Your analogy is poor, and you are missing a very important fact.
Most of the human written code, in places where that code needs to make money, is decidable either entirely or in large parts. I.e without running the code, you can take a domain of inputs and build a complete range of outputs solely by looking at the code.
The way that works in your head is that you are effectively doing a compilation to a logical like structure, which then you can use to infer what the output will be from what the input is, and its a direct mapping that is invertible and separable, so if you know what the output should be, you know what the input is, you can pinpoint the exact location where it breaks. Thats how humans write code.
If thats not clear, imagine a piece of code that splits strings by spaces, deletes the empty strings, and returns the number of words in a string. The fact that you can say that if you want 3 words, there should be maximum 2 sequences of continous spaces between words, is you effectively transpiling that program into a latent space inside your brain neurons and inverting it.
LLMs essentially do this, with the added advantage of having been trained on a HUGE number of codebases, so they can recognize patterns that a human cant.
Where LLMs struggle is complex behavior - they can't simulate things like a human can and choose the best course of action. Even harnesses for agentic loops that can auto run and debug code can't match what a human can do in this regard (hence why self driving still sucks rn).
So moving forward, being a good coder isn't going to be about writing code, or even about prompting LLMs. Its going to be all about whether or not you can design good custom agentic loops, which necessarily involves knowledge of the model at hand (i.e what words you have to use to get it to do the right thing). This will be especially true as investment into "private" inference grows where companies will be using smaller models that have less detailed RL and thus will need much more guidance to do the right thing.
I disagree, to keep it short, what you're describing is not understanding, it's superstition. And I think it's a wrong direction of engineering, relying on some sort of irreproducible expert intuition, one that has been successfully replaced by enlightenment and scientific method.
There are 3 major obstacles in understanding LLMs:
1. They use inscrutable internal language of embeddings
2. They communicate in natural language which is itself ambiguous
3. The weights and training inputs are being hidden as a "trade secret"
"with the added advantage of having been trained on a HUGE number of codebases"
This doesn't really mean much unless we understand what is the quality and relevance of these sources for the problem at hand. Without this understanding it's just a superstition.
The difference is, the total cost for a human to write/modify code files is much higher than for an LLM. A lot of time has been spent designing program languages so that human time is more optimized. LLMS dgaf if the code is structured cleanly or is a mess.
So even if you end up with a mess of a codebase, as long as you define your test cases and they all pass, what is in the middle doesn't really matter.
Im fully on the spectrum, but my first thought when I hear about stuff like this is what would be the moral implication of "recreational gang war" experience. I.e you pay a lot of money, you get guns and armor, you go into gang territory, they shoot arrows at you, you defend yourself. Seems like you would have the right to use deadly force, but you would obviously be at a big advantage.
There is no moral or philosophical debate to be had here. It's morally wrong to do this and in my opinion that's clear as day. Putting yourself in a dangerous situation and provoking a "self-defense" scenario like that only serves as a thinly veiled excuse to murder people you consider inferior.
Do you often have violent power fantasies? Do you find yourself daydreaming about them regularly?
Spare me the psycho analysis please, its pretty cringe.
Explain to me how is this different from safari hunting trips where you can get attacked by a wild animal (often an agressive one thatvattacks villagers) and have to shoot it to defend yourself? And the argument of them being human versus safari trip featuring an animal is not something that is given, you have to show that their lives are worth more than the animal life
There being a market for something doesn't make it morally right, but it highlights the demand aspect where someone is willing to pay money, which means that there is some level of it not being universally bad.
In the case of a Safari, there is this moral middle ground - according to the countries policy, if an animal attacks a human, it generally is a candidate for being put down because of perceived danger to humans in the area, but on the flip side, its also morally wrong to kill an animal just because it does that.
In the case of the gang warfare, you are doing a service by reducing the number of gang members and crime, but its also morally wrong to take someones life in offense for personal entertainment? (If its not clear, the difference in doing that in PNG versus any other country is because specifically of the lack of law enforcement that is the way of life there)
Another hypothetical scenario - say you have a country where age of consent is low, and certain type of porn that is highly illegal in the rest of the world is freely available (which is not far from reality, scarily). You can't stop the creation or distribution of that porn without basically invading the country and changing its law. However, what if you used said porn and supplied to convicted pdf files in your home country so that they are less likely to exploit young kids irl? Is that morally wrong, or is it morally justified?
Did you know there have been literal slave markets in the past? And they still exist today (e.g. in Lybia). Please explain how this is "not universally bad". Actually don't. Spare us your disgusting mental gymnastics that you shamelessly try to disguise as autism.
Not that you've explicitly claimed it's authentic, but the very idea of a pirate hunting tour is a hoax that resurfaces across the Internet every so often in different forms:
Not at all. But the world we live in is pretty fucked up from that perspective. There are ex military contractor guys on youtube talking about their weapons and how they used a certain rifle to fight off invaders in the middle east. Nobody seems to care that the people on the other end of those bullets may have families.
My interest in this is purely philosophical. If you have a what is effectively a lawless country, where native populace has the desire to kill you, to what extent could you engage with that in a moral framework?
but do you have any evidence beyond intuition that those denials differ significantly in reversal rate? one would guess that things that are never going to be covered get denied and then never appealed but one would also guess that things that are never going to be covered don't get submitted in the first place. assuming that the rate of legitimate denials is higher among those claims never appealed, what rate of illegitimate denials is acceptable? why would denying even one person care that they're contractually entitled to receive be okay? why shouldn't insurance companies have to do what they promise to do and accept money for?
CAD is not just drawing. CAD packages are often integrated with other software that does things like stress analysis, BOM, thermal simulation, and so on.
Its an entire ecosystem that streamlines a lot of things, so it makes sense to invest in it.
The reason for Windowsis because a lot of the CAD software utilizes gfx drivers heavily for rendering. You can do things like raytracing to simulate how the product will look.
- In the winter, if there is sun, and you have like black leather interior, it actually helps warm the car up.
-For trucks, having a glass roof is almost essential for seeing if you will fit into a parking garage. Having lived in Seattle, Im glad my truck had one.
- Opening the roof and setting AC to use outside air is the fastest way to vent the car. It doesn't create as many vorticies as opening the windows.
For trucks, having a glass roof is almost essential for seeing if you will fit into a parking garage.
If your truck is that tall, I suggest measuring it and recording the height. After that, you'll find that just about every low-hanging garage, drive-up, bridge, etc. has a height listed. Over 9.5'? Don't try and drive through the Arboretum, as the signs will tell you.
Opening the roof...
That's a sunroof, not a panoramic glass roof being discussed.
Pedantic, but more likely to be a moonroof. A sunroof is opaque, while a moonroof is the clear or mild-to-moderately tinted glass panel we've all grown used to.
I currently live in Seattle and drive a 4WD Toyota Tacoma with a rooftop tent on the back and frequently enter parking garages. I know the height of my truck and was careful to get a setup where the rooftop tent isn't gratuitously high. I think everything is still a little under 6' which is fine in just about every parking garage.
Even if my setup was taller, garages here are pretty good at having the height listed when you enter and anyone with a tall truck really should know exactly how tall it is.
I don't think a sunroof would help me. If I was ever in a situation where it did, I would already be taking too much of a risk of damaging something.
> Turns out that all of the things you listed as examples are stuff that most consumers don’t actually want or will use.
The thing that nobody really realized is that setup is a big factor that isn't worth to a lot of people. For example, to turn off lights with Alexa, you have to do a whole bunch of setup that involves installing special lights, installing app, connecting the app to Alexa. Even though relatively this is simple, its just not worth the effort to a lot of people.
Amazon or Apple could easily overcome this by offering Alexa or Siri connected devices as incentive to builders building houses or apartments. That way, when you plug in Alexa in your new apartment, it can auto detect everything it can control including lights, thermostat appliances, e.t.c.
There are endless posts on HN about how Macs are good for LLMs. Somoene is running some model on their Mac for local inference.
What the posts dont mention is how unusable that experience is with dogshit slow tokens/second. To make a local model usefull you need to run the highest parameter models at 100+ tok/sec, otherwise you are just better off paying for cloud inference. So either all those people are dumb as hell, or Apple is doing clever advertising. And I personally have more faith in the tech sector.
Most of the human written code, in places where that code needs to make money, is decidable either entirely or in large parts. I.e without running the code, you can take a domain of inputs and build a complete range of outputs solely by looking at the code.
The way that works in your head is that you are effectively doing a compilation to a logical like structure, which then you can use to infer what the output will be from what the input is, and its a direct mapping that is invertible and separable, so if you know what the output should be, you know what the input is, you can pinpoint the exact location where it breaks. Thats how humans write code.
If thats not clear, imagine a piece of code that splits strings by spaces, deletes the empty strings, and returns the number of words in a string. The fact that you can say that if you want 3 words, there should be maximum 2 sequences of continous spaces between words, is you effectively transpiling that program into a latent space inside your brain neurons and inverting it.
LLMs essentially do this, with the added advantage of having been trained on a HUGE number of codebases, so they can recognize patterns that a human cant.
Where LLMs struggle is complex behavior - they can't simulate things like a human can and choose the best course of action. Even harnesses for agentic loops that can auto run and debug code can't match what a human can do in this regard (hence why self driving still sucks rn).
So moving forward, being a good coder isn't going to be about writing code, or even about prompting LLMs. Its going to be all about whether or not you can design good custom agentic loops, which necessarily involves knowledge of the model at hand (i.e what words you have to use to get it to do the right thing). This will be especially true as investment into "private" inference grows where companies will be using smaller models that have less detailed RL and thus will need much more guidance to do the right thing.
reply