not every single operation, but in general.

byby··on Legend of Zelda game sells 10M copies in three days
> I was equally immersed 20 years ago playing Gothic and Oblivion as I am playing recent games, graphically speaking.

But you wouldn't be immersed on Gothic now. By playing gothic you set a new bar for yourself and that's how the industry works. Audiences everywhere get more and more sophisticated so much so that what worked back then won't work today and what works today won't work tomorrow. Can't wait until the time RDR2 becomes an "old" game with shitty graphics.

byby··on Legend of Zelda game sells 10M copies in three days
I bet you didn't like elden ring or any of the dark souls games either. In general the clunky controls are a style of game and promotes a certain type of game play.

It turns off people who want instantaneous responsiveness, but there's another style of action game genre your missing out on if you don't develop the patience to master those controls.

byby··on Legend of Zelda game sells 10M copies in three days
It is for most customers. They're not buying a movie, they're buying a movie player that only plays Disney movies.

For me and you we can likely buy every console but most of this rivalry stems from the fact that most people buy one console.

In this case the PC is the overall winning choice imo, but it's too expensive for a top of the line model. In that case the PS5 takes the next spot and the overall best choice when affordability and simplicity* are factored in.

(PC can play pirated games and emulate many past consoles leading to vast overall savings but it requires a bit more technical know how then most consumers are used to. Also many consumers are not into piracy, but it is an option here)

byby··on Legend of Zelda game sells 10M copies in three days
Yeah probably not. If you prefer really snappy controls and fluidity then games like RDR2 or elden ring or souls like games are not for you. That being said, in general, these games are not games that are generally considered to have bad controls.
byby··on Legend of Zelda game sells 10M copies in three days
>I'd have to think about it, but I might defend the NES, SNES, and Gamecube as the best consoles of their generations

NES yes, it's competitor was atari and atari had games that weren't that great. The NES launched games as mainstream, there was no competition at the time... it was the best.

SNES was also the best but it was only a slight edge over sega. The genesis was comparable.

Gamecube was not the winner. PS2 ruled this era. Starting with the gamecube and maybe you could say the N64... nintendo became more and more reliant on it's IP rather then a platform all developers wanted to work on.

>I'd put the N64 at a tie with the Playstation. The N64 looked better and had four-player support out of the box, which made a big difference in how many games supported four players.

No playstation wins this one hands down. N64 graphics had blurry textures and lack of digital media. Squaresoft switched their entire nintendo led RPG line (Final Fantasy 7) to the playstation because nintendo hardware couldn't handle it. Nintendo came out with some heavy hitters but overall in terms of game selection Sony won.

>It's only really with the Wii that they stopped credibly trying to keep up (and IMO the Wii's various gimmicks don't make up for its deficiencies, among which I count some of those same gimmicks)

I wouldn't say they stopped trying to keep up. Nintendo took a risk, and in the end for games that are less casual they sort of lost the market. But in the end it was the right choice in terms of sales.

byby··on Legend of Zelda game sells 10M copies in three days
Hmm maybe it's a bad term. I was trying to convey the fact that gamers that enjoy something like fallout or Diablo or Mass effect or RDR2, those types of gamers lose out when they lock themselves onto nintendo.

I for one like every genre but I find it unfortunate nintendo has a wierd focus on casual age friendly games.

byby··on Legend of Zelda game sells 10M copies in three days
So did I, I bought all the consoles and the wii system is the most focused in terms of genre. Very casual and age friendly.
byby··on Legend of Zelda game sells 10M copies in three days
Metroid is one of the least casuals but that's as casual as nintendo will get. Simple stories no blood, no dark hard themes like in RDR2 or dead space.

>What do God of War, RDR2, and Dead Space have in common that makes them not casual? I'm really curious about what makes a game casual or not.

Things that will make a movie casual. That's what I meant. Not casual in the sense that tetris is casual. Casual in the sense that a disney movie is casual while inception or Evil Dead is not casual.

Here's a better way to put it, thematically casual and age friendly is what nintendo is all about.

byby··on Cargo Cult AI
We hit a clock limit ages ago and switched to parallelism.

I still remember the days there things were in mhz and we were going 90, 100, 200, 300, 500, 1ghz, 2, 3 ... and we'd be at hundreds of ghzs if we didn't hit a limit. Now we switched to cores.

byby··on Cargo Cult AI
>"Emergent abilities" in that paper is an extraordinarily poor communication of the idea that larger models can do more than smaller models, which should be a surprise to no one.

Nothing on the face of the earth can be a sum of more than their parts. It's called conservation of mass and conservation of energy. So based off of your logic how can the term "emergent ability" even exist? Why do we even use the term?

You're getting hung up on a language quirk. Abilities aren't being built out of thin air, the term "emergent ability" is just an expression for "unexpected abilities". Even you as a human are a sum of it's parts.

There's no nonsense being spread here. Experts and eminent researchers including the father of modern AI (Hinton) ALL use the term "emergent abilities". Are you an expert? Why is your opinion on this "spreading of nonsense" and why is it directly contradictory to expert opinions?

Clearly you have logic that these experts and researchers haven't thought about. Can you spell it out step by step exactly what makes you more utterly clear about the "nonsense" then the experts who in actuality claim they don't fully understand what's going on?

I'm not being sarcastic here, there is a huge contingent of people on HN who are totally dismissing LLMs and it would be a disservice not for them and you to not allow them to spell out their logic clearly. But what I'm seeing in your paragraph is just an investigation on vocabulary on what is meant by "emergent abilities"

byby··on Cargo Cult AI
Except this doesn't make any sense. Moores law is beginning to fail size has a physical limit in that we can't get much smaller than an atom.

The potential of AI on the other hand is clear. Look at your self. If someone as intelligent as a human can be physically realized then the possibility of building something as intelligent as a human exists by simple logic.

It's easy to see the convoluted argument on your end. Why bet on something that is physically impossible to continue when you can bet on AI that is not only theoretically possible, but walking and talking versions of intelligence are all around us everyday.

byby··on Legend of Zelda game sells 10M copies in three days
I played on PC. Mouse and keyboard. The controls were excellent.
← PreviousPage 2 of 3Next →